proposed methodology climate change

Advertisement

Strategies for mitigation of climate change: a review

• Open access
• Published: 30 July 2020
• Volume 18 , pages 2069–2094, ( 2020 )

Cite this article

You have full access to this open access article

• Samer Fawzy 1 ,
• Ahmed I. Osman   ORCID: orcid.org/0000-0003-2788-7839 1 ,
• John Doran 2 &
• David W. Rooney 1  

222k Accesses

626 Citations

66 Altmetric

Explore all metrics

Climate change is defined as the shift in climate patterns mainly caused by greenhouse gas emissions from natural systems and human activities. So far, anthropogenic activities have caused about 1.0 °C of global warming above the pre-industrial level and this is likely to reach 1.5 °C between 2030 and 2052 if the current emission rates persist. In 2018, the world encountered 315 cases of natural disasters which are mainly related to the climate. Approximately 68.5 million people were affected, and economic losses amounted to $131.7 billion, of which storms, floods, wildfires and droughts accounted for approximately 93%. Economic losses attributed to wildfires in 2018 alone are almost equal to the collective losses from wildfires incurred over the past decade, which is quite alarming. Furthermore, food, water, health, ecosystem, human habitat and infrastructure have been identified as the most vulnerable sectors under climate attack. In 2015, the Paris agreement was introduced with the main objective of limiting global temperature increase to 2 °C by 2100 and pursuing efforts to limit the increase to 1.5 °C. This article reviews the main strategies for climate change abatement, namely conventional mitigation, negative emissions and radiative forcing geoengineering. Conventional mitigation technologies focus on reducing fossil-based CO 2 emissions. Negative emissions technologies are aiming to capture and sequester atmospheric carbon to reduce carbon dioxide levels. Finally, geoengineering techniques of radiative forcing alter the earth’s radiative energy budget to stabilize or reduce global temperatures. It is evident that conventional mitigation efforts alone are not sufficient to meet the targets stipulated by the Paris agreement; therefore, the utilization of alternative routes appears inevitable. While various technologies presented may still be at an early stage of development, biogenic-based sequestration techniques are to a certain extent mature and can be deployed immediately.

Similar content being viewed by others

Strategies for Combating Climate Change

The Politics and Governance of Negative Emissions Technologies

Anthropogenic Climate Change and Countermeasures: Chances and Risks of Weather Modification Techniques and Climate Engineering (CE)

Explore related subjects.

• Environmental Chemistry

Avoid common mistakes on your manuscript.

Introduction

Status of climate change.

Climate change is defined as the shift in climate patterns mainly caused by greenhouse gas emissions. Greenhouse gas emissions cause heat to be trapped by the earth’s atmosphere, and this has been the main driving force behind global warming. The main sources of such emissions are natural systems and human activities. Natural systems include forest fires, earthquakes, oceans, permafrost, wetlands, mud volcanoes and volcanoes (Yue and Gao 2018 ), while human activities are predominantly related to energy production, industrial activities and those related to forestry, land use and land-use change (Edenhofer et al. 2014 ). Yue and Gao statistically analysed global greenhouse gas emissions from natural systems and anthropogenic activities and concluded that the earth’s natural system can be considered as self-balancing and that anthropogenic emissions add extra pressure to the earth system (Yue and Gao 2018 ).

GHG emissions overview

The greenhouse gases widely discussed in the literature and defined by the Kyoto protocol are carbon dioxide (CO 2 ), methane (CH 4 ), nitrous oxide (N 2 O), and the fluorinated gases such as hydrofluorocarbons (HFCs), perfluorocarbons (PFCs) and sulphur hexafluoride (SF 6 ) (UNFCCC 2008 ). According to the emissions gap report prepared by the United Nations Environment Programme (UNEP) in 2019, total greenhouse gas emissions in 2018 amounted to 55.3 GtCO 2 e, of which 37.5 GtCO 2 are attributed to fossil CO 2 emissions from energy production and industrial activities. An increase of 2% in 2018 is noted, as compared to an annual increase of 1.5% over the past decade for both total global greenhouse gas and fossil CO 2 emissions. The rise of fossil CO 2 emissions in 2018 is mainly driven by higher energy demand. Furthermore, emissions related to land-use change amounted to 3.5 GtCO 2 in 2018 (UNEP 2019 ). Together in 2018, fossil-based and land-use-related CO 2 emissions accounted for approximately 74% of the total global greenhouse gas emissions. Methane (CH 4 ), another significant greenhouse gas, had an emission rate increase of 1.7% in 2018 as compared to an annual increase of 1.3% over the past decade. Nitrous oxide (N 2 O) emissions, which are mainly influenced by agricultural and industrial activities, saw an increase of 0.8% in 2018 as compared to a 1% annual increase over the past decade. A significant increase was, however, noted in the fluorinated gases during 2018 at 6.1% as compared to a 4.6% annual increase over the past decade (UNEP 2019 ). To put these numbers into perspective, a recent Intergovernmental Panel on Climate Change (IPCC) report demonstrated that anthropogenic activities so far have caused an estimated 1.0 °C of global warming above the pre-industrial level, specifying a likely range between 0.8 and 1.2 °C. It is stated that global warming is likely to reach 1.5 °C between 2030 and 2052 if the current emission rates persist (IPCC 2018 ).

Climate change impacts, risks and vulnerabilities

An understanding of the severe impact of climate change on natural and human systems as well as the risks and associated vulnerabilities is an important starting point in comprehending the current state of climate emergency. Changes in climate indicators, namely temperature, precipitation, seal-level rise, ocean acidification and extreme weather conditions have been highlighted in a recent report by the United Nations Climate Change Secretariat (UNCCS). Climate hazards reported included droughts, floods, hurricanes, severe storms, heatwaves, wildfires, cold spells and landslides (UNCCS 2019 ). According to the Centre for Research on the Epidemiology of Disasters (CRED), the world encountered 315 cases of natural disasters in 2018, mainly climate-related. This included 16 cases of drought, 26 cases of extreme temperature, 127 cases of flooding, 13 cases of landslides, 95 cases of storms and 10 cases of wildfire. The number of people affected by natural disasters in 2018 was 68.5 million, with floods, storms and droughts accounting for 94% of total affected people. In terms of economic losses, a total of $131.7 billion was lost in 2018 due to natural disasters, with storms ($70.8B), floods ($19.7B), wildfires ($22.8B) and droughts ($9.7B) accounting for approximately 93% of the total costs. CRED also provides data on disasters over the past decade, which shows even higher annual averages in almost all areas, except for wildfire cases. The economic losses attributed to wildfires in 2018 alone are approximately equal to the collective losses from wildfires incurred over the past decade, which is quite alarming (CRED 2019 ). Moreover, wildfires are a direct source of CO 2 emissions. Although wildfires are part of the natural system, it is clear that human-induced emissions are directly interfering and amplifying the impact of natural system emissions. It is evident that human-induced climate change is a major driving force behind many natural disasters occurring globally.

Furthermore, climate risks such as temperature shifts, precipitation variability, changing seasonal patterns, changes in disease distribution, desertification, ocean-related impacts and soil and coastal degradation contribute to vulnerability across multiple sectors in many countries (UNCCS 2019 ). Sarkodie et al. empirically examined climate change vulnerability and adaptation readiness of 192 United Nations countries and concluded that food, water, health, ecosystem, human habitat and infrastructure are the most vulnerable sectors under climate attack while pointing out that Africa is the most vulnerable region to climate variability (Sarkodie and Strezov 2019 ). It is also important to note the interconnected nature of such sectors and the associated impacts.

The 15 th edition of the global risks report 2020 prepared by the world economic forum thoroughly presented a number of climate realities, laying out areas that are greatly affected. The risks included loss of life due to health hazards and natural disasters, as well as excessive stress on ecosystems, especially aquatic/marine systems. Moreover, food and water security are other areas that are highly impacted. Increased migration is anticipated due to extreme weather conditions and disasters as well as rising sea levels. Geopolitical tensions and conflicts are likely to arise as countries aim to extract resources along water and land boundaries. The report also discusses the negative financial impact on capital markets as systematic risks soar. Finally, the impact on trade and supply chains is presented (WEF 2020 ).

An assessment, recently presented in an Intergovernmental Panel on Climate Change (IPCC) special report, covered the impacts and projected risks associated with 2 levels of global warming, 1.5 °C and 2 °C. The report investigated the negative impact of global warming on freshwater sources, food security and food production systems, ecosystems, human health, urbanization as well as poverty and changing structures of communities. The report also investigated climate change impact on key economic sectors such as tourism, energy and transportation. It is evident that most of the impacts assessed have lower associated risks at 1.5 °C compared to 2 °C warming level. We would likely reach 1.5 °C within the next 3 decades and increases in warming levels beyond this point would amplify risk effects; for example, water stress would carry double the risk under a 2 °C level compared to 1.5 °C. An increase of 70% in population affected by fluvial floods is projected under the 2 °C scenario compared to 1.5 °C, especially in USA, Europe and Asia. Double or triple rates of species extinction in terrestrial ecosystems are projected under the 2 °C level compared to 1.5 °C (IPCC 2018 ). It can be simply concluded that the world is in a current state of climate emergency.

Global climate action

Acknowledgement of climate change realities started in 1979 when the first world climate conference was held in Geneva. The world climate conference was introduced by the World Meteorological Organization in response to the observation of climatic events over the previous decade. The main purpose was to invite technical and scientific experts to review the latest knowledge on climate change and variability caused by natural and human systems as well as assess future impacts and risks to formulate recommendations moving forward (WMO 1979 ). This was possibly the first of its kind conference discussing the adverse effects of climate change. In 1988, the Intergovernmental Panel on Climate Change (IPCC) was set up by the World Meteorological Organization in collaboration with the United Nations Environment Programme (UNEP) to provide governments and official bodies with scientific knowledge and information that can be used to formulate climate-related policies (IPCC 2013 ).

Perhaps, the most critical step taken, in terms of action, was the adoption of the United Nations Framework Convention on Climate Change (UNFCCC) in 1992, which then went into force in 1994. Since then, the UNFCCC has been the main driving force and facilitator of climate action globally. The main objective of the convention is the stabilization of greenhouse gas concentrations in the atmosphere to prevent severe impacts on the climate system. The convention set out the commitments to all parties involved, putting major responsibilities on developed countries to implement national policies to limit anthropogenic emissions and enhance greenhouse gas sinks. The target was to reduce emissions by the year 2000 to the levels achieved in the previous decade. Moreover, committing developed country parties to assist vulnerable developing country parties financially and technologically in taking climate action. The convention established the structure, reporting requirements and mechanism for financial resources, fundamentally setting the scene for global climate policy (UN 1992 ). The convention is currently ratified by 197 countries (UNCCS 2019 ).

During the third UNFCCC conference of the parties (COP-3) in 1997, the Kyoto protocol was adopted and went into force in 2005. The Kyoto protocol introduced the emission reduction commitments for developed countries for a five-year commitment period between 2008 and 2012. The protocol laid out all related policies, monitoring and reporting systems, as well as introduced three market-based mechanisms to achieve those targets. The protocol introduced two project-based mechanisms, clean development mechanism and joint implementation mechanism. The clean development mechanism allows developed country parties to invest and develop emission reduction projects in developing countries, to drive sustainable development in the host country as well as offset carbon emissions of the investing party. Joint implementation projects allow developed country parties to develop similar projects, however, in other developed countries that are protocol parties, offsetting excess emissions of the investing party. Furthermore, the protocol introduced an emissions trading mechanism as a platform to facilitate the trading of annually assigned emissions that are saved by protocol members to those that exceed their limits (UNFCCC 1997 ). Emission reduction has mainly been achieved through the introduction of renewable energy, energy efficiency and afforestation/reforestation-related projects.

The Kyoto protocol defines four emission saving units, each representing one metric ton of CO 2 equivalent and are all tradeable (UNFCCC 2005 ).

Certified emissions reduction unit, obtained through clean development mechanism projects.

Emission reduction unit, obtained through joint implementation projects.

Assigned amount unit, obtained through the trading of unused assigned emissions between protocol parties.

Removal unit, obtained through reforestation-related projects.

The Kyoto units and general framework introduced laid the structural foundation of a carbon emissions market and the concept of carbon pricing. Many national and regional governments introduced emissions trading schemes; some are mandatory while others are voluntary. In some cases, such schemes are linked to Kyoto commitments and regulations. The largest emissions trading scheme introduced thus far is the European emissions trading scheme (Perdan and Azapagic 2011 ). Villoria-Saez et al. empirically investigated the effectiveness of greenhouse gas emissions trading scheme implementation on actual emission reductions covering six major emitting regions. The investigation presented a number of findings; first, it is possible to reduce greenhouse gas emissions by approximately 1.58% annually upon scheme implementation. Furthermore, after 10 years of implementation, approximately 23.43% of emissions reduction can be achieved in comparison with a scenario of non-implementation (Villoria-Sáez et al. 2016 ). Another emission abatement instrument widely discussed in the literature is carbon taxation. There is growing scientific evidence that carbon taxation is an effective instrument in reducing greenhouse gas emissions; however, political opposition by the public and industry is the main reason delaying many countries in adopting such mechanism (Wang et al. 2016 ).

In 2012, the Doha amendment to the Kyoto protocol was adopted, mainly proposing a second commitment period from 2013 to 2020 as well as updating emissions reduction targets. The amendment proposed a greenhouse gas emissions reduction target of at least 18% below 1990 levels. The amendment has not yet entered into force since it has not been ratified by the minimum number of parties required to this date (UNFCCC 2012 ).

During the twenty-first UNFCCC conference of the parties (COP-21) held in Paris in 2015, the Paris agreement was adopted and entered into force in 2016. The Paris agreement added further objectives, commitments, enhanced compliance and reporting regulations, as well as support mechanisms to the existing climate change combat framework in place. The main objective of the agreement is to limit the global temperature increase to 2 °C by 2100 and pursue efforts to limit the increase to 1.5 °C. The agreement aims to reach global peaking of greenhouse gases as soon as possible as to strike a balance between human-induced emission sources and greenhouse gas sinks and reservoirs between 2050 and 2100. The agreement also introduced new binding commitments, asking all parties to deliver nationally determined contributions and to enforce national measures to achieve, and attempt to exceed such commitments. Enhanced transparency, compliance and clear reporting and communication are advocated under the agreement. Furthermore, the agreement encourages voluntary cooperation between parties beyond mandated initiatives. Moreover, financial support and technological support, as well as capacity building initiatives for developing countries, are mandated by the agreement. Such obligations are to be undertaken by developed country parties to promote sustainable development and establish adequate mitigation and adaptation support measures within vulnerable countries. Perhaps, one of the most important goals established under the agreement is that of adaptation and adaptive capacity building concerning the temperature goal set (UN 2015 ).

Under article 6 of the agreement, two international market mechanisms were introduced, cooperative approaches and the sustainable development mechanism. These mechanisms are to be utilized by all parties to meet their nationally determined contributions. Cooperative approaches are a framework that allows parties to utilize internationally transferred mitigation outcomes (ITMOs) to meet nationally determined contribution goals as well as stimulate sustainable development. On the other hand, the sustainable development mechanism is a new approach that promotes mitigation and sustainable development and is perceived as the successor of the clean development mechanism. There is still much debate and negotiations on such mechanisms moving forward (Gao et al. 2019 ).

Nieto et al. conducted an in-depth systematic analysis of the effectiveness of the Paris agreement policies through the evaluation of 161 intended nationally determined contributions (INDCs) representing 188 countries. The study investigated sectoral policies in each of these countries and quantified emissions under such INDCs. The analysis concluded that a best-case scenario would be an annual global emission increase of approximately 19.3% in 2030 compared to the base period (2005–2015). In comparison, if no measures were taken a 31.5% increase in global emissions is projected. It is concluded that if the predicted best-case level of emissions is maintained between 2030 and 2050 a temperature increase of at least 3 °C would be realized. Furthermore, a 4 °C increase would be assured if annual emissions continue to increase (Nieto et al. 2018 ).

To meet the 1.5 °C target by the end of the century, the IPCC stated that by 2030 greenhouse gas emissions should be maintained at 25–30 GtCO 2 e year −1 . In comparison, the current unconditional nationally determined contributions for 2030 are estimated at 52–58 GtCO 2 e year −1 . Based on pathway modelling for a 1.5 °C warming scenario, a 45% decline in anthropogenic greenhouse gas emissions must be reached by 2030 as compared to 2010 levels, and net-zero emissions must be achieved by 2050. To maintain a 2 °C global warming level by the end of the century, emissions should decline by approximately 25% in 2030 as compared to 2010 levels and net-zero emissions should be achieved by 2070 (IPCC 2018 ). There is growing evidence that confirms that current mitigation efforts, as well as future emissions commitments, are not sufficient to achieve the temperature goals set by the Paris agreement (Nieto et al. 2018 ; Lawrence et al. 2018 ). Further measures and new abatement routes must be explored if an attempt is to be made to achieve such goals.

Climate change mitigation strategies

There are three main climate change mitigation approaches discussed throughout the literature. First, conventional mitigation efforts employ decarbonization technologies and techniques that reduce CO 2 emissions, such as renewable energy, fuel switching, efficiency gains, nuclear power, and carbon capture storage and utilization. Most of these technologies are well established and carry an acceptable level of managed risk (Ricke et al. 2017 ; Victor et al. 2018 ; Bataille et al. 2018 ; Mathy et al. 2018 ; Shinnar and Citro 2008 ; Bustreo et al. 2019 ).

A second route constitutes a new set of technologies and methods that have been recently proposed. These techniques are potentially deployed to capture and sequester CO 2 from the atmosphere and are termed negative emissions technologies, also referred to as carbon dioxide removal methods (Ricke et al. 2017 ). The main negative emissions techniques widely discussed in the literature include bioenergy carbon capture and storage, biochar, enhanced weathering, direct air carbon capture and storage, ocean fertilization, ocean alkalinity enhancement, soil carbon sequestration, afforestation and reforestation, wetland construction and restoration, as well as alternative negative emissions utilization and storage methods such as mineral carbonation and using biomass in construction (Lawrence et al. 2018 ; Palmer 2019 ; McLaren 2012 ; Yan et al. 2019 ; McGlashan et al. 2012 ; Goglio et al. 2020 ; Lin 2019 ; Pires 2019 ; RoyalSociety 2018 ; Lenzi 2018 ).

Finally, a third route revolves around the principle of altering the earth’s radiation balance through the management of solar and terrestrial radiation. Such techniques are termed radiative forcing geoengineering technologies, and the main objective is temperature stabilization or reduction. Unlike negative emissions technologies, this is achieved without altering greenhouse gas concentrations in the atmosphere. The main radiative forcing geoengineering techniques that are discussed in the literature include stratospheric aerosol injection, marine sky brightening, cirrus cloud thinning, space-based mirrors, surface-based brightening and various radiation management techniques. All these techniques are still theoretical or at very early trial stages and carry a lot of uncertainty and risk in terms of practical large-scale deployment. At the moment, radiative forcing geoengineering techniques are not included within policy frameworks (Lawrence et al. 2018 ; Lockley et al. 2019 ).

Conventional mitigation technologies

As previously discussed, energy-related emissions are the main driver behind the increased greenhouse gas concentration levels in the atmosphere; hence, conventional mitigation technologies and efforts should be focused on both the supply and demand sides of energy. Mitigation efforts primarily discussed in the literature cover technologies and techniques that are deployed in four main sectors, power on the supply side and industry, transportation and buildings on the demand side. Within the power sector, decarbonization can be achieved through the introduction of renewable energy, nuclear power, carbon capture and storage as well as supply-side fuel switch to low-carbon fuels such as natural gas and renewable fuels. Furthermore, mitigation efforts on the demand side include the efficiency gains achieved through the deployment of energy-efficient processes and sector-specific technologies that reduce energy consumption, as well as end-use fuel switch from fossil-based fuels to renewable fuels, and, moreover, the integration of renewable power technologies within the energy matrix of such sectors (Mathy et al. 2018 ; Hache 2015 ). This section will review the literature on decarbonization and efficiency technologies and techniques that cover those four main sectors introduced. Figure  1 depicts the conventional mitigation technologies and techniques discussed in the literature and critically reviewed in this paper.

Major decarbonization technologies which focus on the reduction of CO 2 emissions related to the supply and demand sides of energy. Conventional mitigation technologies include renewable energy, nuclear power, carbon capture and storage (CCS) as well as utilization (CCU), fuel switching and efficiency gains. These technologies and techniques are mainly deployed in the power, industrial, transportation and building sectors

Renewable energy

According to a recent global status report on renewables, the share of renewable energy from the total final energy consumption globally has been estimated at 18.1% in 2017 (REN21 2019 ). An array of modern renewable energy technologies is discussed throughout the literature. The most prominent technologies include photovoltaic solar power, concentrated solar power, solar thermal power for heating and cooling applications, onshore and offshore wind power, hydropower, marine power, geothermal power, biomass power and biofuels (Mathy et al. 2018 ; Shinnar and Citro 2008 ; Hache 2015 ; REN21 2019 ; Hussain et al. 2017 ; Østergaard et al. 2020 ; Shivakumar et al. 2019 ; Collura et al. 2006 ; Gude and Martinez-Guerra 2018 ; Akalın et al. 2017 ; Srivastava et al. 2017 ).

In terms of power production, as of 2018, renewable energy accounted for approximately 26.2% of global electricity production. Hydropower accounted for 15.8%, while wind power’s share was 5.5%, photovoltaic solar power 2.4%, biopower 2.2% and geothermal, concentrated solar power and marine power accounted for 0.46% of the generated electricity (REN21 2019 ). While large-scale hydropower leads in terms of generation capacity as well as production, there has been a significant capacity increase in photovoltaic solar power and onshore wind power over the past decade. By the end of 2018, a total of 505 GW of global installed capacity for photovoltaic solar power has been noted as compared to 15 GW in 2008. Regarding wind power, 591 GW of global installed capacity is recorded in 2018 as compared to 121 GW in 2008. Global biopower capacity has been estimated at 130 GW in 2018 with a total 581 TWh of production in that year. China has maintained its position as the largest renewable energy producing country, from solar, wind and biomass sources. The total share of renewable energy in global power capacity has reached approximately 33% in 2018 (REN21 2019 ).

Besides the power sector, renewable energy can be deployed within the industry, transportation and building sectors. Photovoltaic and thermal solar energy as well as industrial end-use fuel switch to renewable fuels such as solid, liquid and gaseous biofuels for combined thermal and power production are examples of decarbonization efforts through renewables. Buildings can also benefit from solar as well as biomass-based technologies for power, heating and cooling requirements. In relation to the transportation sector, end-use fuel switch is a determinant to sector decarbonization. Some examples of biofuels are biodiesel, first- and second-generation bioethanol, bio-hydrogen, bio-methane and bio-dimethyl ether (bio-DME) (Srivastava et al. 2020 ; Chauhan et al. 2009 ; Hajilary et al. 2019 ; Osman 2020 ). Furthermore, hydrogen produced through electrolysis using renewable energy is a potential renewable fuel for sector decarbonization. Another example of sector decarbonization through renewable energy deployment is electric vehicles using renewable power (Michalski et al. 2019 ). Other mitigation measures within these sectors will be further discussed in the following section.

Variable renewables, such as solar and wind, are key technologies with significant decarbonization potential. One of the main technological challenges associated is the intermittent nature/variability in power production. This has been overcome by integrating such technologies with storage as well as other renewable baseload and grid technologies. Sinsel et al. discuss four specific challenge areas related to variable renewables, namely quality, flow, stability and balance. Furthermore, they present a number of solutions that mainly revolve around flexibility as well as grid technologies for distributed as well as centralized systems (Sinsel et al. 2020 ).

Economic, social and policy dimensions play an influencing role in renewable energy technology innovation and deployment. Pitelis et al. investigated the choice of policy instruments and its effectiveness in driving renewable energy technology innovation for 21 Organization for Economic Co-operation and Development (OECD) countries between 1994 and 2014. The study classified renewable energy policies into three categories: technology-push, demand-pull and systemic policy instruments. Furthermore, the study investigated the impact of each policy classification on innovation activity of various renewable energy technologies: solar, wind, biomass, geothermal and hydro. The study concluded that not all policy instruments have the same effect on renewable energy technologies and that each technology would require appropriate policies. However, the study suggested that demand-pull policy instruments are more effective in driving renewable energy innovation compared to alternative policy types (Pitelis et al. 2019 ). On barriers and drivers of renewable energy deployment, Shivakumar et al. highlighted various dimensions that may hinder or enable renewable energy project development. The main points highlighted revolve around policy, financial access, government stability and long-term intentions, administrative procedures and support framework or lack thereof, as well as the profitability of renewable energy investments (Shivakumar et al. 2019 ). Seetharaman et al. analysed the impact of various barriers on renewable energy deployment. The research confirms that regulatory, social and technological barriers play a significant role in renewable energy deployment. The research does not find a significant direct relationship between economic barriers and project deployment; however, the interrelated nature between the economic dimension with regulatory, social and technological barriers affects deployment, however, indirectly (Seetharaman et al. 2019 ).

In terms of the relationship between financial accessibility and renewable energy deployment, Kim et al. empirically investigated such relationship by analysing a panel data set of 30 countries during a 13-year period from 2000 to 2013. Statistical evidence shows the positive impact of well-developed financial markets on renewable energy deployment and sector growth. Furthermore, the study confirms a positive and significant relationship between market-based mechanisms, such as clean development mechanism, with renewable energy deployment. There is a strong impact on photovoltaic solar and wind technologies, while the impact is marginal under biomass and geothermal technologies (Kim and Park 2016 ).

Pfeiffer et al. studied the diffusion of non-hydro renewable energy (NHRE) technologies in 108 developing countries throughout a 30-year period from 1980 to 2010. Based on the results, economic and regulatory policies played a pivotal role in NHRE deployment, as well as governmental stability, higher education levels and per capita income. On the other hand, growth in energy demand, aid and high local fossil fuel production hindered NHRE diffusion. In contrast with Kim et al., the study finds weak support to show that international financing mechanisms and financial market development positively influenced diffusion (Pfeiffer and Mulder 2013 ). The reason may be related to how the analysis was constructed, different data sets, periods and statistical methods.

Decarbonization through renewable energy deployment is extremely significant. Development of renewable energy projects should be seen as a top priority. The areas that would drive decarbonization through renewable energy and should be focused upon by policymakers, financiers and market participants include policy instruments, financial support and accessibility, and market-based mechanisms to incentivize project developers. Moreover, governmental support frameworks, public education for social acceptance as well as research and development efforts for technological advances and enhanced efficiencies are important focus areas.

Nuclear power

According to the latest report prepared by the international atomic energy agency (IAEA), as of 2018, 450 nuclear energy plants are operational with a total global installed capacity of 396.4 GW. It is projected that an increase of 30% in installed capacity will be realized by 2030 (from a base case of 392 GW in 2017). As a low-case projection scenario, it is estimated that by 2030 a 10% dip might be realized based on the 2017 numbers. On the long term, it is projected that global capacity might reach 748 GW by 2050, as a high-case scenario (IAEA 2018 ). Pravalie et al. provide an interesting review of the status of nuclear power. The investigation demonstrates the significant role nuclear power has played in terms of contribution to global energy production as well as its decarbonization potential in the global energy system. The study presents an estimation of approximately 1.2–2.4 Gt CO 2 emissions that are prevented annually from nuclear power deployment, as alternatively the power would have been produced through coal or natural gas combustion. The paper suggests that to be in line with the 2 °C target stipulated by the Paris agreement, nuclear plant capacity must be expanded to approximately 930 GW by 2050, with a total investment of approximately $ 4 trillion (Prăvălie and Bandoc 2018 ).

Although nuclear energy is considered as a low-carbon solution for climate change mitigation, it comes with a number of major disadvantages. First, the capital outlay and operating costs associated with nuclear power development are quite significant. Furthermore, risk of environmental radioactive pollution is a major issue related to nuclear power, which is mainly caused through the threat of reactor accidents as well as the danger associated with nuclear waste disposal (Prăvălie and Bandoc 2018 ; Abdulla et al. 2019 ). While conventional fission-based nuclear plants are suggested to be phased out in future, the introduction of enhanced fusion-based nuclear technology may positively contribute to mitigation efforts in the second half of the century. Fusion power is a new generation of nuclear power, which is more efficient than the conventional fission-based technology and does not carry the hazardous waste disposal risk associated with conventional fission-based nuclear technology. Furthermore, fusion power is characterized as a zero-emission technology (Prăvălie and Bandoc 2018 ; Gi et al. 2020 ).

Carbon capture, storage and utilization

Carbon capture and storage is a promising technology discussed in the literature as a potential decarbonization approach to be applied to the power as well as the industrial sectors. The technology consists of separating and capturing CO 2 gases from processes that rely on fossil fuels such as coal, oil or gas. The captured CO 2 is then transported and stored in geological reservoirs for very long periods. The main objective is the reduction in emission levels while utilizing fossil sources. Three capturing technologies are discussed in the literature: pre-combustion, post-combustion and oxyfuel combustion. Each technology carries a specific process to extracting and capturing CO 2 . Post-combustion capture technologies, however, are the most suitable for retrofit projects and have vast application potential. Once CO 2 has been successfully captured, it is liquified and transported through pipelines or ships to suitable storage sites. Based on the literature, storage options include depleted oil and gas fields, coal beds and underground saline aquifers not used for potable water (Vinca et al. 2018 ). Some of the main drawbacks of carbon capture and storage include safety in relation to secured storage and the possibility of leakage. Negative environmental impacts that may result from onshore storage locations that undergo accidental leakage have been investigated by Ma et al. The investigation focused on the impact of leakage on agricultural land (Ma et al. 2020 ). Risk of leakage and associated negative impacts have also been pointed out by Vinca et al. ( 2018 ). Other issues related to this technology include public acceptance (Tcvetkov et al. 2019 ; Arning et al. 2019 ) as well as the high deployment costs associated (Vinca et al. 2018 ). Another pathway post-carbon capture is the utilization of the CO 2 captured in the production of chemicals, fuels, microalgae and concrete building materials, as well as utilization in enhanced oil recovery (Hepburn et al. 2019 ; Aresta et al. 2005 ; Su et al. 2016 ; Qin et al. 2020 ).

Large-scale deployment of carbon capture storage and utilization technologies is yet to be proven. According to the international energy agency, there are only 2 carbon capture and storage projects under operation as of 2018, with a combined annual capture capacity of 2.4 MtCO 2 . There are 9 more carbon capture projects under development and are projected to increase capacity to 11 MtCO 2 by 2025; however, a significant deviation exists from the sustainable development scenario targeted by the international energy agency for 2040 which is a capacity of 1488 MtCO 2 (IEA 2019a ).

Fuel switch and efficiency gains

Fuel switching in the power sector from coal to gas, in the short-term, has been discussed extensively in the literature as a potential approach to economically transition to a low-carbon and hopefully a zero-carbon economy in future (Victor et al. 2018 ; Wendling 2019 ; Pleßmann and Blechinger 2017 ). The move to natural gas is also applicable to industry, transportation and building sectors; however, as discussed previously the switch to renewable fuels is a more sustainable approach creating further decarbonization potential in these sectors.

In addition to fuel switching, efficiency gains are of extreme significance within mitigation efforts. Efficiency gains in the power sector are achieved through improvements in thermal power plants by enhancing the efficiency of fuel combustion as well as improving turbine generator efficiencies. Furthermore, waste heat recovery for additional thermal as well as electric production enhances efficiency. In gas-fired power plants, the utilization of a combined cycle technology enhances the efficiency significantly. Combined heat and power units have also played an interesting role in efficiency gains. Technological advances within transmission and distribution networks also enhance efficiencies by reducing losses (REN21 2019 ).

In industry, there are many potential areas where efficiency gains may be realized. For example, in steel and cement applications, waste heat can be recovered for onsite power and heat production through the installation of waste heat-driven power plants that utilize waste heat from exhaust gases. For industries that utilize process steam, there is an excellent opportunity to utilize waste steam pressure to generate electric power for onsite usage or drive rotating equipment. The application of back pressure steam turbines in areas where steam pressure reduction is required can enhance energy efficiency significantly. The same approach can be deployed in applications where gas pressure reduction is required, however, using turboexpanders. Waste gases from industrial processes can also be utilized to generate onsite heat and power using micro- and small gas turbines. In addition, further efficiency gains can be realized through the deployment of advanced machinery controls in a multitude of processes and industrial sectors.

A number of factors influence energy efficiency within buildings, first the building design as well as materials utilized in construction, e.g. insulation and glazing. Furthermore, appliances, devices and systems used throughout buildings, e.g. heating, cooling and ventilation systems, and lighting, play a pivotal role in energy consumption. Efficiency gains can be realized by utilizing energy-efficient systems and appliances as well as improved construction materials (REN21 2019 ; Leibowicz et al. 2018 ).

In the transportation sector, efficiency gains can be realized through the introduction of enhanced and more efficient thermal engines, hybrid and electric vehicles as well as hydrogen (H 2 ) vehicles (Hache 2015 ). Furthermore, efficiency gains can be achieved through technological advances within aviation, shipping and rail, although rail is currently one of the most energy-efficient modes. Efficiency measures in the transportation sector can also take other forms. For example, travel demand management, to reduce frequency and distance of travel, can be an interesting approach. Moreover, shifting travel to the most efficient modes where possible, such as electrified rail, and reducing dependence on high-intensity travel methods can play an interesting role in enhancing efficiency (IEA 2019b ).

• Negative emissions technologies

Most of the climate pathways that were investigated by the Intergovernmental Panel on Climate Change (IPCC) included the deployment of negative emissions technologies along with conventional decarbonization technologies to assess the feasibility of achieving the targets mandated by the Paris agreement. Only two negative emissions technologies have been included in the IPCC assessments so far, bioenergy carbon capture and storage as well as afforestation and reforestation (IPCC 2018 ).

Gasser et al. empirically investigated the potential negative emissions needed to limit global warming to less than 2 °C. The analysis utilized an IPCC pathway that is most likely to maintain warming at such level and constructed a number of scenarios based on conventional mitigation assumptions in an attempt to quantify the potential negative emissions efforts required. The results indicated that in the best-case scenario, that is under the best assumptions on conventional mitigation efforts, negative emissions of 0.5–3Gt C year −1 and 50–250 Gt C of storage capacity are required. Based on a worst-case scenario, negative emissions of 7–11 Gt C year −1 and 1000–1600 Gt C of storage capacity are required. (1 Gigaton Carbon = 3.6667 Gigaton CO 2 e) The results indicate the inevitable need for negative emissions, even at very high rates of conventional mitigation efforts. Furthermore, the study suggests that negative emissions alone should not be relied upon to meet the 2 °C target. The investigation concluded that since negative emissions technologies are still at an infant stage of development, conventional mitigation technologies should remain focused upon within climate policy, while further financial resources are to be mobilized to accelerate the development of negative emissions technologies (Gasser et al. 2015 ).

It is argued that negative emissions technologies should be deployed to remove residual emissions after all conventional decarbonization efforts have been maximized and that such approach should be utilized to remove emissions that are difficult to eliminate through conventional methods (Lin 2019 ). It is important to note that negative emissions should be viewed as a complementary suite of technologies and techniques to conventional decarbonization methods, and not a substitute (Pires 2019 ).

The significant role of negative emissions in meeting climate targets is understood and appreciated amongst academics, scientists and policymakers; however, there still remains a debate on the social, economic and technical feasibility as well as the risk associated with large-scale deployment (Lenzi 2018 ). This section will carry out an extensive literature review on the main negative emissions technologies and techniques, their current state of development, perceived limitations and risks as well as social and policy implications. Figure  2 depicts the major negative emissions technologies and carbon removal methods discussed in the literature and critically reviewed in this article.

Major negative emissions technologies and techniques which are deployed to capture and sequester carbon from the atmosphere. This approach includes bioenergy carbon capture and storage, afforestation and reforestation, biochar, soil carbon sequestration, enhanced terrestrial weathering, wetland restoration and construction, direct air carbon capture and storage, ocean alkalinity enhancement and ocean fertilization

Bioenergy carbon capture and storage

Bioenergy carbon capture and storage, also referred to as BECCS, is one of the prominent negative emissions technologies discussed widely in the literature. The Intergovernmental Panel on Climate Change (IPCC) heavily relied on bioenergy carbon capture and storage within their assessments as a potential route to meet temperature goals (IPCC 2018 ). The technology is simply an integration of biopower, and carbon capture and storage technologies discussed earlier. The basic principle behind the technology is quite straightforward. Biomass biologically captures atmospheric CO 2 through photosynthesis during growth, which is then utilized for energy production through combustion. The CO 2 emissions realized upon combustion are then captured and stored in suitable geological reservoirs (Pires 2019 ; RoyalSociety 2018 ). This technology can significantly reduce greenhouse gas concentration levels by removing CO 2 from the atmosphere. The carbon dioxide removal potential of this technology varies within the literature; however, a conservative assessment by Fuss et al. presents an estimated range of 0.5–5 GtCO 2 year −1 by 2050 (Fuss et al. 2018 ). In terms of global estimates for storage capacity, the literature presents a wide range from 200 to 50,000 GtCO 2 (Fuss et al. 2018 ). Cost estimates for carbon dioxide removal through bioenergy carbon capture and storage are in the range of $100-$200/tCO 2 (Fuss et al. 2018 ).

The biomass feedstocks utilized for this approach can either be dedicated energy crops or wastes from agricultural or forestry sources. Furthermore, such feedstocks can either be used as dedicated bio-based feedstocks or can be combined with fossil-based fuels in co-fired power plants (RoyalSociety 2018 ). Besides the standard combustion route, the literature suggests that CO 2 can be captured in non-power bio-based applications, such as during the fermentation process in ethanol production or the gasification of wood pulp effluent, e.g. black liquor, in pulp production (McLaren 2012 ; Pires 2019 ).

The main challenge associated with this technology is the significant amount of biomass feedstocks required to be an effective emission abatement approach. Under large-scale deployment, resource demand when utilizing dedicated crops would be quite significant, with high pressure exerted on land, water as well as nutrient resources. A major issue would be the direct competition with food and feed crops for land, freshwater and nutrients (RoyalSociety 2018 ; GNASL 2018 ). Heck et al. empirically investigated the large-scale deployment of bioenergy carbon capture and storage for climate change abatement and demonstrated its impact on freshwater use, land system change, biosphere integrity and biogeochemical flows. Furthermore, the investigation identified the interrelated nature between each of these dimensions as well as the associated impacts when any one dimension is prioritized (Heck et al. 2018 ). A sustainable approach to land use is quite critical in approaching bioenergy carbon capture and storage. Competing with food for arable land and changing forest land to dedicated plantations have serious negative social and environmental effects. Harper et al. argue that the effectiveness of this technology in achieving negative emissions is based on several factors which include previous land cover, the initial carbon gain or loss due to land-use change, bioenergy crop yields, and the amount of harvested carbon that is ultimately sequestered. Their empirical investigation highlights the negative impact of bioenergy carbon capture and storage when dedicated plantations replace carbon-dense ecosystems (Harper et al. 2018 ). Another issue discussed in the literature is the albedo effects of biomass cultivation. This is mainly applicable in high-latitude locations, where biomass replaces snow cover and reduces radiation reflection potential which offsets mitigation efforts (Fuss et al. 2018 ).

In terms of technology readiness, bioenergy technologies are to a certain extent well developed; however, carbon capture and storage are still at an early stage. Technology risk is mainly associated with storage integrity and the potential of leakage as discussed previously on carbon capture and storage. Furthermore, Mander et al. discuss the technical difficulties in scaling deployment within a short period. Besides, they question whether this technology can deliver its abatement potential within the projected time frame. In terms of policy, it is argued that a strong framework, as well as adequate incentives, need to be in place to properly push the technology forward (Mander et al. 2017 ). Commercial logic may not be enough to drive forward global deployment. Financial viability of such projects will depend on a utilitarian carbon market that caters for negative emissions as well as an appropriate carbon price that incentivizes deployment (Hansson et al. 2019 ). Therefore, policy should look at ways to strengthen carbon pricing mechanisms and introduce negative emissions as a new class of tradeable credits (Fajardy et al. 2019 ).

Afforestation and reforestation

During tree growth, CO 2 is captured from the atmosphere and stored in living biomass, dead organic matter and soils. Forestation is thus a biogenic negative emissions technology that plays an important role within climate change abatement efforts. Forestation can be deployed by either establishing new forests, referred to as afforestation, or re-establishing previous forest areas that have undergone deforestation or degradation, which is referred to as reforestation. Depending on tree species, once forests are established CO 2 uptake may span 20–100 years until trees reach maturity and then sequestration rates slow down significantly. At that stage, forest products can be harvested and utilized. It is argued that forest management activities and practices have an environmental impact and should be carefully planned (RoyalSociety 2018 ). Harper et al. discuss several advantages and co-benefits that are associated with forest-based mitigation which include biodiversity, flood control as well as quality improvement for soil, water and air (Harper et al. 2018 ).

Carbon can be stored in forests for a very long time; however, permanence is vulnerable due to natural and human disturbances. Natural disasters such as fire, droughts and disease or human-induced deforestation activities are all risks that negatively impact storage integrity. In general, biogenic storage has a much shorter lifespan than storage in geological formations, such as in the case of bioenergy carbon capture and storage (Fuss et al. 2018 ). Another issue related to forestation is land requirement as well as competition with other land use. Significant amounts of land are required to achieve effective abatement results (RoyalSociety 2018 ). Fuss et al. discuss another issue and that is the albedo effect. Forests in high latitudes would actually be counterproductive, accelerating local warming as well as ice and snow cover loss. They argue that tropical areas would be the most suitable zones to host forestation projects. However, competition with agriculture and other sectors for land will be another problem. Based on global tropical boundary limitations, an estimated total area of 500 Mha is argued to be suitable for forestation deployment. This would allow for a global carbon dioxide removal potential of 0.5–3.6 GtCO 2 year −1 by 2050. Removal costs are estimated at $5–$50/tCO 2 (Fuss et al. 2018 ).

In terms of technology readiness, afforestation and reforestation have already been widely adopted on a global level and have already been integrated within climate policies through the Kyoto protocol’s clean development mechanism programme since the 1990s. To drive forward forest-based mitigation efforts, the protocol introduced removal units which allowed forestation projects to yield tradeable credits. Despite the early policy measures, forest-based mitigation efforts accounted for a small fraction of emissions at that time. Forest-based abatement projects have also been introduced through national regulations as well as voluntary systems such as the reducing emissions from deforestation and forest degradation (REDD+) programme that was introduced by the United Nations in 2008. However, carbon sequestration through forestation remained insignificant, as it only accounted for 0.5% of the total carbon traded in 2013 (Gren and Aklilu 2016 ). The effectiveness of the REDD+ programme is argued in the literature after more than 10 years of its introduction. Hein et al. present a number of arguments around the programme’s poor track record in achieving its intended purpose of emissions reduction. However, despite the uncertainty and weaknesses discussed, REDD+ implementation intentions have been indicated by 56 countries in their INDC submissions under the Paris agreement (Hein et al. 2018 ). Permanence, sequestration uncertainty, the availability of efficient financing mechanisms as well as monitoring, reporting and verification systems are all difficulties associated around forest-based abatement projects (Gren and Aklilu 2016 ).

Biochar has recently gained considerable recognition as a viable approach for carbon capture and permanent storage and is considered as one of the promising negative emissions technologies. Biochar is produced from biomass, e.g. dedicated crops, agricultural residues and forestry residues, through a thermochemical conversion process. It is produced through pyrolysis, a process of heating in the absence of oxygen, as well as through gasification and hydrothermal carbonization (Matovic 2011 ; Oni et al. 2020 ; Osman et al. 2020a , b ). The carbon captured by biomass through CO 2 uptake during plant growth is then processed into a char that can be applied to soils for extended periods. The conversion process stores biomass carbon in a form that is very stable and resistant to decomposition. Stability in soils is perhaps the most important property of biochar that makes it a solid carbon removal technology. Although considered more stable than soil organic carbon, there are certain uncertainties around decomposition rates of various types of biochar, which depends on the feedstock used and process conditions utilized (Osman et al. 2019 ; Chen et al. 2019 ). Depending on the feedstock used, it is estimated that this technology can potentially remove between 2.1 and 4.8 tCO 2 /tonne of biochar (RoyalSociety 2018 ). Carbon removal potential, as well as costs, varies greatly in the literature; however, a conservative range is provided by Fuss et al. It is estimated that by 2050 global carbon reduction removal potential achieved through biochar can be in the range of 0.3–2 Gt CO 2  year −1 , with costs ranging from $90 to $120/tCO 2 (Fuss et al. 2018 ).

In terms of resource requirements, biochar production would require vast amounts of land to have an effective impact on greenhouse gas concentration levels. Land is required for feedstock cultivation, as well as for biochar dispersal acting as a carbon sink. While land for dedicated biomass cultivation may create competition issues with agriculture and other land-use sectors, same as the case of bioenergy carbon capture and storage, there would be no issues with areas required for biochar dispersal. This would be the case as long as the biochar is technically matched with the type of crop, soil and growing conditions related to the specific cropping system. Besides soil, Schmidt et al. introduced other carbon sink applications for biochar which include construction materials, wastewater treatment and electronics, as long as the product does not thermally degrade or oxidize throughout its life cycle (Schmidt et al. 2019 ). Furthermore, it has been argued in the literature that marginal and degraded lands can potentially be utilized for dedicated plantations, relieving pressure on land that can be used for other purposes. Moreover, using waste biomass eliminates the need for land and provides a waste disposal solution; however, competition over waste for other purposes increases feedstock availability risk as well as price volatility. Biomass availability is one of the limiting factors to successful large-scale deployment of biochar projects (RoyalSociety 2018 ).

In addition to the beneficial effect of capturing and storing CO 2 from the atmosphere, there is growing evidence in the literature that biochar also has an impact on other greenhouse gas emissions such as CH 4 and N 2 O. Although the literature shows a positive impact in many occasions, in terms of reduced emissions, Semida et al. present mixed results, where the application of biochar has positive as well as negative effects on CH 4 and N 2 O emissions. This is specific to the cropping system as well as the type of biochar utilized and its processing conditions (Semida et al. 2019 ). Xiao et al. also present conflicting results regarding biochar application, which is very specific to the condition of the soils amended with biochar (Xiao et al. 2019 ). Impact on greenhouse gas emissions should, therefore, be studied on a case-by-case basis.

Another benefit that is widely discussed in the literature is the positive effects associated with biochar application to soils. It is argued that soil quality and fertility are significantly enhanced. Improvement in nutrient cycling, reduction in nutrient leaching from the soil and an increase in water and nutrient retention as well as stimulation of soil microbial activity are all co-benefits associated with biochar application. However, this is mainly dependent on biochar physical and chemical properties. Such properties are defined by the type of feedstock utilized, pyrolysis conditions, as well as other processing conditions. Furthermore, despite the general perception that biochar positively impacts plant growth and production, which is true in a large number of cases, there is evidence that biochar application may hinder plant growth in certain cropping systems. This is based on the type of biochar, the quantity applied and the specific crops under cultivation and sometimes management practices. The evidence is mixed, and therefore careful analysis should be carried out to successfully match biochar with appropriate carbon sinks (Oni et al. 2020 ; Semida et al. 2019 ; El-Naggar et al. 2019 ; Maraseni 2010 ; Purakayastha et al. 2019 ; Xu et al. 2019 ).

Concerning the risks associated with large-scale deployment, albedo effect is mentioned in the literature. With high application rates of biochar to the soil surface, e.g. 30–60 tons/ha, it is argued that a decrease in surface reflectivity would increase soil temperature, which in turn would reduce the beneficial effect of carbon sequestration through this route (RoyalSociety 2018 ; Fuss et al. 2018 ). Other risks and challenges associated include the risk of reversibility and challenges in monitoring, reporting and verification. Moreover, limited policy incentives and support, as well as lack of carbon pricing mechanisms that incorporate CO 2 removal through biochar (Ernsting et al. 2011 ), hinder this technology’s potential for large-scale commercialization. Pourhashem et al. examined the role of government policy in accelerating biochar adoption and identified three types of existing policy instruments that can be used to stimulate biochar deployment in the USA: commercial financial incentives, non-financial incentives and research and development funding (Pourhashem et al. 2019 ). With the current technological advancements, in particular blockchain, a number of start-ups are developing carbon removal platforms to drive forward voluntary carbon offsets for consumers and corporations. A Finnish start-up, Puro.earth, has introduced biochar as a net-negative technology. Once verified through the company’s verification system, the carbon removal certificates generated by biochar producers are auctioned to potential offset parties. However, until carbon removal is adequately monetized and supported through sufficient policy instruments, biochar project development will probably not reach the scale required to have a profound impact within the time frame mandated by international policy.

Soil carbon sequestration

Soil carbon sequestration is the process of capturing atmospheric CO 2 through changing land management practices to increase soil carbon content. The level of carbon concentration within the soil is determined by the balance of inputs, e.g. residues, litter, roots and manure, and the carbon losses realized through respiration which is mainly influenced by soil disturbance. Practices that increase inputs and/or reduce losses drive soil carbon sequestration (RoyalSociety 2018 ; Fuss et al. 2018 ). It is well noted in the literature that soil carbon sequestration promotes enhanced soil fertility and health as well as improves crop yields due to organic carbon accumulation within soils (Fuss et al. 2018 ). Various land management practices that promote soil carbon sequestration are discussed in the literature which include cropping system intensity and rotation practices, zero-tillage and conservation tillage practices, nutrient management, mulching and use of crop residues and manure, incorporation of biochar, use of organic fertilizers and water management (RoyalSociety 2018 ; Srivastava 2012 ; Farooqi et al. 2018 ). Furthermore, the impact of perennial cropping systems on soil carbon sequestration is well documented in the literature. Agostini et al. investigated the impact of herbaceous and woody perennial cropping systems on soil organic carbon and confirmed an increase in soil organic carbon levels by 1.14–1.88 tCha −1  year −1 for herbaceous crops and 0.63–0.72 tCha −1  year −1 for woody crops. It is reported that these values are well above the proposed sequestration requirement (0.25 tCha −1  year −1 ) to make the crop carbon neutral once converted to biofuels (Agostini et al. 2015 ). The positive impact of perennial cropping systems on soil carbon sequestration is supported and documented in the literature by several other investigations (Nakajima et al. 2018 ; Sarkhot et al. 2012 ).

The main issues related to this approach revolve around permanence, sink saturation as well as the impact on other greenhouse gas emissions. According to Fuss et al., the potential of carbon removal through soil carbon sequestration is time-limited. Once soils reach a level of saturation, further sequestration is no longer achieved. This may take 10–100 years depending on soil type and climatic conditions. However, the Intergovernmental Panel on Climate Change (IPCC) defined a default saturation period of 20 years (Fuss et al. 2018 ). Once saturation is reached, land management practices need to be maintained indefinitely to mitigate reversal. A disadvantage to this would be the ongoing costs with no further removal benefits. Risks of reversibility are significant and weaken this approach’s storage integrity. Another negative effect discussed in the literature is the impact of soil carbon sequestration on other greenhouse gas emissions, mainly CH 4 and N 2 O; however, this effect is reported to be negligible (Fuss et al. 2018 ).

By 2050, the global carbon dioxide removal potential discussed in the literature is estimated between 2.3 and 5.3 GtCO 2  year −1 at costs ranging from $0 to $100 t/CO 2 (Fuss et al. 2018 ). While soil carbon sequestration is ready for large-scale deployment, since many of such practices are already being used, lack of knowledge, resistance to change as well as lack of policy and financial incentives are identified as barriers for scalability. Challenges around monitoring, reporting and verification, as well as concerns about sink saturation and potential reversibility, have been the main reasons behind slow policy action. However, non-climate policies have mainly promoted land management practices to improve soil quality, fertility and productivity as well as prevent land degradation (RoyalSociety 2018 ). While policy and market-based mechanisms are required to push this approach forward, international voluntary carbon removal platforms are emerging. A US-based platform (Nori) is based on the concept of soil carbon sequestration and operates by linking consumers and businesses that wish to offset their carbon footprint with farmers that offer carbon removal certificates that have been audited through an independent verification party. Using blockchain technology, this company is one step further in fighting the challenges associated with monitoring, reporting and verification systems.

Direct air carbon capture and storage

Direct air carbon capture and storage, also referred to as DACCS in the literature, is emerging as a potential synthetic CO 2 removal technology. The underlying principle behind this technology is the use of chemical bonding to remove atmospheric CO 2 directly from the air and then store it in geological reservoirs or utilize it for other purposes such as the production of chemicals or mineral carbonates. CO 2 is captured from the air by allowing ambient air to get in contact with chemicals known as sorbents. Furthermore, the sorbents are then regenerated by applying heat or water to release the CO 2 for storage or utilization. There are mainly two processes by which sorbents work: first through absorption, where the CO 2 dissolves in the sorbent material, typically using liquid sorbents such as potassium hydroxide or sodium hydroxide; second through adsorption, whereby the CO 2 adheres to the sorbent, typically using solid materials such as amines (Pires 2019 ; GNASL 2018 ; Gambhir and Tavoni 2019 ; Liu et al. 2018 ). Both processes require thermal energy to regenerate the sorbent and release the CO 2 ; however, it is important to note that less energy is required under the adsorption route (Gambhir and Tavoni 2019 ). A key issue widely discussed in the literature is the significant energy required by direct air carbon capture and storage plants. Besides the energy required for sorbent regeneration, energy is required for fans, pumps as well as compressors for pressurizing the CO 2 . It is of course very important to utilize low-carbon energy sources, preferably renewable energy as well as sources of waste heat, to drive the operation (Fuss et al. 2018 ). Another major drawback highlighted in the literature is the significant cost associated with developing direct air carbon capture and storage projects (Fuss et al. 2018 ). The major risk associated with this technology is CO 2 storage integrity, similar to that of carbon capture and storage and bioenergy carbon capture and storage (RoyalSociety 2018 ).

Gambhir et al. compare direct air carbon capture and storage to carbon capture and storage and explain that the former technology is more energy- and material-intensive due to the fact that capturing CO 2 from ambient air is much more difficult compared to capturing CO 2 from highly concentrated flue gas streams. Direct air carbon capture is three times energy-intensive compared to conventional carbon capture per ton of CO 2 removed (Gambhir and Tavoni 2019 ). However, direct air carbon capture and storage plants are more flexible and can be located anywhere, provided that low-carbon energy and adequate transportation and storage facilities are available. In terms of technology readiness, a lot of processes are currently being developed and are either under laboratory-scale or pilot-scale phases. Technology developers are mainly working on reducing energy requirements as this is one of the main challenges to deployment and scalability (RoyalSociety 2018 ).

The global potential for carbon dioxide removal has been estimated by Fuss et al. to be in the range of 0.5–5 GtCO 2  year −1 by 2050, and this may potentially go up to 40 GtCO 2  year −1 by the end of the century if the unexpected challenges associated with large-scale deployment are overcome. Furthermore, CO 2 removal costs are estimated at $600–$1000/tCO 2 initially, moving down to the range of $100–$300/tCO 2 as the technology matures (Fuss et al. 2018 ). Currently, there are no policy instruments to support this technology, similar to many of the negative emissions technologies discussed (RoyalSociety 2018 ).

Ocean fertilization

Ocean fertilization is the process of adding nutrients, macro such as phosphorus and nitrates as well as micro such as iron, to the upper surface of the ocean to enhance CO 2 uptake by promoting biological activity. Microscopic organisms, called phytoplankton, found at the surface layer of oceans are an important contributor to the concept of oceanic carbon sequestration. The sequestered CO 2 , in the form of organic marine biomass, is naturally transported to the deep ocean; this process is termed “the biological pump”. It is important to note that this downward flow is to a certain extent balanced by oceanic carbon respiration. Similar to land-based plants, phytoplankton utilizes light, CO 2 as well as nutrients to grow. In the natural system, nutrients are available in the ocean as a consequence of death and decomposition of marine life. Hence, marine production is limited by the availability of recycled nutrients in the ocean. The idea behind ocean fertilization is to introduce additional nutrients to increase the magnitude of biological production, which in turn increases CO 2 uptake rate as compared to the natural rate of respiration creating a carbon-negative atmospheric balance (RoyalSociety 2018 ; Williamson et al. 2012 ). Although there is not much information in the literature regarding carbon removal potential, it is estimated that ocean fertilization can potentially sequester up to 3.7 GtCO 2  year −1 by 2100 with a total global storage capacity of 70–300 GtCO 2 (RoyalSociety 2018 ). In terms of potential abatement costs, a range between $2 and $457/tCO 2 has been estimated in the literature (Fuss et al. 2018 ).

Side effects of ocean fertilization that are discussed in the literature include ocean acidification, deep and mid-water oxygen decrease or depletion, increase in production of further greenhouse gases, unpredictable impact on food cycles, creation of toxic algal blooms as well as mixed effects on the seafloor and upper ocean ecosystems (Fuss et al. 2018 ; Williamson et al. 2012 ). Furthermore, the environmental, economic and social effects as well as the energy and material resources associated with fertilizer production, transportation and distribution are significant. Moreover, according to Fuss et al., uncertainty around permeance is a major drawback. Permanence depends on whether the sequestered carbon, in organic form, remains dissolved in the different layers of the ocean or whether sedimentation allows it to settle within long-term oceanic compartments for extended periods (Fuss et al. 2018 ). The issue with permeance, impact on ecosystems, low sequestration efficiency, as well as lack of adequate monitoring, reporting and verification systems, do not support the concept that ocean fertilization is an effective climate change abatement approach (Fuss et al. 2018 ; Williamson et al. 2012 ).

Enhanced terrestrial weathering

In the natural system, silicate rocks decompose; this is a process termed weathering. This chemical reaction consumes atmospheric CO 2 and releases metal ions as well as carbonate and/or bicarbonate ions. The dissolved ions are transported through groundwater streams through to rivers and eventually end up in the ocean where they are stored as alkalinity, or they precipitate in the land system as carbonate minerals. Enhanced weathering is an approach that can accelerate this weathering process to enhance CO 2 uptake on a much shorter timescale. This is achieved through milling silicate rocks to increase its reactive surface and enhance its mineral dissolution rate. The ground material is then applied to croplands providing a multitude of co-benefits (RoyalSociety 2018 ; Bach et al. 2019 ). Kantola et al. discuss the potential of applying this approach to bioenergy cropping systems (Kantola et al. 2017 ). According to Fuss et al., enhanced weathering promotes the sequestration of atmospheric carbon in two forms, inorganic and organic. Inorganic carbon is sequestered through the production of alkalinity and carbonates, as discussed above. Organic carbon, on the other hand, is sequestered when additional carbon sequestration is realized from enhanced biomass production, through photosynthesis, as a result of the nutrients that are naturally released from the rocks (Fuss et al. 2018 ).

Besides the carbon removal potential associated with enhanced weathering, the literature presents a number of positive side effects. This includes favourable impact on soil hydrological properties, a source for plant nutrients allowing lower dependence on conventional fertilizers, increase in water pH, enhanced soil health, increase in biomass production and an opportunity to reduce dependence on conventional pesticides. Such benefits depend on the type of rock and its application rate, climate, soil and cropping system (RoyalSociety 2018 ; Fuss et al. 2018 ; de Oliveira Garcia et al. 2019 ; Strefler et al. 2018 ).

In terms of technology readiness, enhanced weathering can be practically deployed at the moment. Current land management practices incorporate the application of granular materials, e.g. lime. Existing equipment can be utilized with no additional investment in equipment or infrastructure. The technologies related to quarrying, crushing and grinding are well developed, and there would not be issues with scalability. However, under large-scale deployment, the energy required for extraction, production and transportation would be quite significant (RoyalSociety 2018 ). Careful attention should be paid to the carbon footprint of enhanced weathering operations to assess actual sequestration potential. Lefebevre et al. investigated carbon sequestration through EW in Brazil by conducting a life cycle assessment to identify the carbon removal potential using basalt on agricultural land in Sao Paolo. The investigation presented several key findings, first, that the operation emits 75 kg of CO 2 per ton of CO 2 removed through enhanced weathering and 135 kg of CO 2 per ton of CO 2 removed through carbonation. This is based on a distance of 65 km between the production site and the field on which the ground rock is applied. The results indicate a maximum road travel distance of 540 km for carbonation and 990 km for enhanced weathering, above which the emissions offset the potential benefits realized from such activity. It is concluded that transportation is a major drawback which places limitations on the potential viability of this technology. Furthermore, the results suggest a capture rate of approximately 0.11–0.2 tCO 2 e/ton of basaltic rock applied (Lefebvre et al. 2019 ).

Another approach to reducing pressure on the resources required for extraction is to utilize silicate wastes from various industries. Potential materials include wastes from mining operations, cement, steel, aluminium, and coal or biomass combustion activities (Renforth 2019 ). However, this needs to be carefully assessed as potentially there is a risk of releasing heavy metals into soils if inappropriate materials are used (Fuss et al. 2018 ). Another risk associated with enhanced weathering is the potential health risk from the respiration of fine dust in the production and application of finely ground rock materials (Strefler et al. 2018 ). Furthermore, uncertainties about the impacts of enhanced weathering on microbial and marine biodiversity require further investigation (RoyalSociety 2018 ).

In terms of permanence, the sequestered CO 2 can be stored in several earth pools. Initially, CO 2 can be stored as dissolved inorganic carbon, alkalinity, in soils as well as in groundwater. Depending on conditions, precipitation of carbonate minerals in the soil can take place and such minerals can be stored for an extended period (in the order of 10 6  years) (Fuss et al. 2018 ). If precipitation does not take place, the dissolved inorganic carbon will be transported to the ocean through water streams, where it would be stored as alkalinity, providing a number of additional benefits and challenges to the oceanic pool. Based on an extensive literature assessment, Fuss et al. estimate global carbon removal potential of 2–4 GtCO 2  year −1 by 2050 at a cost ranging from $50 to $200/tCO 2 (Fuss et al. 2018 ). Strefler et al. conducted a techno-economic investigation on the carbon removal potential and costs of enhanced weathering using two rock types (dunite and basaltic rock). The results are inline and support the estimates presented by Fuss et al. in terms of removal potential as well as costs. Furthermore, the investigation highlighted the dimensions that influence removal potential and cost, mainly being rock grain size and weathering rates. Finally, the study indicated that climates that are warm and humid with lands that lack sufficient nutrients are the most appropriate areas for enhanced weathering activities (Strefler et al. 2018 ).

At the moment, enhanced weathering is not included in any carbon markets and does not have any policy support. Further research on social and environmental implications as well as adequate monitoring, reporting and verification systems needs to be developed for this approach to gain traction (RoyalSociety 2018 ). Moreover, integration within carbon markets and adequate carbon pricing are required to incentivize deployment.

Ocean alkalinity enhancement

Ocean alkalinity enhancement has been discussed in the literature as a potential route to inorganic carbon capture and storage within the ocean. The ocean already absorbs a significant amount of atmospheric CO 2 annually, mainly through two routes. First, through the diffusion of CO 2 from the atmosphere into the water, based on the differences of CO 2 partial pressure between the atmosphere and the ocean. The second route is through photosynthesis of phytoplankton discussed earlier. This section will mainly focus on CO 2 oceanic uptake through diffusion that is governed by the oceanic partial pressure of CO 2 . When CO 2 moves from the atmosphere into the ocean, the gas reacts with water to form carbonic acid, which further dissociates into bicarbonate and carbonate ions, where dissolved inorganic carbon is stored. This reaction also releases hydrogen ions, which increases the ocean’s acidity (Renforth and Henderson 2017 ). It is discussed in the literature that oceanic pH has a significant impact on CO 2 partial pressure for a given inorganic carbon content, which is the sum of carbon concentrations in carbonic acid, carbonate and bicarbonate ions (Kheshgi 1995 ). Increasing ocean alkalinity is argued to decrease the surface ocean partial pressure, promoting further oceanic CO 2 uptake, with a major positive side effect of reducing ocean acidification. As alkalinity increases, more carbonic acid is converted to bicarbonate and carbonate ions and greater amounts of carbon are stored in inorganic form (Renforth and Henderson 2017 ).

There are several approaches discussed in the literature on how an increase in oceanic alkalinity can be achieved. The concept of enhanced weathering is the first approach to increase alkalinity within oceans. As previously discussed, dissolved inorganic carbon in the form of bicarbonate and carbonate ions is a product of enhanced terrestrial weathering. If precipitation does not occur, the bicarbonate and carbonate ions are transported through water streams and end up in the ocean, increasing its alkalinity. Another approach is the addition of alkaline silicate rocks directly into the ocean, whereby finely ground rocks are added to the seawater for CO 2 uptake and carbon storage in the form of bicarbonate and carbonate ions, further enhancing alkalinity as well as inducing additional atmospheric CO 2 absorption (Bach et al. 2019 ). Another approach to increasing alkalinity was proposed by Kheshgi in the mid-1990s and that is the addition of lime (CaO) to the ocean surface. The main drawback of this approach is the energy required for the calcination of limestone as well as the CO 2 emissions realized (Kheshgi 1995 ). Another approach discussed in the literature is the accelerated weathering of limestone. This concept includes utilizing a reactor and reacting limestone (CaCO 3 ) with seawater and a gas stream that is high in CO 2 concentration to facilitate mineral dissolution. The main drawback of this approach is the excessive water requirement (Renforth and Henderson 2017 ). Finally, the last approach to enhancing alkalinity was introduced by House et al. whereby an alkaline solution is produced through an electrochemical method (House et al. 2009 ). Besides the challenges associated with each of the approaches presented, challenges around the impact of alkalinity enhancement on the oceanic ecosystem is still an area that needs further investigation. Furthermore, issues are raised around monitoring and regulations related to oceanic modifications (Renforth and Henderson 2017 ).

In terms of permanence, carbon can be stored for extended periods, in the order of 10 4  years, in the form of dissolved inorganic carbon. The ocean currently stores approximately 140,000 GtCO 2 , and with some changes in its chemistry, it may be able to store in the order of trillions of tons of CO 2 (Renforth and Henderson 2017 ). There is, however, a risk of reversal pointed out if mineral precipitation takes place, reducing the carbon carrying capacity of the water (RoyalSociety 2018 ). According to Renforth et al., the cost of removing CO 2 through ocean alkalinity enhancement is estimated between $10 and $190/tCO 2 , depending on the approach utilized in producing, transporting and distributing the alkaline material (Renforth and Henderson 2017 ). Currently, no policies or carbon pricing mechanisms incentivize the pursuit of climate change abatement through this technique, and there is still a need for field trials before deploying such approach on a large scale.

Wetland restoration and construction

Wetlands are high carbon density ecosystems that facilitate atmospheric carbon sequestration through photosynthesis and subsequent storage in above-ground and below-ground biomass as well as soil organic matter (Villa and Bernal 2018 ). Examples of wetlands include peatlands as well as coastal habitats such as mangrove forests, tidal marshes and seagrass meadows, also referred to as blue carbon ecosystems. Furthermore, constructed wetlands have been discussed in the literature as a valid solution to wastewater treatment. While peatlands and coastal wetlands are estimated to store between 44 and 71% of the world’s terrestrial biological carbon, such carbon stocks are vulnerable to deterioration due to habitat degradation. Risks leading to carbon loss, similar to forests, are caused by anthropogenic activities as well as natural disasters. Restoration efforts usually revolve around rewetting the ecosystems as well as further applicable measures (RoyalSociety 2018 ). A major drawback discussed in the literature is the substantial emissions of non-CO 2 greenhouse gases such as CH 4 and N 2 O associated with wetland habitats. A number of investigations emphasize the importance of incorporating the negative impact of non-CO 2 greenhouse gases in evaluating the sequestration benefits associated with a specific wetland restoration or construction project, as a specific site can either be a net carbon sink or a greenhouse gas source. This is based on various environmental and habitat management conditions (de Klein and van der Werf 2014 ; Gallant et al. 2020 ). Pindilli et al. conducted an empirical investigation on the impact of peatland restoration and management on the carbon sequestration potential of a 54,000 ha protected habitat over a 50-year period. The research modelled four scenarios: the first scenario included no management, the second added the impact of a catastrophic fire under no management, the third incorporated current management practices, while the final scenario promoted increased management activities. The results derived from this investigation showed that under the first two scenarios the peatland is declared a net source of CO 2 emissions, emitting 2.4 MtCO 2 and 6.5 MtCO 2 , respectively. Under the third and fourth scenarios, the peatland is declared a net carbon sink with significant sequestration rates of 9.9 MtCO 2 and 16.5 MtCO 2 , respectively, over the entire period of study. This illustrates the high impact of management activities on the carbon sequestration potential of wetland habitats (Pindilli et al. 2018 ).

Carbon sequestration and storage potential vary amongst different types of wetlands; for example, the estimated carbon sequestration rate is 6.3 ± 4.8 tCO 2 e ha −1  year −1 for mangroves, 8.0 ± 8.5 tCO2e ha −1 year −1 for salt marshes and 4.4 ± 0.95 tCO 2 e ha −1 year −1 for seagrass meadows. Within these habitats, the soil organic carbon accumulated in the top one metre amounted to 1060tCO 2 e ha −1 , 917 tCO 2 ha −1 and 500 tCO 2 ha −1 for mangroves, salt marshes and seagrasses, respectively (Sapkota and White 2020 ). The estimated cost of carbon abatement through wetland restoration and construction ranges between $10 and $100/tCO 2 (RoyalSociety 2018 ). According to Sapkota et al., several attempts have been made to include wetland-related offsets within existing voluntary and compliance carbon markets, including the development of protocols and methodologies. A number of methodologies have already been certified in the USA by various voluntary markets. However, despite the efforts, a few wetland restoration carbon offsets have been transacted so far (Sapkota and White 2020 ).

Alternative negative emissions utilization and storage techniques

Mineral carbonation is a process by which CO 2 is chemically reacted with minerals to form stable carbonates that can be safely stored below-ground or utilized in many applications (Olajire 2013 ; Wang et al. 2020 ). It very much resembles the natural weathering process of converting silicate rocks to carbonates, but at a much faster rate. The literature discusses two main routes for mineral carbonation, an ex situ industrial process above-ground that includes grinding and pre-treatment of minerals pre-reaction, or an in situ process with direct injection of CO 2 in silicate rocks below-ground (RoyalSociety 2018 ; Olajire 2013 ; Galina et al. 2019 ). Silicate rocks that contain high concentrations of calcium (Ca), magnesium (Mg) and iron (Fe) are the most suitable elements to react with CO 2 to form stable carbonates. Furthermore, industrial wastes that contain concentrations of such elements such as slag from steel plants and fly ash from coal combustion plants are also adequate materials to utilize for the carbonation process (Galina et al. 2019 ). Cost estimates under ex situ carbonation range from $50 to $300/tCO 2 , while in situ carbonation is estimated at approximately $17/tCO 2 (RoyalSociety 2018 ). An interesting utilization route of mineral carbonates is the replacement of conventional aggregates in concrete production. Substituting aggregates with mineral carbonates in conjunction with CO 2 curing to speed up the curing process and achieve higher strength concrete material is a promising approach to sequester CO 2 in the built environment (RoyalSociety 2018 ). Mineral carbonation using CO 2 that has been captured through direct air carbon capture or bioenergy carbon capture systems can be considered as a carbon-negative process since CO 2 is removed from the atmosphere and safely stored in carbonate form in geological formations, or in the built environment if the carbonates are utilized in construction. It is also important to note that mineral carbonation can also be coupled with carbon capture and storage technologies but would not be considered as a negative emissions technique if the CO 2 utilized is fossil-based.

Another approach discussed in the literature is the utilization of biomass materials in construction, while this is not a new concept, technological advancements in thermal and chemical treatments have mainly focused on increasing the variety and number of materials that can be utilized in different applications within the building industry. The basic principle behind this approach is that carbon is sequestered through photosynthesis, where the resulting biomass can then be utilized in construction allowing carbon to be stored for decades in the built environment, e.g. building structures, insulation and furniture. The potential CO 2 removal is estimated at approximately 0.5–1 GtCO 2  year −1 , through replacing conventional construction materials (RoyalSociety 2018 ). Besides the removal potential, by replacing conventional building materials such as steel and cement further emission reductions can be realized since these are carbon-intensive materials. Estimates of 14–31% reduction in global CO 2 emissions and 12–19% reduction in global fossil fuel consumption can be realized through this approach (RoyalSociety 2018 ). However, significant sustainable forestation projects are required.

• Radiative forcing geoengineering technologies

Radiative forcing geoengineering techniques are a set of technologies that aim to alter the earth’s radiative energy budget to stabilize or reduce global temperatures. This is achieved by either increasing the earth’s reflectivity by increasing shortwave solar radiation that is reflected to space, termed solar radiation management, or by enhancing longwave radiation that is emitted by the earth’s surfaces to space, termed terrestrial radiation management (Lawrence et al. 2018 ). This section briefly describes the various radiative forcing geoengineering techniques discussed in the literature. Figure  3 depicts the main techniques discussed in the literature and reviewed in this article.

Major radiative forcing geoengineering technologies that aim to alter the earth’s radiative energy budget to stabilize or reduce global temperatures. These technologies include stratospheric aerosol injection, marine sky brightening, cirrus cloud thinning, space-based mirrors and surface-based brightening

Stratospheric aerosol injection

Back in 1991, a very large volcanic eruption took place in the Philippines (Mount Pinatubo). During the eruption, a very large amount of sulphur dioxide gas (SO 2 ) was ejected, between 15 and 30 million tons, which induced sunlight reflectively and reduced global temperatures by 0.4–0.5 °C (Zhang et al. 2015 ). Stratospheric aerosol injection is a solar radiation management technology that aims to mimic the cooling effect caused by the volcanic eruption by artificially injecting reflecting aerosol particles in the stratosphere (Lawrence et al. 2018 ; Zhang et al. 2015 ). Through modelling and past volcanic eruption data, the maximum potential cooling from this approach is estimated between 2 and 5 W/m 2 (Lawrence et al. 2018 ). Smith et al. investigated the technology’s tactics and costs during the first 15 years of deployment starting in 2033. They surveyed potential deployment techniques and concluded that an aircraft-based delivery system is the most efficient method to deploy stratospheric aerosol injection. However, a new purpose-built high-altitude aircraft will need to be developed for this purpose as current models, even with modifications will not be sufficient. In an attempt to reduce anthropogenically driven radiative forcing rate by half, Smith et al. calculated initial costs for deployment to be in the range of $3.5 billion with average annual operating costs of $2.25 billion (approximately $1500/t SO 2 injected) (Smith and Wagner 2018 ). The main issue behind this technique is the uncertainty of the side effects and the harmful consequences of deployment, with a specific negative impact on the hydrological cycle as well as stratospheric ozone depletion (Zhang et al. 2015 ). It is important to note that while this approach will provide temporary temperature reduction it should not be considered a long-term solution. This approach is still at a very early stage of research and development (Lawrence et al. 2018 ).

Marine sky brightening

Marine sky brightening, also known as marine cloud brightening or cloud albedo enhancement, is another solar radiation management technology that aims to maintain or reduce global temperatures by enhancing cloud reflectivity. This is achieved through cloud seeding with seawater particles or with chemicals (Zhang et al. 2015 ). The main idea behind this technique is that seawater is sprayed into the air creating small droplets that easily evaporate leaving behind salt crystals that increase low-altitude cloud reflectivity above oceans (Ming et al. 2014 ). The potential cooling effect has been estimated between 0.8 and 5.4 W/m 2 , due to uncertainty, limited knowledge and spatial considerations (Lawrence et al. 2018 ). While this technique seems simple and straightforward, Latham et al. highlighted a number of problems associated with marine sky brightening. This includes the lack of spraying system that is capable of generating seawater particles of the size and quantities required, as well as further technical problems that are associated with the physical outcome of this approach as a result of the complex nature of cloud characteristics. Another challenge would be to undertake extensive trials and properly understand and overcome potential side effects (Latham et al. 2012 ). Again, this approach is still at an infant stage and will require extensive field research and development moving forward.

Space-based mirrors

Sunshade using space-based mirrors is a solar radiation management technique discussed in the literature that aims to reflect part of the incoming solar radiation to reduce global temperatures. For this approach to technically be deployed, space mirrors or reflectors need to be transported into orbit around the earth or placed at the Lagrangian L1 location between the earth and the sun, where the gravitational fields are in balance allowing the reflectors to remain stationary (Zhang et al. 2015 ; Kosugi 2010 ). While this approach can have a considerable cooling effect based on model simulations, development of such technology is still at a very infant stage. The major drawback associated with this approach is the economic feasibility of transporting materials into space. For this technology to be economically feasible, material transport costs need to be reduced from approximately $10,000/kg to less than $100/kg (Lawrence et al. 2018 ). Moreover, risks such as those associated with space debris and asteroid collisions or those associated with technical and communication failures need to be appropriately catered for (Lawrence et al. 2018 ).

Surface-based brightening

Another solar radiation management approach discussed in the literature is the brightening of the earth surface to increase the earth’s albedo and thus reduce global temperatures. This has been suggested through painting urban roofs and roads in white, as well as covering deserts and glaciers with plastic sheets that are highly reflective, and, furthermore, by placing reflective floating panels over water bodies (Ming et al. 2014 ). According to Lawrence et al., based on an extensive literature review, the cooling potential for this approach is too limited. Furthermore, substantial negative side effects are associated, such as disruption of desert ecosystems (Lawrence et al. 2018 ).

Cirrus cloud thinning

Cirrus cloud thinning is a terrestrial radiation management technique that aims to increase longwave radiation that is emitted from the earth’s surface to space to stabilize or reduce global temperatures. Cirrus clouds are high-altitude ice clouds that play a significant role within the earth’s radiation budget, having an impact on the earth’s hydrological cycle as well as surface temperatures. Cirrus clouds absorb terrestrial radiation as well as reflect incoming solar radiation; however, in general, they induce an average net warming effect from the imbalance between incoming and outgoing radiative forcings (Kärcher 2017 ). The basic principle behind this technique is the injection of aerosols into cirrus clouds to reduce its optical thickness as well as its lifetime to increase terrestrial radiation emission to space. This approach would require regular cloud injection, so an efficient and cost-effective delivery method needs to be in places such as dedicated aircrafts or drones. Bismuth triiodide (Bil 3 ) has been proposed as an effective cloud seeding material; however, its toxicity needs to be taken into account. Sea salt is another proposed option, yet it is not found to be as effective as Bil 3 (Lawrence et al. 2018 ). Based on model simulations, the maximum cooling effect through this approach has been estimated to be in the range of 2–3.5 W/m 2 (Lawrence et al. 2018 ). According to Lawrence et al., there are no published costs for cirrus cloud thinning and this approach still requires further research to understand side effects as well as to conduct appropriate research on potential delivery methods (Lawrence et al. 2018 ).

Miscellaneous radiation management techniques

Ming et al. proposed several theoretical technologies that target terrestrial radiation, mainly by creating thermal bridges to bypass the greenhouse gas insulating layer and be able to transfer thermal radiation out to space. The research paper presented several concepts which include transferring surface hot air to the troposphere, transferring latent and sensible heat to the top of the troposphere, transferring surface-sensible heat to the troposphere, as well as transferring cold air to the earth surface. For each concept, conceptual technologies are proposed. Some of the technologies discussed are systems that transfer heat beyond the earth system while generating energy, termed metrological reactors by the authors (Ming et al. 2014 ). While the idea of thermal bridging is interesting, the technologies and concepts introduced require further research, development and extensive field trials.

Bibliometric analysis of research on climate change mitigation

Bibliometric analysis is a statistical tool that can be used to quantitatively analyse the current state of scientific research, by highlighting gaps in the literature as well as trends. The Web of Science (WoS) core collection database was used in this analysis. The following search methodology was used to retrieve relevant research for further evaluation. Please note that the search was refined to a 5-year timespan from 2015 to 2020 to specifically evaluate scientific research efforts related to climate change mitigation after the Paris agreement in 2015.

Search Methodology:

You searched for: TOPIC: (“Climate change mitigation”) OR TOPIC: (“climate change abatement”) OR TOPIC: (“Decarbonization Technologies”) OR TOPIC: (“Bioenergy Carbon Capture & Storage”) OR TOPIC: (“Afforestation & Reforestation”) OR TOPIC: (“Soil Carbon Sequestration”) OR TOPIC: (“Direct Air Carbon Capture & Storage”) OR TOPIC: (“Ocean Fertilization”) OR TOPIC: (“Enhanced Terrestrial Weathering”) OR TOPIC: (“Ocean Alkalinity Enhancement”) OR TOPIC: (“Wetland Restoration & Construction”) OR TOPIC: (“Stratospheric Aerosol Injection”) OR TOPIC: (“Marine Sky Brightening”) OR TOPIC: (“Space-Based Sunshade/Mirrors”) OR TOPIC: (“Surface-Based Brightening”) OR TOPIC: (“Cirrus Cloud Thinning”) OR TOPIC: (“Carbon Dioxide Removal Techniques”) OR TOPIC: (“Radiative Forcing Geoengineering”)

Timespan: Last 5 years. Indexes: SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, ESCI.

Results: A total of 3993 papers were retrieved (3386 articles, 362 reviews, 201 proceedings papers, 71 early access and 61 editorial materials)

The results obtained were then analysed using VOSviewer software by plotting network and density visualization maps as shown in Fig.  4 . The maps are based on keyword co-occurrences. The visualization maps highlight various trends related to climate change mitigation, where areas related to biomass, carbon sequestration, especially soil carbon sequestration, and biochar have received high attention over the past 5 years. Furthermore, research related to policy, energy and in particular renewable energy has also received much attention. Although research on climate change mitigation is trending, a gap in the literature can be highlighted regarding research related to specific mitigation technologies. It is also evident from the literature that radiative forcing geoengineering technologies have not received much attention.

Bibliometric analysis of research on climate change mitigation: a network visualization map and b density visualization map, showing the recent state of scientific research on the topic of climate change mitigation by highlighting trends and gaps in the literature during 5 years between 2015 and 2020

Based on the current state of climate emergency, immediate development of viable mitigation and adaptation mechanisms is of extreme importance. An extensive literature review covered three main strategies to tackling climate change, conventional mitigation technologies, negative emissions technologies as well as radiative forcing geoengineering technologies. It is important to clarify that there is no ultimate solution to tackle climate change and that all technologies and techniques discussed in this review if technically and economically are viable should be deployed. As previously discussed, decarbonization efforts alone are not sufficient to meet the targets stipulated by the Paris agreement; therefore, the utilization of an alternative abatement approach is inevitable. While the concept of radiative forcing geoengineering in terms of managing the earth’s radiation budget is interesting, it is not a long-term solution, as it does not solve the root cause of the problem. It may, however, buy some time until greenhouse gas concentrations are stabilized and reduced. However, the technologies to be deployed are still to be developed and tested and side effects adequately catered for, which may be a lengthy process. Negative emissions technologies, on the other hand, provide a solid solution in combination with the current decarbonization efforts. While some of the negative emissions technologies presented in the literature review may still be at an early stage of development, biogenic-based sequestration techniques are to a certain extent mature and can be deployed immediately. Capturing CO 2 through photosynthesis is a straightforward and solid process; however, it needs to be effectively integrated within a technological framework as presented in the review. The challenge at the moment is that carbon pricing for negative emissions is at a very infant stage, mainly available through voluntary markets for a very small number of carbon removal methods and technically non-existent for most of the technologies discussed. Currently, carbon pricing would be insufficient to economically sustain carbon removal projects, apart from the existing framework for afforestation and reforestation projects. As carbon markets mature and offer incentives for carbon removal, this may change in near future. In order to aggressively drive negative emissions projects, policymakers and governments should devise appropriate policy instruments and support frameworks with a special focus on carbon pricing. Furthermore, the financial industry should provide enhanced financial support and accessibility as well as introduce efficient market-based mechanisms to incentivize project developers to establish carbon removal projects. At the very moment, biogenic-based sequestration projects are in a good position to efficiently utilize financial resources and policy support as most of the related technologies can be deployed immediately; however, efficient carbon pricing mechanisms that focus on carbon removal need to be aggressively developed and introduced. Furthermore, funding for technology research and development is also a very important aspect moving forward.

Abbreviations

Bio-dimethyl ether

Bismuth triiodide

Carbon dioxide

Carbon dioxide equivalent

Centre for Research on the Epidemiology of Disaster

Hydrofluorocarbons

Intended nationally determined contributions

Intergovernmental Panel on Climate Change

International Atomic Energy Agency

Internationally transferred mitigation outcomes

Megahectare

Million tons

Nitrous oxide

Non-hydro renewable energy

Organization for Economic Co-operation and Development

Perfluorocarbons

Per hectare

Reducing emissions from deforestation and forest degradation

Sulphur dioxide

Sulphur hexafluoride

United Nations Environment Programme

United Nations Framework Convention on Climate Change

Watt per square meter

Abdulla A et al (2019) Limits to deployment of nuclear power for decarbonization: insights from public opinion. Energy Policy 129:1339–1346. https://doi.org/10.1016/j.enpol.2019.03.039

Article   Google Scholar  

Agostini F et al (2015) Carbon sequestration by perennial energy crops: is the jury still out? BioEnergy Res 8:1057–1080. https://doi.org/10.1007/s12155-014-9571-0

Article   CAS   Google Scholar  

Akalın MK et al (2017) Supercritical fluid extraction of biofuels from biomass. Environ Chem Lett 15:29–41. https://doi.org/10.1007/s10311-016-0593-z

Aresta M et al (2005) Biotechnology to develop innovative syntheses using CO2. Environ Chem Lett 3:113–117. https://doi.org/10.1007/s10311-005-0009-y

Arning K et al (2019) Same or different? Insights on public perception and acceptance of carbon capture and storage or utilization in Germany. Energy Policy 125:235–249. https://doi.org/10.1016/j.enpol.2018.10.039

Bach LT et al (2019) CO 2 removal with enhanced weathering and ocean alkalinity enhancement: potential risks and co-benefits for marine pelagic ecosystems. Front Clim. https://doi.org/10.3389/fclim.2019.00007

Bataille C et al (2018) A review of technology and policy deep decarbonization pathway options for making energy-intensive industry production consistent with the Paris Agreement. J Clean Prod 187:960–973. https://doi.org/10.1016/j.jclepro.2018.03.107

Bustreo C et al (2019) How fusion power can contribute to a fully decarbonized European power mix after 2050. Fusion Eng Des 146:2189–2193. https://doi.org/10.1016/j.fusengdes.2019.03.150

Chauhan SK et al (2009) Environmental aspects of biofuels in road transportation. Environ Chem Lett 7:289–299. https://doi.org/10.1007/s10311-008-0185-7

Chen H et al (2019) Upcycling food waste digestate for energy and heavy metal remediation applications. Resour Conserv Recycl X 3:100015. https://doi.org/10.1016/j.rcrx.2019.100015

Collura S et al (2006) Miscanthus × Giganteus straw and pellets as sustainable fuels. Environ Chem Lett 4:75–78. https://doi.org/10.1007/s10311-006-0036-3

CRED (2019) Natural disasters 2018. CRED, Brussels. https://emdat.be/sites/default/files/adsr_2018.pdf . Accessed 18 Jan 2020

de Klein JJM, van der Werf AK (2014) Balancing carbon sequestration and GHG emissions in a constructed wetland. Ecol Eng 66:36–42. https://doi.org/10.1016/j.ecoleng.2013.04.060

de Oliveira Garcia W et al (2019) Impacts of enhanced weathering on biomass production for negative emission technologies and soil hydrology. Biogeosci Discuss 2019:1–35. https://doi.org/10.5194/bg-2019-386

Edenhofer ORP-M, Sokona Y, Kadner S, Minx JC, Brunner S, Agrawala S, Baiocchi G, Bashmakov IA, Blanco G, Broome J, Bruckner T, Bustamante M, Clarke L, Conte Grand M, Creutzig F, Cruz-Núñez X, Dhakal S, Dubash NK, Eickemeier P, Farahani E, Fischedick M, Fleurbaey M, Gerlagh R, Gómez-Echeverri L, Gupta S, Harnisch J, Jiang K, Jotzo F, Kartha S, Klasen S, Kolstad C, Krey V, Kunreuther H, Lucon O, Masera O, Mulugetta Y, Norgaard RB, Patt A, Ravindranath NH, Riahi K, Roy J, Sagar A, Schaeffer R, Schlömer S, Seto KC, Seyboth K, Sims R, Smith P, Somanathan E, Stavins R, von Stechow C, Sterner T, Sugiyama T, Suh S, Ürge-Vorsatz D, Urama K, Venables A, Victor DG, Weber E, Zhou D, Zou J, Zwickel T (2014) Technical summary. In: Edenhofer O, Pichs-Madruga R, Sokona Y, Farahani E, Kadner S, Seyboth K, Adler A, Baum I, Brunner S, Eickemeier P, Kriemann B, Savolainen J, Schlömer S, von Stechow C, Zwickel T, Minx JC (eds) Climate change 2014: mitigation of climate change. Contribution of working group III to the fifth assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. https://www.ipcc.ch/site/assets/uploads/2018/02/ipcc_wg3_ar5_technical-summary.pdf . Accessed 17 Dec 2019

El-Naggar A et al (2019) Biochar application to low fertility soils: a review of current status, and future prospects. Geoderma 337:536–554. https://doi.org/10.1016/j.geoderma.2018.09.034

Ernsting A et al (2011) Biochar and carbon markets. Biofuels. 2:9–12. https://doi.org/10.4155/bfs.10.78

Fajardy M et al (2019) Negative emissions: priorities for research and policy design. Front Clim. https://doi.org/10.3389/fclim.2019.00006

Farooqi ZU, Sabir M, Zeeshan N, Naveed K, Hussain MM (2018). Enhancing carbon sequestration using organic amendments and agricultural practices, carbon capture, utilization and sequestration. https://www.intechopen.com/books/carbon-capture-utilization-and-sequestration/enhancing-carbon-sequestration-using-organic-amendments-and-agricultural-practices . IntechOpen. https://doi.org/10.5772/intechopen.79336 . Accessed 9 Feb 2020

Fuss S et al (2018) Negative emissions—part 2: costs, potentials and side effects. Environ Res Lett 13:063002. https://doi.org/10.1088/1748-9326/aabf9f

Galina NR et al (2019) Evolution of carbon capture and storage by mineral carbonation: data analysis and relevance of the theme. Miner Eng 142:105879. https://doi.org/10.1016/j.mineng.2019.105879

Gallant K et al (2020) Measurement and economic valuation of carbon sequestration in Nova Scotian wetlands. Ecol Econ 171:106619. https://doi.org/10.1016/j.ecolecon.2020.106619

Gambhir A, Tavoni M (2019) Direct air carbon capture and sequestration: how it works and how it could contribute to climate-change mitigation. One Earth 1:405–409. https://doi.org/10.1016/j.oneear.2019.11.006

Gao S et al (2019) International carbon markets under the Paris Agreement: basic form and development prospects. Adv Clim Change Res 10:21–29. https://doi.org/10.1016/j.accre.2019.03.001

Gasser T et al (2015) Negative emissions physically needed to keep global warming below 2 °C. Nat Commun 6:7958. https://doi.org/10.1038/ncomms8958

Gi K et al (2020) Potential contribution of fusion power generation to low-carbon development under the Paris Agreement and associated uncertainties. Energy Strategy Rev 27:100432. https://doi.org/10.1016/j.esr.2019.100432

GNASL (2018) Negative emission technologies: what role in meeting Paris Agreement targets? German National Academy of Sciences Leopoldina. https://easac.eu/fileadmin/PDF_s/reports_statements/Negative_Carbon/EASAC_Report_on_Negative_Emission_Technologies.pdf . Accessed 28 Jan 2020

Goglio P et al (2020) Advances and challenges of life cycle assessment (LCA) of greenhouse gas removal technologies to fight climate changes. J Clean Prod 244:118896. https://doi.org/10.1016/j.jclepro.2019.118896

Gren I-M, Aklilu AZ (2016) Policy design for forest carbon sequestration: a review of the literature. For Policy Econ 70:128–136. https://doi.org/10.1016/j.forpol.2016.06.008

Gude VG, Martinez-Guerra E (2018) Green chemistry with process intensification for sustainable biodiesel production. Environ Chem Lett 16:327–341. https://doi.org/10.1007/s10311-017-0680-9

Hache EN, Nathalie A-T, Criqui P, Lavergne J-G, Chevallet L, Gentier S, Le Net E, Menanteau P, Thiais F (2015) Decarbonization wedges. https://www.allianceenergie.fr/wp-content/uploads/2017/06/Decarbonization_Wedges_report_0.pdf . https://doi.org/10.13140/RG.2.1.3040.1360 . Accessed 4 Feb 2020

Hajilary N et al (2019) Biofuel types and membrane separation. Environ Chem Lett 17:1–18. https://doi.org/10.1007/s10311-018-0777-9

Hansson A et al (2019) Preconditions for bioenergy with carbon capture and storage (BECCS) in sub-Saharan Africa: the case of Tanzania. Environ Dev Sustain. https://doi.org/10.1007/s10668-019-00517-y

Harper AB et al (2018) Land-use emissions play a critical role in land-based mitigation for Paris climate targets. Nat Commun 9:2938. https://doi.org/10.1038/s41467-018-05340-z

Heck V et al (2018) Biomass-based negative emissions difficult to reconcile with planetary boundaries. Nat Clim Change 8:151–155. https://doi.org/10.1038/s41558-017-0064-y

Hein J et al (2018) Deforestation and the Paris climate agreement: an assessment of REDD+ in the national climate action plans. For Policy Econ 90:7–11. https://doi.org/10.1016/j.forpol.2018.01.005

Hepburn C et al (2019) The technological and economic prospects for CO 2 utilization and removal. Nature 575:87–97. https://doi.org/10.1038/s41586-019-1681-6

House KZ et al (2009) Electrochemical acceleration of chemical weathering for carbon capture and sequestration. Energy Procedia 1:4953–4960. https://doi.org/10.1016/j.egypro.2009.02.327

Hussain A et al (2017) Emerging renewable and sustainable energy technologies: state of the art. Renew Sustain Energy Rev 71:12–28. https://doi.org/10.1016/j.rser.2016.12.033

IAEA (2018) IAEA annual report 2018, International Atomic Energy Agency. https://www.iaea.org/sites/default/files/publications/reports/2018/gc63-5.pdf . Accessed 5 Feb 2020

IEA (2019a) Tracking report—CCUS in power. International Energy Agency. https://www.iea.org/tcep/power/ccus . Accessed 5 Feb 2020

IEA (2019b) Tracking transport. International Energy Agency. https://www.iea.org/reports/tracking-transport-2019 . Accessed 6 Feb 2020

IPCC (2013) IPCC factsheet: what is the IPCC? https://www.ipcc.ch/site/assets/uploads/2018/02/FS_what_ipcc.pdf . Accessed 21 Jan 2020

IPCC (2018) Global warming of 1.5 °C. In: Masson-Delmotte V, Zhai P, Pörtner H-O, Roberts D, Skea J, Shukla PR, Pirani A, Moufouma-Okia W, Péan C, Pidcock R, Connors S, Matthews JBR, Chen Y, Zhou X, Gomis MI, Lonnoy E, Maycock T, Tignor M, Waterfield T (eds) An IPCC special report on the impacts of global warming of 1.5 °C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty. https://www.ipcc.ch/site/assets/uploads/sites/2/2019/06/SR15_Full_Report_High_Res.pdf . Accessed 22 Dec 2019

Kantola IB et al (2017) Potential of global croplands and bioenergy crops for climate change mitigation through deployment for enhanced weathering. Biol Lett 13:20160714. https://doi.org/10.1098/rsbl.2016.0714

Kärcher B (2017) Cirrus clouds and their response to anthropogenic activities. Curr Clim Change Rep 3:45–57. https://doi.org/10.1007/s40641-017-0060-3

Kheshgi HS (1995) Sequestering atmospheric carbon dioxide by increasing ocean alkalinity. Energy 20:915–922. https://doi.org/10.1016/0360-5442(95)00035-F

Kim J, Park K (2016) Financial development and deployment of renewable energy technologies. Energy Econ 59:238–250. https://doi.org/10.1016/j.eneco.2016.08.012

Kosugi T (2010) Role of sunshades in space as a climate control option. Acta Astronaut 67:241–253. https://doi.org/10.1016/j.actaastro.2010.02.009

Latham J et al (2012) Marine cloud brightening. Phil Trans R Soc A Math Phys Eng Sci 370:4217–4262. https://doi.org/10.1098/rsta.2012.0086

Lawrence MG et al (2018) Evaluating climate geoengineering proposals in the context of the Paris Agreement temperature goals. Nat Commun 9:3734. https://doi.org/10.1038/s41467-018-05938-3

Lefebvre D et al (2019) Assessing the potential of soil carbonation and enhanced weathering through Life Cycle Assessment: a case study for Sao Paulo State, Brazil. J Clean Prod 233:468–481. https://doi.org/10.1016/j.jclepro.2019.06.099

Leibowicz BD et al (2018) Optimal decarbonization pathways for urban residential building energy services. Appl Energy 230:1311–1325. https://doi.org/10.1016/j.apenergy.2018.09.046

Lenzi D (2018) The ethics of negative emissions. Glob Sustain 1:e7. https://doi.org/10.1017/sus.2018.5

Lin A (2019) Carbon dioxide removal after Paris. Ecol Law Q 45:533. https://doi.org/10.15779/Z386M3340F

Liu S et al (2018) High CO 2 adsorption by amino-modified bio-spherical cellulose nanofibres aerogels. Environ Chem Lett 16:605–614. https://doi.org/10.1007/s10311-017-0701-8

Lockley A et al (2019) Geoengineering and the blockchain: coordinating carbon dioxide removal and solar radiation management to tackle future emissions. Front Eng Manag 6:38–51. https://doi.org/10.1007/s42524-019-0010-y

Ma X et al (2020) Farmland degradation caused by radial diffusion of CO 2 leakage from carbon capture and storage. J Clean Prod 255:120059. https://doi.org/10.1016/j.jclepro.2020.120059

Mander S et al (2017) The role of bio-energy with carbon capture and storage in meeting the climate mitigation challenge: a whole system perspective. Energy Procedia 114:6036–6043. https://doi.org/10.1016/j.egypro.2017.03.1739

Maraseni TN (2010) Biochar: maximising the benefits. Int J Environ Stud 67:319–327. https://doi.org/10.1080/00207231003612225

Mathy S et al (2018) After the Paris Agreement: measuring the global decarbonization wedges from national energy scenarios. Ecol Econ 150:273–289. https://doi.org/10.1016/j.ecolecon.2018.04.012

Matovic D (2011) Biochar as a viable carbon sequestration option: global and Canadian perspective. Energy 36:2011–2016. https://doi.org/10.1016/j.energy.2010.09.031

McGlashan N et al (2012) High-level techno-economic assessment of negative emissions technologies. Process Saf Environ Prot 90:501–510. https://doi.org/10.1016/j.psep.2012.10.004

McLaren D (2012) A comparative global assessment of potential negative emissions technologies. Process Saf Environ Prot 90:489–500. https://doi.org/10.1016/j.psep.2012.10.005

Michalski J et al (2019) The role of renewable fuel supply in the transport sector in a future decarbonized energy system. Int J Hydrog Energy 44:12554–12565. https://doi.org/10.1016/j.ijhydene.2018.10.110

Ming T et al (2014) Fighting global warming by climate engineering: is the Earth radiation management and the solar radiation management any option for fighting climate change? Renew Sustain Energy Rev 31:792–834. https://doi.org/10.1016/j.rser.2013.12.032

Nakajima T et al (2018) Carbon sequestration and yield performances of Miscanthus × giganteus and Miscanthus sinensis . Carbon Manag 9:415–423. https://doi.org/10.1080/17583004.2018.1518106

Nieto J et al (2018) Less than 2 °C? An economic-environmental evaluation of the Paris Agreement. Ecol Econ 146:69–84. https://doi.org/10.1016/j.ecolecon.2017.10.007

Olajire AA (2013) A review of mineral carbonation technology in sequestration of CO 2 . J Pet Sci Eng 109:364–392. https://doi.org/10.1016/j.petrol.2013.03.013

Oni BA et al (2020) Significance of biochar application to the environment and economy. Ann Agric Sci. https://doi.org/10.1016/j.aoas.2019.12.006

Osman AI (2020) Catalytic hydrogen production from methane partial oxidation: mechanism and kinetic study. Chem Eng Technol 43:641–648. https://doi.org/10.1002/ceat.201900339

Osman AI et al (2019) Production and characterisation of activated carbon and carbon nanotubes from potato peel waste and their application in heavy metal removal. Environ Sci Pollut Res 26:37228–37241. https://doi.org/10.1007/s11356-019-06594-w

Osman AI et al (2020a) Upcycling brewer’s spent grain waste into activated carbon and carbon nanotubes for energy and other applications via two-stage activation. J Chem Technol Biotechnol 95:183–195. https://doi.org/10.1002/jctb.6220

Osman AI et al (2020b) The production and application of carbon nanomaterials from high alkali silicate herbaceous biomass. Sci Rep 10:2563. https://doi.org/10.1038/s41598-020-59481-7

Østergaard PA et al (2020) Sustainable development using renewable energy technology. Renew Energy 146:2430–2437. https://doi.org/10.1016/j.renene.2019.08.094

Palmer C (2019) Mitigating climate change will depend on negative emissions technologies. Engineering 5:982–984. https://doi.org/10.1016/j.eng.2019.10.006

Perdan S, Azapagic A (2011) Carbon trading: current schemes and future developments. Energy Policy. 39:6040–6054. https://doi.org/10.1016/j.enpol.2011.07.003

Pfeiffer B, Mulder P (2013) Explaining the diffusion of renewable energy technology in developing countries. Energy Econ 40:285–296. https://doi.org/10.1016/j.eneco.2013.07.005

Pindilli E et al (2018) Estimating the societal benefits of carbon dioxide sequestration through peatland restoration. Ecol Econ 154:145–155. https://doi.org/10.1016/j.ecolecon.2018.08.002

Pires JCM (2019) Negative emissions technologies: a complementary solution for climate change mitigation. Sci Total Environ 672:502–514. https://doi.org/10.1016/j.scitotenv.2019.04.004

Pitelis A et al (2019) Fostering innovation in renewable energy technologies: choice of policy instruments and effectiveness. Renew Energy. https://doi.org/10.1016/j.renene.2019.11.100

Pleßmann G, Blechinger P (2017) How to meet EU GHG emission reduction targets? A model based decarbonization pathway for Europe’s electricity supply system until 2050. Energy Strategy Rev 15:19–32. https://doi.org/10.1016/j.esr.2016.11.003

Pourhashem G et al (2019) Policy support for biochar: review and recommendations. GCB Bioenergy 11:364–380. https://doi.org/10.1111/gcbb.12582

Prăvălie R, Bandoc G (2018) Nuclear energy: between global electricity demand, worldwide decarbonisation imperativeness, and planetary environmental implications. J Environ Manag 209:81–92. https://doi.org/10.1016/j.jenvman.2017.12.043

Purakayastha TJ et al (2019) A review on biochar modulated soil condition improvements and nutrient dynamics concerning crop yields: pathways to climate change mitigation and global food security. Chemosphere 227:345–365. https://doi.org/10.1016/j.chemosphere.2019.03.170

Qin Z et al (2020) CO2 reforming of CH4 to syngas over nickel-based catalysts. Environ Chem Lett 18:997–1017. https://doi.org/10.1007/s10311-020-00996-w

REN21 (2019) Renewables 2019—global status report. https://www.ren21.net/wp-content/uploads/2019/05/gsr_2019_full_report_en.pdf . Accessed 4 Feb 2020

Renforth P (2019) The negative emission potential of alkaline materials. Nat Commun 10:1401. https://doi.org/10.1038/s41467-019-09475-5

Renforth P, Henderson G (2017) Assessing ocean alkalinity for carbon sequestration. Rev Geophys 55:636–674. https://doi.org/10.1002/2016RG000533

Ricke KL et al (2017) Constraints on global temperature target overshoot. Sci Rep 7:14743. https://doi.org/10.1038/s41598-017-14503-9

RoyalSociety (2018) Greenhouse gas removal. https://royalsociety.org/-/media/policy/projects/greenhouse-gas-removal/royal-society-greenhouse-gas-removal-report-2018.pdf . Accessed 28 Jan 2020

Sapkota Y, White JR (2020) Carbon offset market methodologies applicable for coastal wetland restoration and conservation in the United States: a review. Sci Total Environ 701:134497. https://doi.org/10.1016/j.scitotenv.2019.134497

Sarkhot DV et al (2012) Total and available soil carbon fractions under the perennial grass Cynodon dactylon (L.) Pers and the bioenergy crop Arundo donax L. Biomass Bioenerg 41:122–130. https://doi.org/10.1016/j.biombioe.2012.02.015

Sarkodie SA, Strezov V (2019) Economic, social and governance adaptation readiness for mitigation of climate change vulnerability: evidence from 192 countries. Sci Total Environ 656:150–164. https://doi.org/10.1016/j.scitotenv.2018.11.349

Schmidt H-P et al (2019) Pyrogenic carbon capture and storage. GCB Bioenergy 11:573–591. https://doi.org/10.1111/gcbb.12553

Seetharaman et al (2019) Breaking barriers in deployment of renewable energy. Heliyon 5:e01166. https://doi.org/10.1016/j.heliyon.2019.e01166

Semida WM et al (2019) Biochar implications for sustainable agriculture and environment: a review. S Afr J Bot 127:333–347. https://doi.org/10.1016/j.sajb.2019.11.015

Shinnar R, Citro F (2008) Decarbonization: achieving near-total energy independence and near-total elimination of greenhouse emissions with available technologies. Technol Soc 30:1–16. https://doi.org/10.1016/j.techsoc.2007.10.006

Shivakumar A et al (2019) Drivers of renewable energy deployment in the EU: an analysis of past trends and projections. Energy Strategy Rev 26:100402. https://doi.org/10.1016/j.esr.2019.100402

Sinsel SR et al (2020) Challenges and solution technologies for the integration of variable renewable energy sources—a review. Renew Energy 145:2271–2285. https://doi.org/10.1016/j.renene.2019.06.147

Smith W, Wagner G (2018) Stratospheric aerosol injection tactics and costs in the first 15 years of deployment. Environ Res Lett 13:124001. https://doi.org/10.1088/1748-9326/aae98d

Srivastava P (2012) Soil carbon sequestration: an innovative strategy for reducing atmospheric carbon dioxide concentration. Biodivers Conserv 21(5):1343–1358. https://doi.org/10.1007/s10531-012-0229-y

Srivastava N et al (2017) Nanomaterials for biofuel production using lignocellulosic waste. Environ Chem Lett 15:179–184. https://doi.org/10.1007/s10311-017-0622-6

Srivastava RK et al (2020) Biofuels, biodiesel and biohydrogen production using bioprocesses. A review. Environ Chem Lett 18:1049–1072. https://doi.org/10.1007/s10311-020-00999-7

Strefler J et al (2018) Potential and costs of carbon dioxide removal by enhanced weathering of rocks. Environ Res Lett 13:034010. https://doi.org/10.1088/1748-9326/aaa9c4

Su T-M et al (2016) Recent advances in the photocatalytic reduction of carbon dioxide. Environ Chem Lett 14:99–112. https://doi.org/10.1007/s10311-015-0528-0

Tcvetkov P et al (2019) Public perception of carbon capture and storage: a state-of-the-art overview. Heliyon 5:e02845. https://doi.org/10.1016/j.heliyon.2019.e02845

UN (1992) United nations framework convention on climate change. https://unfccc.int/files/essential_background/background_publications_htmlpdf/application/pdf/conveng.pdf . Accessed 21 Jan 2020

UN (2015) Paris agreement. https://unfccc.int/files/essential_background/convention/application/pdf/english_paris_agreement.pdf . Accessed 24 Jan 2020

UNCCS (2019) Climate action and support trends, United Nations Climate Change Secretariat. https://unfccc.int/sites/default/files/resource/Climate_Action_Support_Trends_2019.pdf . Accessed 20 Dec 2019

UNEP (2019) Emissions gap report. UN Environment Program, Nairobi. https://wedocs.unep.org/bitstream/handle/20.500.11822/30797/EGR2019.pdf?sequence=1&isAllowed=y . Accessed 22 Dec 2019

UNFCCC (1997) Kyoto Protocol to the United Nations Framework Convention on Climate Change, https://unfccc.int/sites/default/files/resource/docs/cop3/l07a01.pdf . Accessed 21 Jan 2020

UNFCCC (2005) Report of the convergence of the parties serving as the meeting of the parties to the kyoto protocol on its first session. https://unfccc.int/resource/docs/2005/cmp1/eng/08a02.pdf . Accessed 23 Jan 2020

UNFCCC (2008) Kyoto protocol reference manual on accounting of emissions and assigned amount. https://unfccc.int/resource/docs/publications/08_unfccc_kp_ref_manual.pdf . Accessed 22 Dec 2019

UNFCCC (2012) Doha amendment to the kyoto protocol. https://treaties.un.org/doc/Publication/CN/2012/CN.718.2012-Eng.pdf . Accessed 24 Jan 2020

Victor N et al (2018) The U.S. power sector decarbonization: investigating technology options with MARKAL nine-region model. Energy Econ 73:410–425. https://doi.org/10.1016/j.eneco.2018.03.021

Villa JA, Bernal B (2018) Carbon sequestration in wetlands, from science to practice: an overview of the biogeochemical process, measurement methods, and policy framework. Ecol Eng 114:115–128. https://doi.org/10.1016/j.ecoleng.2017.06.037

Villoria-Sáez P et al (2016) Effectiveness of greenhouse-gas emission trading schemes implementation: a review on legislations. J Clean Prod 127:49–58. https://doi.org/10.1016/j.jclepro.2016.03.148

Vinca A et al (2018) The role of carbon capture and storage electricity in attaining 1.5 and 2 °C. Int J Greenh Gas Control 78:148–159. https://doi.org/10.1016/j.ijggc.2018.07.020

Wang Q et al (2016) Distributional effects of carbon taxation. Appl Energy 184:1123–1131. https://doi.org/10.1016/j.apenergy.2016.06.083

Wang B et al (2020) CO2 sequestration: high conversion of gypsum into CaCO 3 by ultrasonic carbonation. Environ Chem Lett 18:1369–1377. https://doi.org/10.1007/s10311-020-00997-9

WEF (2020) Global risks report 2020, World Economic Forum. http://www3.weforum.org/docs/WEF_Global_Risk_Report_2020.pdf . Accessed 18 Jan 2020

Wendling ZA (2019) Bridges beyond renewable energy: decarbonizing the global electricity sector under uncertainty. Energy Res Soc Sci 48:235–245. https://doi.org/10.1016/j.erss.2018.09.020

Williamson P et al (2012) Ocean fertilization for geoengineering: a review of effectiveness, environmental impacts and emerging governance. Process Saf Environ Prot 90:475–488. https://doi.org/10.1016/j.psep.2012.10.007

WMO (1979) Proceedings of the world climate conference—a conference of experts on climate and mankind, World Meteorological Organization. https://library.wmo.int/doc_num.php?explnum_id=8346 . Accessed 21 Jan 2020

Xiao L et al (2019) Biochar promotes methane production at high acetate concentrations in anaerobic soils. Environ Chem Lett 17:1347–1352. https://doi.org/10.1007/s10311-019-00863-3

Xu M et al (2019) Biochar addition to soil highly increases P retention and decreases the risk of phosphate contamination of waters. Environ Chem Lett 17:533–541. https://doi.org/10.1007/s10311-018-0802-z

Yan J et al (2019) Negative emission technologies NETs. Appl Energy 255:113749. https://doi.org/10.1016/j.apenergy.2019.113749

Yue X-L, Gao Q-X (2018) Contributions of natural systems and human activity to greenhouse gas emissions. Adv Clim Change Res 9:243–252. https://doi.org/10.1016/j.accre.2018.12.003

Zhang Z et al (2015) Review of geoengineering approaches to mitigating climate change. J Clean Prod 103:898–907. https://doi.org/10.1016/j.jclepro.2014.09.076

Download references

Acknowledgements

Authors would like to acknowledge the support given by the EPSRC project “Advancing Creative Circular Economies for Plastics via Technological-Social Transitions” (ACCEPT Transitions, EP/S025545/1). The authors wish to acknowledge the support of The Bryden Centre project (Project ID VA5048) which was awarded by The European Union’s INTERREG VA Programme, managed by the Special EU Programmes Body (SEUPB), with match funding provided by the Department for the Economy in Northern Ireland and the Department of Business, Enterprise and Innovation in the Republic of Ireland.

Author information

Authors and affiliations.

School of Chemistry and Chemical Engineering, Queen’s University Belfast, David Keir Building, Stranmillis Road, Belfast, BT9 5AG, Northern Ireland, UK

Samer Fawzy, Ahmed I. Osman & David W. Rooney

The Bryden Centre, Letterkenny Institute of Technology, Letterkenny, Ireland

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Ahmed I. Osman .

Ethics declarations

Conflict of interest.

The author declares no competing financial interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Fawzy, S., Osman, A.I., Doran, J. et al. Strategies for mitigation of climate change: a review. Environ Chem Lett 18 , 2069–2094 (2020). https://doi.org/10.1007/s10311-020-01059-w

Download citation

Received : 04 July 2020

Accepted : 17 July 2020

Published : 30 July 2020

Issue Date : November 2020

DOI : https://doi.org/10.1007/s10311-020-01059-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Climate change mitigation
• Carbon dioxide removal
• Decarbonization technologies
• Find a journal
• Publish with us
• Track your research

The Adaptation Principles: 6 Ways to Build Resilience to Climate Change

STORY HIGHLIGHTS

• Climate risk cannot be reduced to zero, which means governments must take decisive action to help households and businesses manage them.
• A new World Bank report, “The Adaptation Principles: A Guide for Designing Strategies for Climate Change Adaptation and Resilience”, lays out 6 universal principles to help policymakers plan for adaptation…
• … Along with 26 actions, 12 tool boxes and 111 indicators.

Over the past decades, Uganda made remarkable progress in reducing poverty and boosting socio-economic development. In 1992, some 56 percent of the population was living in poverty. By 2016, that figure had fallen to 21 percent . Yet, the global economic ramifications of the COVID-19 pandemic and the effects of climate change are forcing the country to confront new challenges: shocks not only threaten further progress but can reverse hard won successes of the past.

Around 72 percent of Uganda’s labor force works in agriculture – a sector that is highly climate sensitive. Take coffee: Uganda is Africa’s second largest exporter of coffee. Over 17 percent of Uganda’s exports coming from just this high-value crop. Recent droughts, however, are estimated to have destroyed half of all coffee yields. In the coming decades, changing climatic conditions are expected to pose profound challenges to Uganda’s coffee sector : without adaptive measures, only 1 percent of Uganda’s current coffee producing land is expected to be able to continue production. And coffee is just one sector that could face mounting impacts from climate change: around 2.3 million poor people in Uganda also face high levels of flood risk.

In countries around the world, climate change poses a significant risk threatening the lives and livelihoods of people. These risks cannot be reduced to zero, which means governments must take decisive action to help firms and people manage them. Doing so requires planning ahead and putting in place proactive measures that not only reduce climate risk but also accelerate development, and cut poverty, according to a new report, The Adaptation Principles: A Guide for Designing Strategies for Climate Change Adaptation and Resilience .

“Adaptation cannot be an afterthought to development. Instead, by integrating it into policy thinking up front, governments can catalyze robust economic development while also reducing vulnerability to climate change,” says Lead Economist, Stéphane Hallegatte , who co-authored the report with Jun Rentschler and Julie Rozenberg, all of the World Bank.

The report lays out six universal “Principles of Adaptation and Resilience” and 26 concrete actions that governments can use to develop effective strategies. To support the development and design of these actions, it also includes 12 toolboxes with methodologies and data sources that can ensure that strategies are evidence-based.  

1. Build resilient foundations with rapid and inclusive development

Poverty and the lack of access to basic services—including infrastructure, financial services, health care, and social protection—are strong predictors of vulnerability to climate change . To put it another way: the poorer communities are, the more climate change will affect them. No adaptation strategy can be successful without ensuring high-vulnerability populations have the financial, technical, and institutional resources they need to adapt.

2. Help people and firms do their part.

It’s critical to boost the adaptive capacity of households and firms: many already have incentives to adapt, but they need help overcoming obstacles, ranging from a lack of information and financing, to behavioral biases and imperfect markets. Governments can make information on climate risks available, clarify responsibilities and liabilities, support innovation and access to the best technologies , and ensure financing is available to all especially for solutions that come with high upfront costs. And they will also need to provide direct support to the poorest people, who cannot afford to invest in adaptation but are the most vulnerable to experiencing devastating effects of climate change .

3. Revise land use plans and protect critical infrastructure.

In addition to direct support to households and businesses, governments must also play a role in protecting public investments, assets, and services. Power and water outages and transport disruptions are estimated to cost more than $390 billion per year already in developing countries. But if countries have the right data, risk models, and decision-making methods available, the incremental cost of building the resilience of new infrastructure assets is small—only around 3 percent of total investments. Urban and land use plans are also important responsibilities of the public sector, and they influence massive private investments in housing and productive assets, so it is vital these adapt to evolving long-term climate risks to avoid locking people into high-risk areas.

4. Help people and firms recover faster and better.

Risks and impacts cannot be reduced to zero. Governments must develop strategies to ensure that when disasters do occur, people and firms can cope without devastating long-term consequences, and can recover quickly. Preparation such as better hydromet data , early warning and emergency management systems reduces physical damage and economic losses—for example, shuttering windows ahead of a hurricane can reduce damage by up to 50 percent. The benefits of providing universal access to early warning systems globally have been repeatedly found to largely exceed costs, by factors of at least 4 to 10 . And then, financial inclusion, such as access to emergency borrowing, and social protection are essential ways to help firms and people get back on their feet. Adaptive social protection systems , which can be rapidly scaled up to cover more people and provide bigger support after a disaster, are particularly efficient, but they rely on delivery and finance mechanisms that have to be created before a crisis occurs.

5. Manage impacts at the macro level.

Coping with climate change impacts in one economic sector is already complicated. Coping with climate change impacts in all sectors at once requires strategic planning at the highest levels. Through many impacts in many sectors ---  from floods affecting housing prices to changes in ecosystems affecting agriculture productivity --- climate change will affect the macroeconomic situation and tax revenues. Some impacts on major sectors (especially exporting ones) can affect a country’s trade balance and capital flows. And spending needs for adaptation and resilience need to be added on top of existing contingent liabilities and current debt levels to create further pressure on public finances. The combination of these factors may result in new risks for macroeconomic stability, public finances and debt sustainability, and the broader financial sector. Governments will need to manage these risks . Because of the massive uncertainty that surrounds macroeconomic estimates of future climate change impacts, strategies to build the resilience of the economy, especially through appropriate diversification of the economic structure, export composition and tax base, are particularly attractive over the short term.

6. Prioritize according to needs, implement across sectors and monitor progress.

Governments must not only prioritize actions to make them compatible with available resources and capacity; they must also establish a robust institutional and legal framework , and a consistent system for monitoring progress. The main objective of an adaptation and resilience strategy is not to implement stand-alone projects: it is to ensure that all government departments and public agencies adopt and mainstream the strategy in all their decisions, and that governments continuously monitor and evaluate the impact of their decisions and actions, so they can address any challenges and adjust their actions accordingly.

The report provides a range of practical tools that can help governments implement adaptation strategies. For instance, economic analysis methodologies can help to select the most important interventions, and budget tagging methods can ensure spending is consistent with expectations. A set of 111 indicators is also provided to enable governments to track progress toward greater resilience, to identify areas that are lagging behind, and to prioritize effective measures. It also sheds light on how the COVID-19 pandemic and subsequent economic crisis can affect the design of an adaptation and resilience strategy, recognizing how it has changed the development landscape in all countries.

The impacts of climate change are already here and fast increasing and there is no silver bullet to prevent them. Proactive and robust actions ahead of time, however, can go a long way to helping people and communities so that when a natural disaster strikes, not only are they better prepared to respond, but hard-won development gains are not lost.

Join us on Tuesday, December 1 2020, for a discussion on the main findings of this report .

“The Adaptation Principles: A Guide for Designing Strategies for Climate Change Adaptation and Resilience” was produced with financial support from the Global Facility for Disaster Reduction and Recovery .

• Report: The Adaptation Principles - A Guide for Designing Strategies for Climate Change Adaptation and Resilience
• Infographic: The Adaptation Principles at a Glance

This site uses cookies to optimize functionality and give you the best possible experience. If you continue to navigate this website beyond this page, cookies will be placed on your browser. To learn more about cookies, click here .

ORIGINAL RESEARCH article

A framework for methodological options to assess climatic and anthropogenic influences on streamflow.

• 1 State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Nanjing Hydraulic Research Institute, Nanjing, China
• 2 China Energy Investment Corporation Science and Technology Research Institute Co.,Ltd., Nanjing, China

Climate change and human activities are having increasing impacts on the global water cycle, particularly on streamflow. Current methods for quantifying these impacts are numerous and have their merits and limitations. There is a lack of a guide to help researchers select one or more appropriate methods for attribution analysis. In this study, hydrological modeling, statistical analysis, and conceptual approaches were used jointly to develop a methodological options framework consisting of three modules, to guide researchers in selecting appropriate methods and assessing climatic and anthropogenic contributions to streamflow changes. To evaluate its effectiveness, a case study in the Upper Yangtze River Basin (UYRB) of China was conducted. The results suggest that the SWAT-based method is the best approach to quantify the influences of climate change and human activities on streamflow in the UYRB. The comprehensive assessment indicates that climate change is the dominant cause of streamflow changes in the UYRB, and the contribution of climate change, indirect human activities, and direct human activities to streamflow changes is about 7:1:2. The proposed framework is efficient and valuable in assisting researchers to find appropriate methods for attribution analysis of streamflow changes, which can help to understand the water cycle in changing environments.

Introduction

Streamflow is one of the most important elements of the hydrological cycle and is key to understanding hydrological processes at various spatial and temporal scales under changing environments ( Penny et al., 2020 ; Jiang et al., 2021 ; Porter et al., 2021 ; Wright et al., 2021 ). However, streamflow has been significantly altered by the combined effects of climate change and human activities ( Liu et al., 2019 ; Yasarer et al., 2020 ; Zeng et al., 2021 ). Changes in two important climatic variables, precipitation, and potential evapotranspiration, jointly influence the spatial and temporal distribution patterns of water resources ( Borgwardt et al., 2020 ). The sixth assessment report of the IPCC (Intergovernmental Panel on Climate Change) notes that global mean precipitation and evaporation increase with global warming (high confidence). This will undoubtedly accelerate the change in streamflow. Human activities, including land use/land cover changes, water consumption, urbanization, and dam construction, have also altered hydrological processes in many areas ( Zhao et al., 2012 ; Jardim et al., 2020 ). It is worth noting that streamflow changes may have a significant impact on water use patterns in different sectors such as agriculture, domestic, industry, environment, and hydropower generation ( Zhang et al., 2021b ). Therefore, it is important to clarify streamflow changes and their drivers for water resources management under the changing environment.

The existing common methods used to quantitatively estimate the impacts of climate change and human activities on streamflow can be grouped into three categories: hydrological modeling, statistical analysis, and conceptual approaches ( Wang, 2014 ; Dey and Mishra, 2017 ). The hydrological modeling method is based on various hydrological models, such as SWAT (Soil and Water Assessment Tool) model, GBHM (Geomorphology-Based Hydrological Model), and VIC (Variable Infiltration Capacity) model, to simulate the runoff process in a watershed and then analyze the influences of climate change and human activities on streamflow ( Dey and Mishra, 2017 ; Hajihosseini et al., 2020 ). For instance, Zhang et al. (2021a) employed SWAT to evaluate the impacts of climate change and human activities on streamflow changes in the Upstream Yangtze River. The results showed that the main contributions to runoff change are 70% from climate change and 30% from human influence. However, the uncertainty introduced by the parameter estimations, structure, and input data of the model may lead to inaccurate results.

Statistical analysis methods are based on the observed streamflow and meteorological factors, and the contributions of meteorological factors to streamflow changes are calculated by statistical analysis, which requires a certain length of the observations series, and usually the longer the series, the better the results are obtained ( Wang, 2014 ; Dey and Mishra, 2017 ). The common methods include multiple regression models and climate elasticity models, which have been widely applied in many case studies ( Miao et al., 2011 ; Li et al., 2014 ). For instance, Zhang et al. (2020) used the simple linear regression method and sensitivity indicator method to separate impacts of climate variabilities and human activities on streamflow changes in a typical semi-arid basin (Guanting River Basin in China). Nevertheless, these methods have certain limitations. For example, multiple regression models cannot capture the nonlinear characteristics of streamflow changes, and climate elasticity models cannot quantify the effects of extreme hydroclimatic variability and human activities on streamflow changes.

The conceptual approaches mainly include those based on the Budyko framework and those based on the Tomer Schilling framework. The former uses estimated precipitation and potential evapotranspiration elasticity to assess the impact of climate change on streamflow changes, which has certain physical mechanisms, while the calculation is relatively simple and convenient, and is widely used ( Li et al., 2020 ; Todhunter et al., 2020 ; Wang et al., 2020 ). For example, Liu et al. (2021) used the Budyko framework for attribution analysis of global runoff changes and found that other factors than precipitation and potential evaporation are the most significant drivers to global streamflow changes based on observed data. Tomer and Schilling (2009) proposed a conceptual approach that can distinguish the relative contributions of climate change and human activities to the streamflow changes based on a 25-years experiment in a small watershed, namely the Tomer Schilling framework. This method can only distinguish the relative contribution of climate change and human activities, and requires more data on precipitation and actual evapotranspiration ( Wang, 2014 ). Also, Renner et al. (2012) noted that the method does not adhere to the water and energy limits. Thus, this method is not as widely used as the former.

As mentioned above, each method has its merits and limitations. Therefore, it is more valuable to combine and compare the results of multiple methods than a single method. At present, few studies have sorted out and integrated these methods and analyzed and identified the attribution of streamflow changes in the form of a framework. For instance, Wang (2014) compared currently-used methods to analyze the attribution of streamflow changes, assess assumptions and issues of the methods and provided a framework for gauged watersheds. However, the framework is limited to watersheds with negligible groundwater loss, and there is no application case. Also, Zhang et al. (2020) proposed a comprehensive assessment framework, including six Budyko-framework-based methods, hydrological simulation, sensitivity indicator method, and empirical statistics, to separate impacts of climate variability and human activity on streamflow. The difference between the ten attribution methods varied between 5 and 12% in their study, which showed that the multi-method framework can effectively avoid overestimation/underestimation and quantitative uncertainty. However, some of the methods involved in the framework may be inappropriate for some basins, which may easily lead to inaccurate evaluation. Moreover, the methods for attribution analysis of streamflow changes have some criteria and assumptions, which undoubtedly limit their generality ( Dey and Mishra, 2017 ). Therefore, it is crucial to choose a suitable method for attribution analysis of streamflow changes.

Therefore, the main aim of this study is to propose a methodological options framework based on hydrological modeling, statistical analysis, and conceptual approach, which is to guide researchers in selecting appropriate methods and comprehensively assessing the impacts of climate change and human activities on streamflow. Specifically, the framework includes three modules: data preparation (Module Ⅰ), streamflow changes analysis and key driver identification (Module Ⅱ), and attribution analysis method selection and comprehensive assessment (Module Ⅲ). Also, the Upper Yangtze River Basin (UYRB) of China was taken as a case study to confirm the effectiveness of the proposed framework. The rest of this paper is organized as follows. Methodology clarifies the details of methods, including the proposed framework, fundamental techniques, and critical criteria and assumptions. Case Study provides the details of the case study. Results shows the results of the streamflow changes and their attribution analysis. Discussion discusses the results of the three methods and the implications of the proposed framework. Finally, Conclusion presents the conclusions of this study, summarizing the results of the comprehensive analysis and significance of the framework.

Methodology

Overview of the framework.

In this study, we proposed a framework for methodological options to assess climatic and anthropogenic influences on streamflow ( Figure 1 ), which is mainly composed of three modules: (Ⅰ) data preparation, (Ⅱ) streamflow change analysis and key driver identification, and (Ⅲ) attribution analysis method selection and comprehensive assessment.

FIGURE 1 . Flowchart of the framework for methodological options to assess climatic and anthropogenic influences on streamflow.

Data Preparation

First, hydro-meteorological data of the study area were collected. Reference evapotranspiration (ET 0 ) is calculated for each meteorological station by the Penman-Monteith equation recommended by FAO (Food and Agriculture Organization), and the area-averaged precipitation and ET 0 are calculated. The streamflow dataset, precipitation dataset, and ET 0 dataset of the study area are constructed. At the same time, data on human activities were also collected and combined with data on meteorological elements to form a dataset regarding the drivers of streamflow changes.

Streamflow Change Analysis and Key Driver Identification

First, based on the dataset of streamflow, precipitation and ET 0 generated in Module I, time series analysis methods were used to clarify the trends in streamflow and other water cycle elements in the study area, and check whether the streamflow changes have a major abrupt change point, and whether the critical assumption of “time series is long enough” is satisfied, if not, it is considered that a longer time series is needed to assess climatic and anthropogenic influences on streamflow, but if it is satisfied, the steps continue. Second, the study period is divided into base period and change period according to the main abrupt change point of streamflow. Trends in streamflow, precipitation, ET 0 before the abrupt change point are tested to check whether the critical assumption of “base period and change period” is met, if not, the abrupt change point is considered to be lagging behind and needs to be re-divided into base and change periods, if it is met, two sub-periods are determined and the basin streamflow changes in the two sub-periods are calculated. Third, the drivers are analyzed for significant changes between the base and change periods, and the drivers with significant changes are selected to form the key driver set.

Attribution Analysis Method Selection and Comprehensive Assessment

In this study, the SWAT-based method, climate-elasticity-based method and Choudhury-Yang-equation-based method were selected as representatives of the hydrological modeling methods, conceptual methods and statistical methods, respectively. First, based on the research needs and identified key drivers, it is determined whether quantitative assessment of anthropogenic contributions is required; if not, then the climate-elasticity-based method is sufficient, and if so, then the Choudhury-Yang-equation-based or SWAT-based method is selected. Second, to determine whether the quantitative assessment for contributions of direct human activities is needed, if not, the Choudhury-Yang equation-based method is satisfied, if needed, the SWAT-based method is selected. Third, to determine whether the study basin satisfies the assumption of “negligible change in water storage”, if not, the SWAT-based method is selected, and if so, the climate-elasticity-based method and Choudhury-Yang-equation based can be used.

Basic Methods

Time series analysis methods.

In this framework, time series analysis was used to detect the temporal trends and abrupt changes in hydro-meteorological series. Since different statistical methods may produce different results, four time series analysis methods were selected in this study, i.e. , the Mann-Kendall test ( Mann, 1945 ; Kendall, 1975 ), the Spearman’s Rho test ( Lehmann, 1975 ), the Pettitt’s test ( Pettitt, 1979 ), the sequential clustering method ( Dubes and Jain, 1980 ). The non-parametric Mann-Kendall test is highly recommended by the WMO (World Meteorological Organization) and widely used to detect monotonic trends in long-term hydro-meteorological variations. The Spearman’s Rho test is a quick and simple test to determine whether correlation exists between two sub-series in the same series of observations. Both the Mann-Kendall test and the Spearman’s Rho test are used to identify trends in hydro-meteorological series and evaluate the significance of the trends. The non-parametric Pettitt’s test, same as the Mann-Kendall test, is a rank-based and distribution-free test. It is used to identify the occurrence of an abrupt change point in hydro-meteorological series. The sequential clustering method is used to detect the abrupt change point as a reconfirmation for Pettitt’s test in this study.

Attribution Analysis Methods

Generally, the study period is divided into two sub-periods based on historical streamflow changes, i.e., baseline period and change period. The baseline period is affected by the changing environment to a negligible extent, and streamflow is in a natural or quasi-natural state. The streamflow during the change period, on the other hand, is affected by climate change and human activities, and undergoes significant changes under the influence. Based on this, the principles of the currently used attribution analysis methods can be summarized as follows:

where Δ R is the streamflow change; R ¯ a p is the average annual streamflow during change period; R ¯ b p is the average annual streamflow during baseline period. The streamflow change is considered to be caused by the combined effects of climate change and human activities, and their contributions are denoted as Δ R C and Δ R H . R ¯ a p C and R ¯ a p H are the average annual streamflow during change period influenced by climate change only and by human activities only. The contributions of climate change and human activities to streamflow change are estimated by R ¯ a p C and R ¯ a p H minus the average annual streamflow during baseline period R ¯ b p .

Equations 1 – 3 are the general ideas of attribution analysis methods, which are realized in different ways in different methods. The climate-elasticity-based method, Choudhury-Yang-equation-based method and SWAT-based method are used in this study to further clarify the implementation of the basic ideas of attribution analysis.

1) Climate-elasticity-based method

The climate elasticity model ( Schaake, 1990 ) is a classical approach that was originally proposed to estimate the effect of only one driver ( i.e. , precipitation) on streamflow, and the model can be expressed as:

where Δ R i and Δ P i are the deviations of streamflow and precipitation in the i th year from the average annual value R ¯ and P ¯ . ε R P is the elasticity coefficient of streamflow with respect to precipitation.

After this, a two-parameter model incorporating precipitation and temperature ( Fu et al., 2007 ) was proposed and can be expressed as:

where Δ T is deviations of temperature in the i th year from the average annual value T ¯ . ε R T is the elasticity coefficient of streamflow with respect to temperature.

Later, more meteorological factors were introduced and climate elasticity models with more parameters were established ( Yang and Yang, 2011 ). After fitting the elasticity coefficients based on long series of observations, the contribution of each meteorological factor can be estimated. It can be seen that the more complete and longer the long series of observations of streamflow and each meteorological factor, the better the fitting effect and the more accurate the attribution analysis results.

2) Choudhury-Yang-equation-based method

The Budyko hypothesis ( Budyko, 1974 ) can be expressed as the following equation:

where E is the average annual actual evapotranspiration for a basin; P is the average annual precipitation; E 0 is the average annual potential evapotranspiration.

The function F (·) in Eq. 6 is considered to have a general expression, and many scholars have proposed formulas without parameters, with one parameter or with two parameters. Among them, the Choudhury-Yang equation ( Yang et al., 2008 ) is able to describe the interactions between climate, hydrology and watershed with the following expression:

where n is a parameter related to the land surface. Using E = P − R to estimate the E , n can be calculated from E , P and E 0 .

For a closed watershed, the change in the multi-year average water storage can be neglected, and the long-term water balance can be expressed as:

The streamflow can be written as a function of precipitation ( P ), potential evapotranspiration ( E 0 ) and land surface parameter ( n ):

Further the streamflow change can be written in the following differential form:

where ε P = ∂ f ∂ P P R , ε E 0 = ∂ f ∂ E 0 E 0 R , ε n = ∂ f ∂ n n R are elasticity coefficients of precipitation, potential evapotranspiration and land surface parameter.

Finally, the contributions of drivers to streamflow change are estimated as follows:

where Δ R p , Δ R E 0 , Δ R n are the contributions of precipitation P , potential evapotranspiration E 0 and land surface parameter n ; Δ P , Δ E 0 , Δ n are the differences between the three drivers in the baseline period and the change period.

3) SWAT-based method

SWAT ( Arnold et al., 1998 ) is a widely used, physically based distributed hydrological model for simulating surface streamflow, groundwater discharge, evapotranspiration, soil water content ( Zhang et al., 2021a ). It and has been certified as an effective tool for evaluating water resources at a wide range of scales. A two-driver example is given to clarify the implementation of the basic ideas of attribution analysis in the SWAT-based method.

The driving factors are first identified as climate change and human activities, which are expressed as C and H respectively, while the model inputs are abbreviated as C and H . The streamflow output of the model is written as a function of C and H . Then the streamflow change can be expressed in the form of Eq. 1 as:

where Δ W is the streamflow change; W ( C a p , H a p ) ¯ and W ( C b p , H b p ) ¯ are the average annual streamflow simulated by SWAT during the change period and baseline period; ap and bp represent the change period and baseline period; C a p and H a p are the model inputs of climate and human activities during the change period; C b p and H b p are the model inputs of climate and human activities during the baseline period.

Contributions of climate change and human activities to streamflow change can be expressed in the form of Eqs 2 , 3 as

where W ( C a p , H b p ) ¯ is the average annual streamflow simulated with model inputs of climate during the baseline period and human activities during the change period.

Critical Criteria and Assumptions

Baseline period and change period.

The assumption that “the study period can be divided into a baseline period and a change period” is used in all the above three methods. It assumes streamflow is in a natural or quasi-natural state during the baseline period and undergoes significant changes during the change period because of the influence of climate change and human activities. Determining the baseline period and change period is the first step in attribution analysis. The time series analysis was used to detect the abrupt change point of the streamflow series, and the sub-period before the abrupt change point is recognized as the baseline period, and the sub-period after the abrupt change point is recognized as the change period. Commonly used methods include double cumulative curve, Pettitt test, Mann-Kendall test, sliding t -test, ordered clustering, etc . Two problems may arise when using the abrupt change point in streamflow to divide the base period and the change:

1) The abrupt change point in streamflow is later than the cut-off point between the baseline period and the change period. This occurs because human activities and climate change have opposite effects on the increase or decrease in streamflow between the cut-off point and the abrupt change, which cancel each other out at a certain statistical confidence level. From the general knowledge of water cycle, there should be no obvious trend of streamflow, precipitation and potential evapotranspiration in the natural period of a basin. Therefore, the accuracy of the baseline period determination is verified by detecting the trend of streamflow, precipitation and potential evapotranspiration in the baseline period.

2) For sufficiently long streamflow series, the time series analysis sometimes detects multiple abrupt change points, which are then divided into two cases: (a) Several abrupt change points with relatively low confidence level are detected after a major one. In this case, only the main abrupt change point is selected. (b) Multiple abrupt change points are scattered over the study period and it is difficult to determine which is the major one. In this case, the abrupt change points near the beginning and end of the study period are not included in the reliable points because the uncertainty is too large. The remaining points are filtered out one by one according to the streamflow changes or precipitation and potential evapotranspiration changes during sub-periods to find the major abrupt change point.

Impacts of Human Activities

At the watershed scale, human activities can be grouped into two categories: 1) human activities that directly affect streamflow (direct human activities), such as off-channel water use, damming, reservoir storage and discharge, and other activities that act directly on streamflow and have an immediate impact on streamflow once they are implemented. Direct human activities can increase or decrease streamflow, and for a single activity, the effect on streamflow is obvious and can be measured directly. However, as watersheds become larger and river systems increase, direct human activities will become more complex and difficult to observe comprehensively. 2) Human activities that indirectly affect streamflow (indirect human activities), such as land use change, soil and water conservation, urbanization, etc. The impact of an individual indirect human activity on the increase or decrease of streamflow is not obvious for a short duration, and usually needs to be accumulated for a certain period to produce a measurable impact. As the watershed grows larger, the impact of indirect human activities is instead more easily estimated with the help of remote sensing and other technical means.

The climate-elasticity-based method is only able to calculate the contribution of climate change to streamflow change. If the streamflow change minus climatic contribution is attributed to anthropogenic contribution, then the anthropogenic contribution contains an error term, and it is not able to distinguish between contributions of direct and indirect human activities. The Choudhury-Yang-equation-based method attributes streamflow changes exclusively to precipitation, potential evapotranspiration and land surface coefficient, i.e., to climate change and indirect human activities. The SWAT model allows for reservoirs and water extraction points in addition to inputs such as climate, soil, and land use. Therefore, the SWAT-based method is able to assess the effects of climate change, indirect human activities and direct human activities on streamflow. When the method is applied to a large watershed, the direct human activities are always too complex, so streamflow change minus contributions of climate change and indirect human activities is attributed to the effects of direct human activities, but an error term is included.

Length of Time Series

All three methods implicitly assume the length of the time series. That is, it is assumed that “the time series of streamflow is long enough”. 1) The time series is long enough to include natural or quasi-natural period and periods of change, and to include natural and changing characteristics of streamflow. 2) It is long enough to distinguish between gradual and abrupt changes in streamflow, to detect the points of abrupt change, and to identify changes in climatic elements. 3) It is sufficient to reflect the impact of indirect human activities on streamflow. In practice, for a watershed where streamflow is influenced by changing environment, a streamflow series is considered long enough for attribution analysis if it can be detected with significant trend changes and abrupt change points. Therefore, analyzing the characteristics of streamflow changes, detecting the abrupt change points, and analyzing the characteristics of driver changes are also tests of the assumption that the time series length is sufficiently long. In addition, as far as the attribution analysis method is concerned, the time series length is long enough to facilitate the establishment of a mapping relationship between driving factors and streamflow, thus making the attribution analysis more accurate.

Changes in Watershed Water Storage

The Choudhury-Yang-equation-based method uses a simplified water balance equation for a watershed ( Eq (8) . The simplified water balance equation is based on the assumption that “the change in the multi-year average water storage in a closed watershed is negligible”. However, if there is a significant decrease in streamflow and an increase in water storage in a study basin, attribution analysis under this assumption tends to exaggerate the contribution of climate change and underestimate the impact of human activities, while other cases may also produce biased attribution results. The climate-elasticity-based method aims to establish statistical relationships between streamflow and drivers. The watershed water storage is difficult to measure directly, so the climate-elasticity-based method is not able to take into account changes in water storage. As the SWAT model takes into account the watershed water storage and deep groundwater loss, the SWAT-based method can estimate the influence of changes in watershed water storage.

The Upper Yangtze River Basin (UYRB) was selected as a case study area to quantify the contributions of climate change and human activities to its streamflow changes and confirm the effectiveness of the proposed framework. The Yangtze River is the longest river in Asia and the third-longest river in the world. The average annual streamflow of the Yangtze River is 996 km 3 , making up 36% of the total streamflow in China, and the hydropower generation accounts for about 40% of the national total. The UYRB is located in the region between 90 and 112°E and 23 and 35°N, with a total area of 1 million km 2 ( Figure 2 ). The UYRB is the streamflow formation area and hydropower storage area of the Yangtze River, accounting for 45% of basin water resources and 90% of basin hydropower resources (with a developable installed capacity of about 20 × 10 8  kW). It has a mean annual temperature of 16.8°C, mean annual precipitation of 1,130 mm. However, measured streamflow in most regions of the UYRB has been decreasing heavily in recent years, which greatly impacts economic development, social stability, and ecological security ( Zhang et al., 2018 ). Therefore, it is necessary to conduct an attribution analysis of the streamflow changes in the basin.

FIGURE 2 . Map of the UYRB.

Data Processing (Module Ⅰ)

Daily streamflow series (1951–2013) from hydrological stations located at the mainstream and important tributaries ( Figure 2 ) were collected from the China Annual Hydrological Reports. Meteorological data were collected from the National Meteorological Information Centre of China (NMIC), including daily observations of precipitation, air temperatures, relative humidity, wind speed, and sunshine duration for the period from 1951 to 2013. The topography is represented using the digital elevation model (DEM) with a spatial resolution of 90 m, which was obtained from the Consultative Group on International Agricultural Research (CGIAR) Consortium for Spatial Information (CGIAR-CSI) ( http://srtm.csi.cgiar.org ). Geological characteristics data includes land use/cover data and soil type data. Land use/cover maps with a scale of 1: 100000 for 1980, 1990, 1995, 2000, 2005, 2010, and 2013 were collected from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC) ( http://www.resdc.cn ). ET 0 was calculated for each meteorological station by the FAO Penman-Monteith equation. The solar radiation was calculated using the Angstrom formula ( Wang et al., 2021 ), in which the Angstrom coefficients were estimated by the measured radiation within the UYRB ( Figure 2 ). The area-averaged precipitation and ET 0 were further calculated. The streamflow dataset, precipitation dataset, and ET 0 dataset of the UYRB were constructed and combined with data on human activities to form a dataset on the drivers of streamflow changes. At this point the data preparation in Module Ⅰ is complete.

Long-Term Trend and Change in Streamflow (Module II)

Yichang station is the control station for the entire UYRB. Streamflow series observed in the Yichang station reflects the general characteristics of streamflow changes in the UYRB. The observed annual streamflow series of Yichang station is shown in Figure 3 . It is clear that streamflow in the UYRB decreased by a rate of −0.77 km 3 per year during 1951–2013. The long-term trend in streamflow and its significance were identified by the Mann-Kendall test and the Spearman’s Rho test. Results obtained from the two methods are consistent, revealing a significant downward trend (at the 95% confidence level) in streamflow in the UYRB during the study period. A more detailed description and analysis of the spatio-temporal variability of streamflow can be found in our previous study ( Zhang et al., 2021a ).

FIGURE 3 . Annual streamflow series observed in Yichang station (Note: ↓ represents decreasing trend; 95% represents a confidence level of 95%).

The change point in streamflow series was identified by the Pettitt’s test and the sequential clustering method. Both methods show that the abrupt change point of observed annual streamflow series at Yichang station occurred in 1993at 95% confidence level. More detailed results and analysis of the abrupt change can also be found in our previous study ( Zhang et al., 2021a ). According to the change point in streamflow series, the study period can be divided into two sub-periods (i.e., 1951–1993 and 1994–2013).

For baseline period determination, the changes of streamflow, precipitation and ET 0 during 1951–1993 were detected by Mann-Kendall test. The results indicate that there are no obvious changes in streamflow, precipitation or ET 0 series during the first sub-period ( Table 1 ). Therefore, the basic assumption of “baseline period and change period” is valid. The sub-period 1951–1993 is confirmed as the baseline period and the sub-period 1994–2013 is determined as the change period.

TABLE 1 . Characteristics of the streamflow, precipitation and ET 0 series in the UYRB before and after change point (Note: – represents a confidence level lower than 80%).

Meanwhile, the mean value, standard deviation (SD) and coefficient of variation (CV) of the hydro-meteorology series in the baseline and change periods were separately calculated and listed in Table 1 . Compared with the baseline period, the average annual streamflow in the change period decreased by 8.4%. A statistically significant difference can be found between the mean values of the annual streamflow for the two periods at the 95% confidence level, while the difference between the SD or CV is not statistically significant. Differences at the 95% confidence level are also found between the means of the precipitation and ET 0 series for the two periods. No statistically significant differences are found between the SD and CV of the precipitation and ET 0 series for the two periods. At this point in the analysis, it can be concluded that the hydro-meteorology series of the UYRB from 1951 to 2013 meets the two basic assumptions of “baseline period and change period” and “sufficient length of time series”. In this case study, the SWAT-based method, climate-elasticity-based method and Choudhury-Yang-equation-based method are all applicable.

Driving Factors Identification (Module II)

Precipitation, evapotranspiration and streamflow are the basic elements of the water cycle. Precipitation is the water income item and evapotranspiration is the output item. Actual evapotranspiration is influenced by available water (i.e., precipitation, land surface conditions) on the one hand, and available energy (i.e., potential evapotranspiration) on the other. The potential evapotranspiration is mainly influenced by meteorological factors such as maximum temperature (Tmax), minimum temperature (Tmin), wind speed (WS), relative humidity (RH), sunshine hour (SH). In addition, streamflow is also influenced by indirect human activities such as land use/cover change (LUCC) and direct human activities. Therefore, P, Tmax, Tmin, RH, SH, WS, and LUCC data were initially selected to form the drivers of streamflow changes in Module Ⅰ. It is noted that the direct human activities in the UYRB are too complex to collect detailed information. In this case study, the contribution of direct human activities was estimated by subtracting the contribution of other drivers from the total streamflow change.

In order to identify the key driving factors of streamflow change in the drivers dataset in Module Ⅰ. The mean value of each meteorological factor in baseline period and change period were calculated and listed in Table 2 . The t -test was used to detect differences of meteorological factors for the two periods. Differences at 99% confidence level are found between the mean values of P, Tmax, Tmin, WS, and RH for the two periods. Difference is not obvious in SH. Therefore, P, Tmax, Tmin, WS, and RH were selected as the meteorological driving factors. During the study period, land use/cover in the UYRB underwent obvious changes. As an indirect human activity, LUCC was also selected as a driver. Detailed description of LUCC can be found in our previous study ( Zhang et al., 2021a ).

TABLE 2 . Mean values of the meteorological factors in the UYRB for baseline period and change period.

Climatic and Anthropogenic Contributions to Streamflow Change (Module III)

Based on the judgments in Module III, it is clear that the SWAT-based method is the best method for assessing climatic and anthropogenic influences on streamflow in the UYRB, because it is capable of assessing the effects of climate change, indirect human activities and direct human activities on streamflow. The climate-elasticity-based method is the second-best choice, which does not distinguish between direct and indirect human activities. In this case study, the parameter n was not collected and needed to be estimated, so the Choudhury-Yang-equation-based method is the worst option. In this section, the SWAT-based method was selected, supplemented by the other two methods, to conduct a comprehensive attribution analysis. For the purpose of comparison between the three methods, streamflow changes were attributed to climate change, direct human activities, and indirect human activities.

1) SWAT-based method

According to the general steps of the method, the contributions of climate change ( Δ W C ), LUCC ( Δ W L ) and direct human activities ( Δ W D ) can be calculated by the following equations:

where L a p and L b p are the land use inputs during change period and baseline period; Δ W T is the total streamflow change; W o b s e r v e d , a p ¯ is the average observed streamflow during change period.

The contributions of climate change, LUCC and direct human activities are listed in Table 3 . The results suggest that the main contributions to streamflow change are from climatic variabilities (69%), LUCC (10%) and direct human activities (21%). Climate change appears to be the main cause of streamflow change with a contribution of -25.5 km 3 .

2) Climate-elasticity-based method

TABLE 3 . Climatic and anthropogenic contributions to streamflow changes.

A multi-parameter climate elasticity model was developed with the key driving factors as follows:

where ε P , ε T max , ε T min , ε R H , ε W S are elasticity coefficients of the meteorological factors.

The contributions of climate change and human activities to streamflow were listed in Table 3 . The climatic contribution is −22.6 km 3 . The contribution of climate change and human activities to streamflow change is about 7:3.

3) Choudhury-Yang-equation-based method

The elasticity coefficients of precipitation, potential evapotranspiration and land surface parameter were first calculated according to the steps: ε P = 1.68 , ε E 0 = − 0.68 , ε n = − 0.69 . The larger the absolute value of the elasticity coefficient of the factor, the more sensitive the streamflow is to it. Thus, precipitation is the dominant factor of streamflow change. The contributions of precipitation, potential evapotranspiration and land surface parameter to streamflow were calculated and listed in Table 3 .

Through the analysis in Climatic and Anthropogenic Contributions to Streamflow Change (Module III) , the results based on the SWAT-based method and climate-elasticity-based method are relatively close, and the contribution of climate change and human activities to the streamflow changes are 7 : 3 and 6 : 4, respectively; where the contribution of climate change is −25.5 and −22.6 km 3 , respectively. However, the contribution of climate change based on the Choudhury-Yang-equation-based method is -45.6 km 3 , which is significantly larger than the above two methods. The possible reason for this is that n is calculated from precipitation and evapotranspiration.

Here, we used a multiple regression model, the most direct method for attribution analysis of streamflow changes under changing environment, to further verify the contribution of P and E 0 (climate factors) to streamflow changes. The specific expressions are as follows:

where R , P and E 0 are streamflow, precipitation, and potential evapotranspiration; a , b and c are regression coefficients.

Based on the precipitation, potential evapotranspiration and streamflow data for the baseline period (1951–1993), the values of a, b and c are 0.426, 0.128, and 128.961 (at the 95% confidence level), respectively, with a correlation coefficient of 0.803. The results of calculating the attribution analysis in the UYRB are shown in Table 3 . It can be seen that: the contribution of climate change is −24.4 km 3 , and the contribution ratio of climate change and human activities to streamflow changes is 6.6 : 3.4, which is close to the results of the SWAT-based method and climate-elasticity-based method. Therefore, comprehensive analysis shows that climate change is the dominant driver of streamflow changes in the UYRB, and the contribution ratio of climate change, indirect human activities, and direct human activities to streamflow changes is about 7 : 1 : 2.

In general, the proposed framework was applied to assess climatic and anthropogenic influences on streamflow changes in the UYRB in this study, indicating that the dominant factor of streamflow changes in the basin from 1951 to 2013 was climate change, which is consistent with previous studies ( Lu et al., 2019 ; Ye et al., 2020 ; Shao et al., 2021 ). For example, Wang and Xia (2015) employed water balance model to assess the impacts of climate change and human activities on streamflow changes in the UYRB, finding that 71.43% of streamflow decrease was due to climate change in the basin since 1993. Similarly, Ahmed et al. (2020) noted that climate change played a controlling role in streamflow changes in the UYRB. Recently, Shao et al. (2021) used the water balance equation, double mass curve, and linear regression analysis to quantify the contributions of climate change and anthropogenic activities to streamflow changes in the Jialing River (a main tributary of the upper Yangtze River) during 1960–2017. The results show that climate change had led to a significant reduction in annual streamflow (82.2%).

Different methods of attribution analysis have their own characteristics, limitations, critical criteria and assumptions, which in turn cause large variability and uncertainty in calculation results ( Zhang et al., 2020 ). This is consistent with the findings of this study that the attribution results based on the Choudhury-Yang-equation-based method are significantly different from both the SWAT-based method and climate-elasticity-based method. This further emphasizes the importance of selecting an appropriate methodology to help properly assess the contribution of climate change and human activities to streamflow changes. Therefore, the proposed framework is valuable and effective in helping researchers to select the appropriate method, reduce the uncertainty introduced by the method, and avoid the misperceptions that may result from inappropriate methods. Indeed, the framework has some limitations, which need further research. We chose only one commonly used method as the most representative for the three categories of methods (i.e., hydrological modeling, statistical analysis, and conceptual approaches). However, the calculation results of different methods in the same category may not be consistent. For instance, Zhang et al. (2020) used six Budyko framework-based methods to estimate the contributions of climate change and human activities to streamflow changes. The results showed that the uncertainty between these methods is about 5–7%.

This study aims to provide guidance for the selection of appropriate methods to quantify the climatic and anthropogenic influences on streamflow. Therefore, a method selection framework for attribution analysis was developed using the SWAT-based method, climate-elasticity-based method, and Choudhury-Yang-equation-based method jointly. The proposed framework consists of three modules, namely data preparation (Module Ⅰ), streamflow changes analysis and key driver identification (Module Ⅱ), and attribution analysis method selection and comprehensive assessment (Module Ⅲ). To evaluate its effectiveness, a case study in the UYRB was conducted. Under the framework, a significantly decreasing trend and changes were detected in observed annual streamflow in the UYRB. The study period was divided into a baseline period (1951–1993) and a change period (1994–2013) based on proposed criteria and assumptions. Subsequently, statistical indicators indicated that precipitation, maximum temperature, minimum temperature, wind speed, relative humidity, and LUCC are key drivers of streamflow changes in the UYRB. The analysis in Module Ⅲ indicates that the SWAT-based method is the best approach to assess climatic and anthropogenic influences on streamflow in the UYRB, and the climate-elasticity-based method and Choudhury-Yang-equation-based method are also applicable. The comprehensive attribution analysis suggests that climate change is the dominant cause of streamflow changes in the UYRB, and the contribution of climate change, indirect human activities, and direct human activities to streamflow changes is about 7:1:2. Overall, the proposed framework is efficient and valuable in assisting researchers to find appropriate methods for attribution analysis of streamflow changes under changing environments.

Data Availability Statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

Author Contributions

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

This work was supported by the National Key R&D Program of China (Grant No. 2018YFE0206200); the Project funded by China Postdoctoral Science Foundation (Grant No. 2019M661884); the Natural Science Foundation of Jiangsu Province, China (Grant No. BK20200160); the Special Research Fund of Nanjing Hydraulic Research Institute (Grant No. Y120006).

Conflict of Interest

CX was employed by the China Energy Investment Corporation Science and Technology Research Institute Co., Ltd

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer CS declared a shared affiliation, with no collaboration, with one of the authors, LY, to the handling editor at the time of the review.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Arnold, J. G., Srinivasan, R., Muttiah, R. S., and Williams, J. R. (1998). Large Area Hydrologic Modeling and Assessment Part I: Model Development. J. Am. Water Resour. Assoc 34, 73–89. doi:10.1111/j.1752-1688.1998.tb05961.x

CrossRef Full Text | Google Scholar

Borgwardt, F., Unfer, G., Auer, S., Waldner, K., El-Matbouli, M., and Bechter, T. (2020). Direct and Indirect Climate Change Impacts on Brown Trout in Central Europe: How Thermal Regimes Reinforce Physiological Stress and Support the Emergence of Diseases. Front. Environ. Sci. 8. doi:10.3389/fenvs.2020.00059

Budyko, M. I. (1974). Climate and Life . New York: Academic Press , 508.

Google Scholar

Dey, P., and Mishra, A. (2017). Separating the Impacts of Climate Change and Human Activities on Streamflow: A Review of Methodologies and Critical Assumptions. J. Hydrol. 548, 278–290. doi:10.1016/j.jhydrol.2017.03.014

Dubes, R., and Jain, A. K. (1980). Clustering Methodologies in Exploratory Data Analysis. Adv. Comput. 19, 113–228. doi:10.1016/S0065-2458(08)60034-0

Fu, G., Charles, S. P., and Chiew, F. H. S. (2007). A Two-Parameter Climate Elasticity of Streamflow index to Assess Climate Change Effects on Annual Streamflow. Water Resour. Res. 43 (11), W11419. doi:10.1029/2007WR005890

Hajihosseini, M., Hajihosseini, H., Morid, S., Delavar, M., and Booij, M. J. (2020). Impacts of Land Use Changes and Climate Variability on Transboundary Hirmand River Using SWAT. J. Water Clim. Change 11 (4), 1695–1711. doi:10.2166/wcc.2019.100

Jardim, P. F., Melo, M. M. M., Ribeiro, L. d. C., Collischonn, W., and Paz, A. R. d. (2020). A Modeling Assessment of Large-Scale Hydrologic Alteration in South American Pantanal Due to Upstream Dam Operation. Front. Environ. Sci. 8, 567450. doi:10.3389/fenvs.2020.567450

Jiang, S., Zhou, L., Ren, L., Wang, M., Xu, C.-Y., Yuan, F., et al. (2021). Development of a Comprehensive Framework for Quantifying the Impacts of Climate Change and Human Activities on River Hydrological Health Variation. J. Hydrol. 600, 126566. doi:10.1016/j.jhydrol.2021.126566

Kendall, M. G. (1975). Rank Correlation Methods . London: Griffin .

Lehmann, E. L. (1975). Nonparametrics, Statistical Methods Based on Ranks . California: Holden Day .

Li, B., Su, H., Chen, F., Li, H., Zhang, R., Tian, J., et al. (2014). Separation of the Impact of Climate Change and Human Activity on Streamflow in the Upper and Middle Reaches of the Taoer River, Northeastern China. Theor. Appl. Climatol 118 (1-2), 271–283. doi:10.1007/s00704-013-1032-8

Li, H., Shi, C., Zhang, Y., Ning, T., Sun, P., Liu, X., et al. (2020). Using the Budyko Hypothesis for Detecting and Attributing Changes in Runoff to Climate and Vegetation Change in the Soft sandstone Area of the Middle Yellow River basin, China. Sci. Total Environ. 703, 135588. doi:10.1016/j.scitotenv.2019.135588

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, J., You, Y., Zhang, Q., and Gu, X. (2021). Attribution of Streamflow Changes across the globe Based on the Budyko Framework. Sci. Total Environ. 794, 148662. doi:10.1016/j.scitotenv.2021.148662

Liu, X., Liu, W., Yang, H., Tang, Q., Flörke, M., Masaki, Y., et al. (2019). Multimodel Assessments of Human and Climate Impacts on Mean Annual Streamflow in China. Hydrol. Earth Syst. Sci. 23 (3), 1245–1261. doi:10.5194/hess-23-1245-2019

Lu, C.-h., Dong, X.-y., Tang, J.-l., and Liu, G.-c. (2019). Spatio-temporal Trends and Causes of Variations in Runoff and Sediment Load of the Jinsha River in China. J. Mt. Sci. 16 (10), 2361–2378. doi:10.1007/s11629-018-5330-6

Mann, H. B. (1945). Nonparametric Tests against Trend. Econometrica 13, 245–259. doi:10.2307/1907187

Miao, C., Yang, L., Liu, B., Gao, Y., and Li, S. (2011). Streamflow Changes and its Influencing Factors in the Mainstream of the Songhua River basin, Northeast China over the Past 50 Years. Environ. Earth Sci. 63 (3), 489–499. doi:10.1007/s12665-010-0717-x

Penny, G., Srinivasan, V., Apoorva, R., Jeremiah, K., Peschel, J., Young, S., et al. (2020). A Process‐based Approach to Attribution of Historical Streamflow Decline in a Data‐scarce and Human‐dominated Watershed. Hydrological Process. 34 (8), 1981–1995. doi:10.1002/hyp.13707

Pettitt, A. N. (1979). A Non-parametric Approach to the Change-point Problem. Appl. Stat. 28 (2), 126–135. doi:10.2307/2346729

Porter, M. E., Hill, M. C., Harris, T., Brookfield, A., and Li, X. (2021). The DiscoverFramework Freeware Toolkit for Multivariate Spatio-Temporal Environmental Data Visualization and Evaluation. Environ. Model. Softw. 143, 105104. doi:10.1016/j.envsoft.2021.105104

Renner, M., Seppelt, R., and Bernhofer, C. (2012). Evaluation of Water-Energy Balance Frameworks to Predict the Sensitivity of Streamflow to Climate Change. Hydrol. Earth Syst. Sci. 16 (5), 1419–1433. doi:10.5194/hess-16-1419-2012

Schaake, J. C. (1990). “From Climate to Flow,” in Climate Change and U.S.Water Resources . Editor P. E. Waggoner (New York: JohnWiley ), 177–206.

Shao, Y., He, Y., Mu, X., Zhao, G., Gao, P., and Sun, W. (2021). Contributions of Climate Change and Human Activities to Runoff and Sediment Discharge Reductions in the Jialing River, a Main Tributary of the Upper Yangtze River, China. Theor. Appl. Climatol 145 (3-4), 1437–1450. doi:10.1007/s00704-021-03682-1

Todhunter, P. E., Jackson, C. C., and Mahmood, T. H. (2020). Streamflow Partitioning Using the Budyko Framework in a Northern Glaciated Watershed under Drought to Deluge Conditions. J. Hydrol. 591, 125569. doi:10.1016/j.jhydrol.2020.125569

Tomer, M. D., and Schilling, K. E. (2009). A Simple Approach to Distinguish Land-Use and Climate-Change Effects on Watershed Hydrology. J. Hydrol. 376 (1-2), 24–33. doi:10.1016/j.jhydrol.2009.07.029

Wang, M., Zhang, Y., Lu, Y., Gong, X., and Gao, L. (2021). Detection and Attribution of Reference Evapotranspiration Change (1951-2020) in the Upper Yangtze River Basin of China. J. Water Clim. Change , jwc2021011. doi:10.2166/wcc.2021.011

Wang, W., Zhang, Y., and Tang, Q. (2020). Impact Assessment of Climate Change and Human Activities on Streamflow Signatures in the Yellow River Basin Using the Budyko Hypothesis and Derived Differential Equation. J. Hydrol. 591, 125460. doi:10.1016/j.jhydrol.2020.125460

Wang, X. (2014). Advances in Separating Effects of Climate Variability and Human Activity on Stream Discharge: An Overview. Adv. Water Resour. 71, 209–218. doi:10.1016/j.advwatres.2014.06.007

Wright, M. S. P., Santelmann, M. V., Vaché, K. B., and Hulse, D. W. (2021). Modeling the Impact of Development Policies and Climate on Suburban Watershed Hydrology Near Portland, Oregon. Landscape Urban Plann. 214, 104133. doi:10.1016/j.lurbplan.2021.104133

Yang, H., and Yang, D. (2011). Derivation of Climate Elasticity of Runoff to Assess the Effects of Climate Change on Annual Runoff. Water Resour. Res. 47, W07526. doi:10.1029/2010WR009287

Yang, H., Yang, D., Lei, Z., and Sun, F. (2008). New Analytical Derivation of the Mean Annual Water-Energy Balance Equation. Water Resour. Res. 44, W034103. doi:10.1029/2007WR006135

Yasarer, L. M. W., Taylor, J. M., Rigby, J. R., and Locke, M. A. (2020). Trends in Land Use, Irrigation, and Streamflow Alteration in the Mississippi River Alluvial Plain. Front. Environ. Sci. 8, 66. doi:10.3389/fenvs.2020.00066

Ye, X., Zhang, Z., Xu, C.-Y., and Liu, J. (2020). Attribution Analysis on Regional Differentiation of Water Resources Variation in the Yangtze River Basin under the Context of Global Warming. Water 12 (6), 1809. doi:10.3390/w12061809

Zeng, X., Schnier, S., and Cai, X. (2021). A Data-Driven Analysis of Frequent Patterns and Variable Importance for Streamflow Trend Attribution. Adv. Water Resour. 147, 103799. doi:10.1016/j.advwatres.2020.103799

Zhang, K., Ruben, G. B., Li, X., Li, Z., Yu, Z., Xia, J., et al. (2020). A Comprehensive Assessment Framework for Quantifying Climatic and Anthropogenic Contributions to Streamflow Changes: A Case Study in a Typical Semi-arid North China basin. Environ. Model. Softw. 128, 104704. doi:10.1016/j.envsoft.2020.104704

Zhang, Y., Wang, M., Chen, J., Zhong, P.-a., Wu, X., and Wu, S. (2021a). Multiscale Attribution Analysis for Assessing Effects of Changing Environment on Runoff: Case Study of the Upstream Yangtze River in China. J. Water Clim. Change 12 (2), 627–646. doi:10.2166/wcc.2020.155

Zhang, Y., Yu, L., Wu, S., Wu, X., Dai, J., Xue, W., et al. (2021b). A Framework for Adaptive Control of Multi-Reservoir Systems under Changing Environment. J. Clean. Prod. 316, 128304. doi:10.1016/j.jclepro.2021.128304

Zhang, Y., Zhong, P.-a., Wang, M., Xu, B., and Chen, J. (2018). Changes Identification of the Three Gorges Reservoir Inflow and the Driving Factors Quantification. Quat. Int. 475, 28–41. doi:10.1016/j.quaint.2016.02.064

Zhao, Q., Liu, S., Deng, L., Dong, S., Yang, J., and Wang, C. (2012). The Effects of Dam Construction and Precipitation Variability on Hydrologic Alteration in the Lancang River Basin of Southwest China. Stoch Environ. Res. Risk Assess. 26, 993–1011. doi:10.1007/s00477-012-0583-z

Keywords: attribution analysis, climate change, human activities, streamflow changes, Upper Yangtze River Basin

Citation: Zhang Y, Wu X, Wu S, Dai J, Yu L, Xue W, Wang F, Gao A and Xue C (2021) A Framework for Methodological Options to Assess Climatic and Anthropogenic Influences on Streamflow. Front. Environ. Sci. 9:765227. doi: 10.3389/fenvs.2021.765227

Received: 26 August 2021; Accepted: 15 September 2021; Published: 29 October 2021.

Reviewed by:

Copyright © 2021 Zhang, Wu, Wu, Dai, Yu, Xue, Wang, Gao and Xue. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Xiufeng Wu, [email protected] ; Shiqiang Wu, [email protected] ; Jiangyu Dai, [email protected] ; Lei Yu, [email protected]

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

• Ways to Give
• Contact an Expert
• Explore WRI Perspectives

Filter Your Site Experience by Topic

Applying the filters below will filter all articles, data, insights and projects by the topic area you select.

• All Topics Remove filter
• Climate filter site by Climate
• Cities filter site by Cities
• Energy filter site by Energy
• Food filter site by Food
• Forests filter site by Forests
• Freshwater filter site by Freshwater
• Ocean filter site by Ocean
• Business filter site by Business
• Economics filter site by Economics
• Finance filter site by Finance
• Equity & Governance filter site by Equity & Governance

Search WRI.org

Not sure where to find something? Search all of the site's content.

Methodology Underpinning the State of Climate Action Series: 2023 Update

This technical note describes the State of Climate Action 2023 ’s methodology for identifying sectors that must transform, translating these transformations into global mitigation targets primarily for 2030 and 2050 and selecting indicators with datasets to monitor annual change. It also outlines the report’s approach for assessing progress made toward near-term targets and comparing trends over time.

Published under  Systems Change Lab , this Technical Note features analysis from Climate , Energy , Food , Forests , WRI Ross Center for Sustainable Cities and Finance . Reach out to Joel Jaeger for more information.

• Joel Jaeger

State of Climate Action

State of climate action 2023.

• Methodology Underpinning the State of Climate Action Series
• State of Climate Action 2022
• State of Climate Action 2021: Systems Transformations Required to Limit Global Warming to 1.5°C
• State of Climate Action: Assessing Progress toward 2030 and 2050

Limiting global temperature rise to 1.5°C requires transformational change across power, buildings, industry, transport, forests and land and food and agriculture as well as the immediate scale-up of carbon removal technologies and climate finance. The State of Climate Action series provides an overview of the world’s collective efforts to accelerate these far-reaching transitions. We first translate each sectoral transformation into a set of actionable, 1.5°C-aligned targets for 2030 and 2050, with associated indicators and datasets. Annual installments of the report then compare recent progress made toward (or away from) these mitigation goals with the pace of change required to achieve 2030 targets to quantify the global gap in climate action. While a similar effort is warranted to evaluate adaptation efforts, we limit this series’ scope to tracking progress made in reducing greenhouse gas emissions and removing carbon dioxide from the atmosphere.

This technical note accompanies the State of Climate Action 2023 . It describes our methods for identifying sectors that must transform, translating these transformations into global mitigation targets primarily for 2030 and 2050 and selecting indicators with datasets to monitor annual change. It also outlines our approach for assessing the world’s progress made toward near-term targets and categorizing recent efforts as on track, off track, well off track, heading in the wrong direction or insufficient data. Finally, it details how we compare trends over time, as well as limitations to our methodology.

This year’s technical note features several changes to the technical note we published alongside the State of Climate Action 2022 . Key updates include the following: revisions to 15 targets to reflect the best available science, including scenarios from the integrated assessment models included in the Intergovernmental Panel on Climate Change’s Sixth Assessment Report and recently published literature; the inclusion of interim targets for 2035 or 2040 where possible; the addition of four indicators (share of new buildings that are zero-carbon in operation, share of electric vehicles in two- and three-wheeler sales, the GHG emissions intensity of agricultural production and ratio of investment in low-carbon to fossil fuel energy supply); and the removal of two indicators (carbon intensity of land-based passenger transport and GHG emissions from agricultural production). We also refined our methodology for assessing the progress of indicators that we can reasonably expect to follow an S-curve, and added new methods for comparing each indicator’s most recent data point to recent trends. Finally, we removed the discussion of enabling conditions from this technical note, as the series no longer covers them.

Preview image by Kamal Preet Kaur/Unsplash

Connected to this report

Tracking climate action: how the world can still limit warming to 1.5 degrees c, we’re not on track for 1.5 degrees c. what will it take, climate action must progress far faster to achieve 1.5 c goal.

Systems Change Lab

Monitoring, learning from and accelerating the transformational changes required to protect both people and the planet

Primary Contacts

Senior Research Associate, Systems Change Lab

How You Can Help

WRI relies on the generosity of donors like you to turn research into action. You can support our work by making a gift today or exploring other ways to give.

Stay Informed

World Resources Institute 10 G Street NE Suite 800 Washington DC 20002 +1 (202) 729-7600

© 2024 World Resources Institute

Envision a world where everyone can enjoy clean air, walkable cities, vibrant landscapes, nutritious food and affordable energy.

An official website of the United States government

Here’s how you know

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( Lock A locked padlock ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

JavaScript appears to be disabled on this computer. Please click here to see any active alerts .

Strategies for Climate Change Adaptation

Select a particular area of interest or explore a list of all adaptation strategies. Learn more about adaptation strategies .

Public Health

All strategies, about adaptation strategies.

Climate changes can make it more difficult for communities to maintain air quality that protect human health and the environment. The Adaptation Strategies below offer possible ways to address anticipated climate risks to outdoor and indoor air quality.

• Outdoor Air

Climate change can make it more difficult for communities to provide drinking water and wastewater services, protect water quality, and maintain healthy aquatic environments. The Adaptation Strategies below offer possible ways to address anticipated climate risks to water management.

• Water Utility
• Water Quality
• Ecosystem Protection

Climate change can make it more difficult to properly manage hazardous and non-hazardous wastes. The pages below offer possible ways to address anticipated climate risks to contaminated site management and disaster debris management.

• Waste Facility Protection adaptation strategies
• Develop a Waste Management Plan

Climate change can make it more difficult for communities to maintain public health. The pages below offer possible ways to address public health risks to air quality, water quality, extreme heat, and waste issues related from anticipated climate changes.

• Public Health Strategies

The adaptation strategies provided on this site are intended to inform and assist communities in identifying potential alternatives. They are illustrative and are presented to help communities consider possible ways to address anticipated current and future threats resulting from the changing climate. In particular, it is important to note:

• The strategies presented are NOT a comprehensive or exhaustive list of resiliency or adaptation actions that may be relevant.
• None of the provided alternatives are likely to be appropriate in all circumstances; the appropriateness of each alternative should be considered in the local context for which it is being considered.
• The potential strategies are largely drawn from EPA and other federal resources. Before adopting any particular strategy, it should be considered in the context provided by the primary source document from which it originated. Source document(s) are indicated.
• The presented strategies should not be relied on exclusively in conducting risk assessments, developing response plans, or making adaptation decisions.
• This information is not a substitute for the professional advice of an environmental or climate change professional or attorney.
• Climate Change Adaptation Resource Center (ARC-X) Home
• Your Climate Adaptation Search
• Implications of Climate Change
• Adaptation Planning
• Adaptation Strategies
• Case Studies
• Federal Funding & Technical Assistance
• Underlying Science
• EPA Contacts & State Websites

share this!

August 29, 2024

This article has been reviewed according to Science X's editorial process and policies . Editors have highlighted the following attributes while ensuring the content's credibility:

fact-checked

trusted source

written by researcher(s)

How 'climate mainstreaming' can address climate change and further development goals

by Steven Lam and Gloria Novović, The Conversation

Canada's first National Adaptation Strategy urges Canadians to consider climate change impacts in their everyday decisions.

The strategy calls such an approach "climate mainstreaming." The approach states that: "as climate impacts become more severe and frequent, and the costs mount, incorporating adaptation considerations in health, social, environmental, infrastructure and economic decision-making is critical to ensure that our collective efforts keep pace."

Similar statements are outlined in the 2023 press release of the Intergovernmental Panel on Climate Change (IPCC). The IPCC Chair Hoesung Lee stated, "mainstreaming effective and equitable climate action will not only reduce losses and damages for nature and people, it will also provide wider benefits."

Global greenhouse gas emissions need to be cut 43% from 2019 to 2030 to limit global warming to 1.5 C. At the 2023 United Nations climate conference (COP28) in Dubai, parties were deemed off track in meeting their Paris Agreement goals.

A rapid and meaningful expansion of climate mainstreaming—the integration of climate considerations into all development programs and policies—is vital for addressing the urgent global climate crisis.

Why mainstream climate change?

Mainstreaming climate considerations ensures that responses to climate change are systemically embedded in all policies and actions, rather than treated as a separate issue. This integration allows for more comprehensive and cost-effective interventions by addressing multiple issues at once.

For instance, within an ongoing program focused on improving food safety in informal, outdoor markets through enhanced hygienic practices, mainstreaming might entail additional activities related to climate adaptation, such as raising awareness among food vendors about the importance of refrigeration during heat waves to prevent bacterial growth .

Failing to mainstream climate considerations can hinder climate action as well as result in maladaptation, which occurs when well-intentioned development actions inadvertently increase climate impacts. For example, seawalls can protect people and property from damage in the short term. However, if they are not part of a long-term plan that can adapt to changing conditions, they can trap communities in risky situations and increase their exposure to climate risks over time .

While attention to climate mainstreaming calls for the prioritization of climate considerations across all policy arenas, progress remains slow and uneven due primarily to an institutional resistance to change. Climate action is often seen as the responsibility of a single sector rather than the collective, and incremental changes are inferior to transformative ones.

Furthermore, climate mainstreaming is often narrowly interpreted as simply the addition of climate to existing structures and initiatives. Often derisively dubbed a "just add climate and stir" approach.

To help address these preconceptions, our research has explored how climate mainstreaming challenges resemble similar decades-long struggles to mainstream gender equality across international and national public policy agendas. The question we have asked is: what can climate mainstreaming learn from gender mainstreaming?

Insights from gender mainstreaming

The longer history of gender mainstreaming, including institutional investments dating back to the 1990s, offers lessons about policy and institutional bottlenecks of mainstreaming. These lessons can help tackle political and institutional challenges of climate mainstreaming. The UN system, with clear gender and climate mainstreaming targets, offers a suitable arena for analysis.

In a new study published in 2024 , we reviewed documents of United Nations agencies working in the food and agriculture sector, which is strongly impacted by climate change. We found varying degrees of gender and climate mainstreaming across selected UN agencies.

Key areas where climate mainstreaming fell short compared to gender mainstreaming included: strategic planning, leadership, organizational culture and accountability.

Our review showed ways to improve climate mainstreaming. Here are three actions governments, development partners and industries can take now:

• Use multiple strategies: draw upon gender mainstreaming good practices to adopt both broad climate initiatives and specific interventions.
• Build institutional accountability: establish strong mechanisms to track progress in climate mainstreaming. The UN's framework for gender mainstreaming can act as a useful model. This would help ensure transparency, monitoring and a stronger commitment to climate action.
• Adopt a climate justice perspective: uphold the needs of climate change-vulnerable populations and prioritize collective human and environmental rights over economic growth. Ensure diverse stakeholders participate across all levels of decision-making.

Accountable and integrated climate justice interventions are prerequisites for a more sustainable and resilient future. Financing is another.

Financing is key

While mainstreaming is important, it is nothing without adequate financing. The 2015 Paris Agreement requires high-income countries to contribute $100 billion annually . However, this goal has not been met , and the existing funds are unevenly distributed.

Historically disadvantaged countries are the least responsible for yet the most impacted by climate crisis . These countries are largely left to balance development and climate action investments in a generally unjust international financial system.

In 2022, official development assistance reached US$204 billion , but this still left nearly half of the humanitarian requirements unmet. Rich countries spent only 0.36% of their total income on aid—slightly up from 0.33% in 2021, but still much lower than the 0.7% promised back in 1970.

With the financing to back it up, a climate mainstreaming perspective may just be the solution to addressing both global development and climate goals.

Provided by The Conversation

Explore further

Feedback to editors

Chemists achieve PFAS-free synthesis of fluorinated pharmaceutical and agrochemical compounds

55 minutes ago

Number of fish species at risk of extinction five-fold higher than previous estimates according to new prediction

Tuberculosis under the sea: A marine sponge microbe provides insights into the bacterium's evolution

Ancient gene gives spiders their narrow waist, research reveals

Catalytic process vaporizes plastic bags and bottles, yielding gases to make new, recycled plastics

Bacteria able to overcome cost of vancomycin resistance in lab setting

Uncovering the secret communication of marmoset monkeys: They have names

How a salt giant radically reshaped Mediterranean marine biodiversity

Neutron scattering instrument represents a new dawn for AI-powered atomic-scale 3D imaging

2 hours ago

New moves for self defense—how plants can inspire future soft robotic design

Relevant physicsforums posts, alaska - pedersen glacier: landslide triggered tsunami.

Aug 23, 2024

Iceland warming up again - quakes swarming

The secrets of prof. verschure's rosetta stones, shiveluch volcano erupts on kamchatka peninsula.

Aug 18, 2024

A very puzzling rock or a pallasite / mesmosiderite or a nothing burger

Aug 14, 2024

M6.8 and M6.3 east of Mindanao, Philippines

Aug 13, 2024

More from Earth Sciences

Related Stories

Researchers say nature recovery must be integrated across all sectors to bend the curve of biodiversity loss

May 2, 2024

Experts highlight benefits of mainstreaming nature in U.S. policymaking

Aug 1, 2024

Key steps for global health, climate post-COP28

Jun 26, 2024

Integrating health and climate agendas

Jun 24, 2020

Dutch people adapt their climate beliefs more easily than their behavior, find researchers

Mar 5, 2024

Four in five people want more climate action: UN survey

Jun 20, 2024

Recommended for you

Labor day crowds temporarily impact local streams, research shows

7 hours ago

Land-sea 'tag-team' devastated ocean life millions of years ago, reveal scientists

10 hours ago

Global timber supply threatened as climate change pushes cropland northwards

Study reveals crucial role of mixing Atlantic and Arctic waters in global ocean circulation

5 hours ago

NASA study tallies carbon emissions from massive Canadian fires

Aug 28, 2024

Let us know if there is a problem with our content

Use this form if you have come across a typo, inaccuracy or would like to send an edit request for the content on this page. For general inquiries, please use our contact form . For general feedback, use the public comments section below (please adhere to guidelines ).

Please select the most appropriate category to facilitate processing of your request

Thank you for taking time to provide your feedback to the editors.

Your feedback is important to us. However, we do not guarantee individual replies due to the high volume of messages.

E-mail the story

Your email address is used only to let the recipient know who sent the email. Neither your address nor the recipient's address will be used for any other purpose. The information you enter will appear in your e-mail message and is not retained by Phys.org in any form.

Newsletter sign up

Get weekly and/or daily updates delivered to your inbox. You can unsubscribe at any time and we'll never share your details to third parties.

More information Privacy policy

Donate and enjoy an ad-free experience

We keep our content available to everyone. Consider supporting Science X's mission by getting a premium account.

E-mail newsletter

Effectiveness of 1,500 global climate policies ranked for first time

The world can take a major step to meeting the goals of the Paris Climate Accord by focusing on 63 cases where climate policies have had the most impact, new research has revealed. The findings have been published today in Science .

Our results inform contentious policy debates in three main ways. First, we show evidence for the effectiveness of policy mixes. Second our findings highlight that successful policy mixes vary across sectors and that policy-makers should focus on sector-specific best practices. Third our results stress that effective policies vary with economic development. Study co-author Dr Moritz Schwarz , an Associate at the Climate Econometrics Programme at the University of Oxford

The study, led by Climate Econometricians at the University of Oxford, the Potsdam Institute for Climate Impact Research (PIK), and the Mercator Research Institute on Global Commons and Climate Change (MCC), analysed 1,500 observed policies documented in a novel, high quality, OECD climate policy database for effectiveness. It is the first time a global dataset of policies has been compared and ranked in this way.

Using a methodology developed by Climate Econometrics at The Institute for New Economic Thinking at the Oxford Martin School (INET Oxford), the researchers measured ‘emission breaks’ that followed policy interventions. The break detection methodology, called indicator saturation estimation, developed at Climate Econometrics, allows break indicators for all possible dates to be examined objectively using a variant of machine learning.

The results were sobering: Across four sectors, 41 countries, two decades and 1,500 policies, only 63 successful policy interventions with large effects were identified, which reduced total emissions between 0.6 and 1.8 Gt CO2.

However, the authors say the good news is that policymakers can learn from the 63 effective cases where climate policies had led to meaningful reductions to get back on track.

The researchers have made the data available to policy-makers across the world, and have produced a sector by sector, country by country data visualisation in a dashboard .

Overall, the Team concluded:

• Climate policies are more effective as part of a mix:  In most cases, effect sizes of climate policies are larger if a policy instrument is part of a policy mix rather than implemented alone –for example combining carbon pricing with a subsidy.
• Developed and developing countries have different climate policy needs:  In developed countries, carbon pricing stands out as an effective policy, whereas in developing countries, regulation is the most powerful policy.
• The Paris emissions gap can be closed:  Focusing on the 63 cases of effective climate policies would close the current emissions gap to meet the Paris Targets by 26% -41%, a significant contribution.

Scaling up good practice policies identified in this study to other sectors and other parts of the world can in the short term be a powerful climate mitigation strategy…The dashboard that we make available to policy-makers provides an accessible platform to conduct country-by-country, sector-by-sector comparisons and to find a suitable policy mix for different situations. Study co-author Professor Felix Pretis , Co-Director of the Climate Econometrics Programme at Nuffield College, University of Oxford

Study co-author Ebba Mark , researcher at the Calleva Project at INET Oxford, said the world needed to get back on track to meeting the Paris Climate Accord targets. ‘Meeting the Paris Climate objectives necessitates decisive policy action as we are still falling short -  data from the UN estimates that there remains a median emissions gap of 23 billion tonnes of CO2 equivalent by 2030 . It is now clear that the persistence of this emissions gap is not only attributable to an ambition gap but also to a gap in the real versus expected outcomes of implemented policies. The 63 success stories identified in this study provide key information about how we can bridge the emissions gap more meaningfully going forward.'

What works: Examples from the UK and USA

The country by country analysis showed that the UK has made very successful progress in the electricity sector, with two adjacent breaks detected following the mid-2013 introduction of a carbon price floor that imposed a minimum price for UK power producers. However, the study did not find in other UK sectors any major emission reductions following a policy intervention beyond what would be expected based on long-term economic and population dynamics.

The US has managed to reduce carbon emissions in the transport sector following actions taken in the aftermath of the financial crisis. While successful policy implementation in the transport sector is generally difficult and hence can be viewed as a positive example for the climate policy globally, the lack of any further climate policy successes in other sectors points to huge remaining challenges in the power sector or industry.

Dr Anupama Sen ,  Head of Policy Engagement at the Oxford Smith School of Enterprise and the Environment said: ' In more than 80%  of investments the total lifetime cost of a clean technology is considerably lower than that of a fossil technology. While the new UK government’s policies are moving in the right direction, they need to go further and faster to unlock these lower costs. New Oxford research now provides evidence that an optimal mix of policies can achieve this, and rapidly lower a country’s emissions.'

Further analysis can be found in INET Oxford’s accompanying Insight brief .

The study ‘Climate policies that achieved major emission reductions: Global evidence from two decades’ has been published in Science .

DISCOVER MORE

• Support Oxford's research
• Partner with Oxford on research
• Study at Oxford
• Research jobs at Oxford

You can view all news or browse by category

American Climate Policy Opinions

Aug. 27, 2024

Jon A. Krosnick and Bo MacInnis

Publication

Reading time

Introduction

In Climate Insights 2024: American Understanding of Climate Change, we showed that huge majorities of Americans believe that the earth has been warming, that the warming has been caused by human activity, that warming poses a significant threat to the nation and the world—especially to future generations—and that governments, businesses, and individuals should be taking steps to address it.

In this report, we turn to specific federal government opportunities to reduce future greenhouse gas emissions, often referred to as climate change mitigation. Policies to accomplish this goal fall into several categories, including:

• Consumer incentives that reward people for taking steps that reduce their use of fossil fuels and, by extension, reduce their carbon footprint
• Carbon pricing policies that require emitters to pay for their carbon emissions, such as a carbon tax (which would require carbon emitters to pay a tax for each ton of carbon they emit), or a cap-and-trade program (which would require businesses to have a permit for each ton of carbon they emit)
• Regulations that require manufacturers to increase energy efficiency of their products
• Tax incentives that encourage manufacturers to increase the energy efficiency of their products

This 2024 survey asked Americans about their opinions on a wide array of such policies, which allows us not only to assess current opinions, but to track changes in those opinions over the past two decades through comparisons with responses to comparable questions asked in earlier national surveys.

Explore the Data

Click here to explore the report's findings using our interactive data tool.

Overall emissions reduction strategies

In 2024, we asked for the first time whether Americans prefer using “carrots” to reduce emissions or “sticks.” The former entails offering incentives to reward companies for achieving desired outcomes, and the latter involves penalizing companies that fail to reach desired goals. 59 percent of Americans prefer a carrot approach in which government lowers taxes for companies that reduce emissions, and 35 percent prefer a stick approach such that government raises taxes on companies that do not reduce emissions (see Figure 1).

Overall emissions reduction principles

Over the past decades, a consistently large majority of Americans has wanted the government to reduce greenhouse gas emissions by US businesses. In 2024, 74 percent of Americans endorse this mitigation policy principle (see Figure 2). This number is not significantly different from the 77 percent seen in 2020 and is about the same as it has been since 1997 when this series of surveys was launched.

Most Popular Policies (>60 percent approval)

Taxing imported emissions.

In 2024, we asked about import taxes tied to emissions; respondents were asked whether they would favor taxing foreign companies for importing products that put out more greenhouse gases than a comparable US product. A huge majority of Americans, 84 percent , favor the special tax (see Figure 3).

Assisting with job transitions

In 2024, we asked whether the federal government should spend money to help people who lose jobs due to a transition from fossil-based electricity generation to electricity generated from renewable sources. 78 percent of Americans favor the government paying those people to learn to do other kinds of work.

Filling abandoned oil wells

In 2024, we asked whether the federal government should spend money to close off abandoned oil wells that emit greenhouse gases; 76 percent of Americans favor the government spending money to fill these old wells.

Shifting energy generation to renewable power

Huge numbers of Americans favor government effort to shift electricity generation away from fossil fuels and toward renewable energy sources.

In 2024, 72 percent of Americans believe that the US government should offer tax breaks to utilities in exchange for making more electricity from water, wind, and solar sources. However, this is a statistically significant decline from the 85 percent seen in 2020, and a record low since 2006 (see Figure 6).

13 percentage points fewer Americans believe that the US government should offer tax breaks to utilities in exchange for making more electricity from renewable sources in 2024 than 2020.

Slight changes to the question wording yielded similar results: 76 percent of Americans in 2024 favor either mandates or tax breaks for utilities to reduce greenhouse gas emissions from power plants. This number is not significantly different from the 82 percent seen in 2020 and is about the same as it has been since 2009. Before 2009, this proportion was slightly higher: 86 percent and 88 percent in 2006 and 2007, respectively (see Figure 7).

Increasing the energy efficiency of products

About two-thirds of Americans favor government efforts through tax breaks or mandates to improve the energy efficiency of various consumer products (see Figure 8).

Specifically, 62 percent of Americans in 2024 favor increasing the fuel efficiency of automobiles, a statistically significant drop from the 72 percent seen in 2020.

68 percent favor increasing the energy efficiency of appliances, similar to the 71 percent observed in 2020.

69 percent favor increasing the energy efficiency of new buildings, a statistically significant decline from the 76 percent in 2020.

Sequestering carbon

In 2024, 63 percent of Americans favor reducing emissions by sequestering (i.e., capturing and storing) carbon released by burning coal. This level of support has been steady over the past 15 years (see Figure 9).

84% of respondents favor import taxes tied to emissions, making it the most popular policy we surveyed.

Moderately Popular Policies (50–60 percent approval)

Reducing subsidies for fossil fuels.

In 2024, we asked for the first time whether the federal government should continue its long-standing practice of offering subsidies to oil and natural gas companies by reducing their taxes.

61 percent of Americans favor ending government reduction of oil companies’ taxes, and 37 percent believe these subsidies should continue.

42 percent of Americans favor ending government reduction of natural gas companies’ taxes, and 56 percent believe that these subsidies should continue.

Taxing greenhouse gases

When asked whether companies should be charged a tax for every ton of greenhouse gases they emit, 54 percent of respondents were in favor in 2024, a statistically significant decline from the 66 percent observed in 2020 (figure 10).

Creating a cap-and-trade program

Although economists generally assert that a carbon tax incentivizes companies to reduce emissions (Baumol and Oates, 1971; Climate Leadership Council, 2019; Marron and Toder, 2014; Montgomery, 1972; World Bank, 2017), a carbon tax does not guarantee that such emissions reductions will happen.

A cap-and-trade or cap-and-dividend policy, on the other hand, are alternative policies in which a government sets a limit, or ‘cap,’ on emissions. The cap is imposed by government-issued permits that limit emissions. The government gives, sells, or auctions the permits to companies, creating an opportunity to generate revenue. A cap-and-dividend program would return this revenue to consumers through a rebate.

The logic in asking this question about cap and trade is to assess whether more Americans would favor a greenhouse gas tax if assured that it would result in emissions reductions. However, we show cap-and-trade and cap-and-dividend policies are not notably more popular than straightforward taxes.

In 2024, 52 percent of Americans favor a cap-and-dividend policy, a statistically significant decline from the 63 percent observed in 2020 (see FIgure 11).

Subsidizing solar panels

In 2024, we asked respondents whether the federal government should spend money to help people install solar panels on houses and apartment buildings. Respondents were randomly assigned to be asked one of four versions of the question. Two versions asked about the government paying all of the installation costs, and the other two versions asked about the government paying some of the costs.

For half of each group (chosen randomly), the question was preceded by this introduction:

“Solar panels can generate electricity when the sun is shining, and that electricity can be stored in batteries to be used when the sun is not out. However, companies that make electricity cannot install enough solar panels to make all of the electricity needed in the country. People can put solar panels on the roofs of many houses and apartment buildings so much more of America’s electricity can be made from the sun. But it is expensive to do this, and most people cannot afford to pay that amount of money.”

Among people who did not hear the introduction, 51 percent favor the government paying some of the cost, and 42 percent favor the government paying all of the costs.

Among people who did hear the introduction, 77 percent favor the government paying some of the cost, and 74 percent favor the government paying all of the costs.

Permitting reform

In 2024, we asked whether the federal government should expedite the process of granting permits to build new power plants that make electricity from sources other than coal and petroleum. 52 percent of Americans favor expediting this process.

Least Popular Policies (<50 percent approval)

Nuclear power tax breaks.

Although nuclear power does not directly emit greenhouse gases, tax breaks for the construction of new nuclear power plants are among the least popular policies asked about in 2024. 47 percent of Americans favor this policy; however, it is notable that this is a statistically significant increase from the 37 percent observed in 2020 (see FIgure 12).

All-electric vehicle tax breaks

In 2024, 46 percent of Americans—a record low—think the government should require or give tax breaks to companies to build all-electric vehicles, a statistically significant decline from the 60 percent observed in 2015 when this question was last asked (see Figure 13).

Taxes on consumers

The least popular policies impose new taxes on consumers to incentivize them to consume less fossil fuel. Few Americans favor increasing taxes on retail gasoline and electricity purchases for this purpose. 15 percent approve increasing taxes on electricity, a statistically significant decline from the 28 percent observed in 2020. Likewise, 28 percent approve increasing taxes on gasoline, a statistically significant decline from the 41 percent observed in 2020 (see Figure 14).

Economic Effects of Mitigation Policies

Perceived effect on the economy.

Implementing many policies to reduce greenhouse gas emissions will cost consumers and companies in the short term. Implementing such policies may also increase the cost of American-made goods and services relative to the costs of those goods and services produced elsewhere. This has led some observers to urge caution about implementing greenhouse gas emissions reduction policies (e.g., Cassidy, 2023; Gross, 2021), because they may result in undesirable economic side effects.

However, this argument does not appear to have taken hold with the majority of Americans. For example, only 36 percent of Americans in 2024 believe that taking action to address global warming will hurt the US economy, about the same as was observed in 2013 (30 percent), though this is a statistically significant increase from the 29 percent observed in 2020 (see FIgure 15). Likewise, in 2024, 34 percent believed that these efforts would hurt their state economy, a statistically significant increase from the 24 percent observed in 2020.

More Americans believe that climate action will help the economy. 44 percent of Americans believe this in 2024, about the same as was observed in 2020 (48 percent) and 15 years ago (46 percent) (see Figure 15). 39 percent of Americans believe that efforts to reduce global warming will help their state economy in 2024, a statistically significant decrease from 46 percent in 2020.

Job availability

A similar picture emerged regarding beliefs about how climate action will affect job availability. Only 27 percent of Americans believe that efforts to reduce emissions will reduce the number of jobs in the nation—the same as was observed in 2020 (see Figure 16). And in 2024, 35 percent of Americans believe that climate change action will increase the number of jobs in the country, similar to the 39 percent observed in 2020.

28 percent of Americans believe that climate change action will reduce the number of jobs in their state, similar to the 23 percent observed in 2020 (see Figure 16). And 32 percent of Americans in 2024 believe that climate action will increase the number of jobs in their state, about the same as the 35 percent observed in 2020.

27 percent of Americans believe that efforts to reduce emissions will reduce the number of jobs in the nation

Personal economic impacts

In 2024, we asked respondents about the likely impact of mitigation efforts on their own personal economic situation. A majority of Americans believe that they will have the same amount of money regardless of mitigation efforts (54 percent). 36 percent believe their wealth will decrease, a statistically significant increase over the 20 percent observed in 2020 (see Figure 17). But 8 percent believe that climate change mitigation will increase their wealth, similar to the 10 percent observed in 2020.

Likewise, a majority (64 percent) believe that mitigation efforts will have no impact on their chance of getting a good-paying job. 17 percent believe that mitigation efforts will make them less likely to get a good-paying job, a statistically significant increase from the 12 percent observed in 2020 (see Figure 17). 17 percent believe that mitigation efforts will increase their ability to get a good-paying job, similar to the 16 percent observed in 2020.

Voting in the 2024 Election

Are the many policy preferences outlined in this report just talk, or do they inspire action in the voting booth? We turn to that question next and describe the findings from a test.

Respondents were read a statement by a hypothetical candidate running for a seat in the US Senate and were asked whether hearing that statement makes the respondents more likely to vote for the candidate, less likely to vote for the candidate, or has no impact.

One statement expressed “green views” that summarized opinions expressed by majorities of Americans:

“I believe that global warming has been happening for the past 100 years, mainly because we have been burning fossil fuels and putting out greenhouse gases. Now is the time for us to be using new forms of energy that are made in America and will be renewable forever. We can manufacture better cars that use less gasoline and build better appliances that use less electricity. We need to transform the outdated ways of generating energy into new ones that create jobs and entire industries and stop the damage we’ve been doing to the environment.”

The other statement proposed expanding production of energy from traditionally used fossil fuels:

“The science on global warming is a hoax and is an attempt to perpetrate a fraud on the American people. I don’t buy into the whole man-caused global warming mantra. We must spend no effort to deal with something that is not a problem at all. We should not invest in windmills and solar panels as alternative energy sources. Instead, we should continue to focus on our traditional sources of energy: coal, oil, and natural gas. We should expand energy production in our country, including continuing to mine our coal and doing more drilling for oil here at home.”

Hearing a green view makes 57 percent of Americans more likely to vote for the candidate, a statistically significant decline from the 65 percent observed in 2020 (see Figure 18). Democrats are significantly more likely to be attracted to a “green” candidate (83 percent) than are Independents (56 percent) and Republicans (23 percent) (see Figure 19).

Hearing the green statement makes only 18 percent of Americans less likely to vote for the candidate (see Figure 18).

Hearing the candidate make a “not-green” statement makes only 21 percent more likely to vote for the candidate (see Figure 18). 43 percent of Republicans are more likely to support the candidate, compared to 20 percent of Independents and 5 percent of Democrats (see Figure 19).

Hearing the “not-green” statement makes 63 percent of Americans less likely to vote for the candidate, similar to the 66 percent observed in 2020 (see FIgure 18). This proportion is greatest among Democrats (88 percent), smaller among Independents (62 percent), and still smaller among Republicans (29 percent) (see Figure 19).

Because almost all Republican citizens vote for Republican candidates and almost all Democratic citizens vote for Democratic candidates, the greatest impact of candidate statements in shaping election outcomes is among Independents. Among them, the same pattern appears that appears among all Americans: taking a “green” position helped a candidate and taking a “not-green” position hurt the candidate.

Taken together, these results point to climate change mitigation policies that may be pursued in the future with widespread public support (such as efforts to reduce emissions from power plants). Furthermore, these results also identify a few policy directions that are well received by few Americans, despite being plausible in theory and in practice (like taxes on electricity and gasoline).

For decades, one school of thought commonly followed by some scholars and policymakers is that economic growth and environmental protection are incompatible, and that any efforts to grow the economy must, of necessity, take resources away from helping the environment. Such a presumption creates an “either the economy or the environment” mindset. This mindset has been reinforced by survey questions asking Americans (e.g., Mildenberger & Leiserowitz, 2017), for example: “With which one of these statements about the environment and the economy do you most agree? ‘Protection of the environment should be given priority, even at the risk of curbing economic growth.’ Or: ‘Economic growth should be given priority, even if the environment suffers to some extent’” (Gallup, 2024).

If Americans do perceive this trade-off as inevitable, the COVID-19 pandemic and associated economic crisis a few years ago might have tilted them away from environmental protection generally and away from efforts to mitigate climate change; the subsequent recovery might have increased support for such efforts. The present study refutes that notion resoundingly. In fact, we see small changes in the opposite direction: slightly less public support for some emissions-reducing policies than four years ago.

Few people believe that taking steps to reduce emissions will hurt the national economy, their state’s economy, or their personal finances, and more Americans believe that such policies will improve these economic outcomes.

Finally, we saw that the policy positions candidates take on this issue are likely to influence the votes of many Americans. Thus, policymakers and their challengers have opportunities to use these issues to help assemble the coalitions needed to accomplish electoral victories. By taking “green” positions, candidates gain considerably more votes than they lost.

Download the Report

Click here to view the full report with citations.

PDF — 5.2 MB

Download the Technical Report

Click here to read more about the methodology and survey questions.

PDF — 716.3 KB

Click here to read the report press release.

• Federal Climate Policy
• Comprehensive Climate Strategies
• Climate Insights

Jon A. Krosnick

University Fellow

Jon A. Krosnick is an RFF university fellow and Stanford University-based social psychologist who does research on attitude formation, change, and effects, on the psychology of political behavior, and on survey research methods.

Bo MacInnis

Lecturer, Political Psychology Research, Stanford University

Related Content

Report — Jul 15, 2024

Climate Insights 2024: American Understanding of Climate Change

This first report in the 2024 Climate Insights report series by researchers at Stanford University and RFF examines American public opinion on issues related to climate change.

Policy Leadership Series — Oct 8, 2024

The Future of Environmental and Energy Justice: A Policy Leadership Series Event with Shalanda Baker

Press Release — Aug 27, 2024

Most Americans Approve of Government-Led Climate Action, but Support for Some Policies is Cooling

A new survey of American opinions on climate has found that Americans widely support government action to tackle climate change—but support for some policies has dropped since 2020.

In Focus — Aug 27, 2024

In Focus: American Opinions on Climate Policies

This video shares insights about a new survey that dives into American opinions on federal climate policy, with RFF University Fellow and Stanford University Professor Jon A. Krosnick.

We use cookies to provide you with a better service. Carry on browsing if you're happy with this, or find out more.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Int J Environ Res Public Health

A Methodological Integrated Approach to Analyse Climate Change Effects in Agri-Food Sector: The TIMES Water-Energy-Food Module

Maria maddalena tortorella.

1 School of Agricultural, Forestry, Food and Environmental Sciences, University of Basilicata, 85100 Potenza, Italy; [email protected] (M.V.); [email protected] (M.C.); [email protected] (S.R.)

2 Institute of Methodologies for Environmental Analysis-National Research Council of Italy (CNR-IMAA) C.da S. Loja, 85050 Tito Scalo (PZ), Italy; [email protected] (S.D.L.); [email protected] (C.C.); [email protected] (F.P.); [email protected] (M.S.)

Senatro Di Leo

Carmelina cosmi, patrícia fortes.

3 CENSE (Center for Environmental and Sustainability Research), NOVA School of Science and Technology, NOVA University Lisbon, 2829-516 Caparica, Portugal; [email protected]

Mauro Viccaro

Mario cozzi, filomena pietrapertosa, monica salvia, severino romano.

The European Union’s 2030 climate and energy policy and the 2030 Agenda for Sustainable Development underline the commitment to mitigate climate change and reduce its impacts by supporting sustainable use of resources. This commitment has become stricter in light of the ambitious climate neutrality target set by the European Green Deal for 2050. Water, Energy and Food are the key variables of the “Nexus Thinking” which face the sustainability challenge with a multi-sectoral approach. The aim of the paper is to show the methodological path toward the implementation of an integrated modeling platform based on the Nexus approach and consolidated energy system analysis methods to represent the agri-food system in a circular economy perspective (from the use of water, energy, biomass, and land to food production). The final aim is to support decision-making connected to climate change mitigation. The IEA-The Integrated MARKAL-EFOM System (TIMES) model generator was used to build up the Basilicata Water, Energy and Food model (TIMES-WEF model), which allows users a comprehensive evaluation of the impacts of climate change on the Basilicata agri-food system in terms of land use, yields and water availability and a critical comparison of these indicators in different scenarios. The paper focuses on the construction of the model’s Reference Energy and Material System of the TIMES model, which integrates water and agricultural commodities into the energy framework, and on the results obtained through the calibration of the model β version to statistical data on agricultural activities.

1. Introduction

The 2030 Agenda for Sustainable Development identifies 17 Sustainable Development Goals (SDGs) to be achieved by 2030, which aim to encourage a change in the current development model regarding the environmental, economic, and social dimensions [ 1 ]. This document represents one of most important global agreement that highlights an integrated and multi-sectoral vision of the different dimensions of sustainable development. It represents an important reference for the nexus approach, addressing the risks and changes associated with the reduced availability of water, energy, and food, in a growing World’s population context (8.6 billion by 2030 and 9.8 billion by 2050) [ 2 ]. In addition, the Paris Agreement commits the signatory parties to reduce drastically their greenhouse gas (GHG) emissions and to take urgent actions to combat climate change and its impacts, supporting a transformation of anthropogenic activities toward more sustainable trajectories [ 3 ]. As concerns Europe, the need to make production and consumption patterns more sustainable is also emphasized by the European Green Deal, which aims to make Europe climate-neutral in 2050. In view of this ambitious goal, the political objectives and targets set for the period 2021 to 2030 in the EU’s 2030 climate and energy policy framework will be made more ambitious to effectively support the transition to a climate-neutral economy [ 4 ] and to implement the Paris agreement commitments.

To reach the policy objectives, it is necessary to implement coordinated actions that can guarantee economic growth and at the same time a drastic reduction of GHG emissions, to mitigate climate change and support environmental protection.

The “Nexus Thinking” and its multi-sector approach are therefore crucial to respond to the sustainability challenge for an effective management of resources in compliance with the SDGs, the Paris Agreement, and the European climate neutrality goal [ 5 ]. Concepts of the Nexus Thinking were successfully applied in developing countries [ 6 , 7 ], and currently also to countries with a more advanced economy [ 8 , 9 , 10 ]. Since 2011, when it was first brought to the attention of the institutions in the opening report of the World Economic Forum [ 11 ], the Nexus approach took on a central role as a method for understanding and modeling the complex interactions between the different resource systems (energy, water, food) [ 12 ]. It has therefore been successfully used both by academics and international organizations (e.g., The Food and Agriculture Organization of the United Nations (FAO), The United Nations Economic Commission for Europe (UNECE), World Wide Fund for Nature (WWF), etc.). In particular, the Nexus approach aims at managing efficiently water, energy, and food systems as a whole, minimizing potential conflicts and strengthening intersectoral integration in order to guarantee a secure and sustainable use of resources [ 13 ]. In fact, the complex interconnections between energy, water and food are difficult to represent due to the numerous variables and phenomena involved. Water affects food production (e.g., crops, livestock) as well as energy production (e.g., hydropower, cooling water). Energy affects food production (e.g., energy for chemical and mineral fertilizers, transportation, and food storage) and also water supply (e.g., water preparation, desalinization, pumping). Agriculture, a major player in food production, is a major user of water (over 70% of all water consumption globally [ 14 ]) and energy (about 30% of the total energy demand). It also affects the water sector through land degradation, changes in runoff and disruption of groundwater discharge. On the other hand, the area available for agricultural activities must also compete for a share of electricity generation from fossil and renewable sources, both in terms of the area required for the installation of power plants and the impact of the related activities (e.g., mining, dams and water flow management, biofuel production, etc.). In particular, cultivation of biofuels, which has a high profitability per hectare and, in many cases, benefits of public incentives, causes an excessive exploitation of territories, generating, indirectly, a potential pressure on the prices of food crops and increasing the competition between land use and water consumption to produce biofuels or food. Rulli et al. [ 15 ] estimated that the worldwide production of biofuel exploits 4% of the land and water used for agriculture, corresponding to an area sufficient to feed about 280 million people if used to grow food.

In this already complex situation, it is necessary to consider the negative effects of climate change that as known, affects the availability of resources and land, with significant impacts on the water-energy-food system. A recent Intergovernmental Panel on Climate Change (IPCC) Report [ 16 ] highlights how climate change increases the rate and extent of ongoing land degradation through two main factors: increased frequency, intensity of heavy rainfall and extreme high-temperature events. Furthermore, global warming will make soil degradation processes more severe in the various geographical areas due to an increased frequency of floods, droughts, cyclones and hurricanes, forest fires and sea levels rise.

The greatest risk in using the nexus approach without considering the data of climate models is to overlook the possible effects of climate change on the balance between the resources involved in the water-energy-food cycle and within their interactions [ 17 ].

In fact, both agriculture and energy production are vulnerable to changes in meteorological parameters and to the occurrence of extreme events such as drought and floods [ 18 ]. Temperatures and precipitation levels can strongly influence the availability of water, and consequently the production of energy and food. Crop productivity may increase in Northern Europe due to prolonged thawing period opening the possibility for new crops cultivation [ 19 ]. In Southern Europe, and in particular in the Mediterranean basin, crop productivity is negatively affected by droughts, heat waves, reduced water availability and other related phenomena such as pest and disease epidemics [ 20 ].

From this perspective, the adoption of a holistic approach that allows representing the interrelationships between the three sectors (water, energy, and food) is a priority to encourage a sustainable and efficient use of resources, reduce risks and define effective integrated policies.

Numerous examples in the literature underline the increasing importance of an integrated approach to water-food-energy challenges and different models have been used to assess WEF interactions ranging from economic to technological, and geographic information systems (GIS) tools. In particular, Endo et al. [ 21 ] analyzed the water, energy, and food nexus by reviewing 37 projects across different world regions, to highlight the current state of art by investigating the nexus keywords and stakeholders to characterize the specific nexus type.

Haji et al. [ 22 ] used a ‘Node’ methodology based on GIS-based approaches, the Analytical Hierarchy Process and resource assessment to evaluate the critical factors that increase the risk in open field farms and to improve water and energy efficiency. In Pakdel et al. [ 23 ] a multi-objective optimization of energy and water management based on GAMS software is performed to minimize both energy and freshwater use, introducing the concept of “energy hubs” and validating the interdependency of energy and water structures. In Lee et al. [ 24 ] a nexus approach is used to analyze the water-food-energy interconnections and their economic implications in the sugar industry in India. Nie et al. [ 25 ] focus on the agricultural system and use a multi-objective procedure and a comprehensive WEF index to select optimal land allocation strategies that can limit stresses in the water-energy-food nexus. Chiodi et al. [ 26 ] integrated the energy and agriculture systems into the IEA-ETSAP methodology to individuate the GHG reduction strategies for Ireland.

Many gaps still need to be filled in the operational application of the nexus concept for decision-making. However, Weitz et al. [ 27 ] point out that despite the still open questions, “a nexus approach promotes policy coherence through identifying optimal policy mixes and governance arrangements across the water, energy and food sectors”.

These considerations and the relevant examples discussed above, have strengthened our motivation to develop a modeling platform focused on the agricultural system that integrates the nexus concept in a framework typically used to support decision-making when different competing goals must be achieved. The outcome of this research, The Integrated MARKAL-EFOM System (TIMES)-WEF model, will be validated in selected areas of Mediterranean Europe, in order to evaluate the robustness of solutions at different spatial scales, and to perform a joint assessment of the effects of climate change and agricultural policies. The scenario analysis will focus on two IPCC pathways (RCP4.5 and RCP8.5) and multi-level (European Union, national and local) agricultural, energy and climate policies to evaluate their effects in terms of availability of land, water and energy as well as other parameters of interest for the agri-food system (e.g., fertilizers, pesticides, etc.).

The paper is structured as follows: Section 2 presents a literature review of the Nexus modeling approach and its application to the development of the TIMES-WEF model; Section 3 describes data requirement, technical assumptions, the preprocessing procedure for the implementation of the model data input and the calibration to the statistical base year data of the TIMES-WEF model for agriculture; finally, Section 4 concludes with the main outcomes and future development of the model.

2. Modeling WEF Nexus

2.1. state of art.

The WEF Nexus currently represents the most advanced methodological and operational approach to address the complexity of sustainable development. It aims to overcome the silo vision and evaluate the interdependencies and management of the different sectors (energy, water, food) as an integrated process. In this way, it is possible to highlight how actions in a sector can influence the management of resources, thus avoiding unwanted consequences and exploiting existing potential synergies.

The conceptualization of the WEF Nexus has become increasingly complex, incorporating a plurality of factors and dimensions e.g., environmental, economic, political, and social. Conversely, there has been a slow development of analytical approaches [ 28 , 29 ] and limited use of modeling tools to assess the correlations between water, energy and food and support an integrated decision-making process [ 30 ].

Table 1 reports a summary of main methods used in the literature to address the WEF Nexus (are included in the table also the methods that analyze only some of its components).

Summary of main methods used in the literature to address the Nexus.

However, many applications only focus on dual sector interactions, for example water-food or water-energy, thus implementing a fragmented vision of the WEF Nexus [ 42 ], or provide a narrow perspective of the interactions between water, energy and food, with a limited ability to capture multi-sectoral interconnections and interdependencies between different systems [ 43 , 44 ].

Few studies are based on innovative methods to quantify the connections and interactions between sectors, in order to better describe the systems included in the WEF Nexus. These are modeling platforms that can support the integration of sectoral models, creating flexible tools that can accommodate new modeling inputs or extensions. In this way, decision support tools are created to combine physical models with scenario analysis, allowing decision makers to compare the impact of different policies or actions on the analyzed system [ 45 , 46 , 47 ].

Based on the above considerations, the proven validity of the Nexus approach to tackle a multi-objective problem such as an integrated management of energy, water, and soil resources [ 48 ] guided our idea to develop an integrated decision support platform, based on existing consolidated models used for policy analysis.

Taking inspiration from the Irish model [ 26 ], the TIMES energy models’ generator [ 38 ] has been used to model the agri-food system within an energy system analysis approach based on technical engineering and economic analysis to ensure a sustainable management of agricultural resources.

2.2. Overview of ETSAP-TIMES

The TIMES (The Integrated MARKAL-EFOM System [ 38 ], (developed by the Energy Technology Systems Analysis Program (ETSAP) of the International Energy Agency (IEA), an autonomous intergovernmental organization born with the 1973–1974 oil crisis and based in Paris, France) is a bottom-up model generator, which uses linear-programming to compute a least-cost energy system, optimized according to several user exogenous constraints, over medium to long-term time horizons [ 38 ]. It is widely used to represent local, national, and multiregional energy systems and to perform scenario analysis, exploring possible energy futures in relation to environmental and technical constraints, such as policy measures.

The TIMES models are driven by the end-use sector demands (Industry, Residential, Commercial, Transport and Agriculture).

The energy system configuration is optimised to provide the least-cost solution that corresponds to the best allocation of resources and technologies, which fulfil end-use demands and the scenarios’ constraints at the minimum total discounted cost of the system. The optimization of the reference scenario provides the baseline for the comparison of solutions in the alternative scenario analysis.

The TIMES model structure is usually described through the Reference Energy and Materials System (REMS), which describes the entire supply–demand chain, providing an accurate representation of energy flows from supply/conversion technologies to demand processes. It allows representing all the components related to energy production and use, including emissions and materials. The supply chain describes the extraction import/export and secondary production of primary resources (typically energy and materials) whereas the demand chain represents in detail the commodity flows through the network of real or dummy technologies (or processes), (e.g., mining processes, import processes, energy transformation plants, end-use devices). Any item produced or consumed by a certain technology is called “commodity” (e.g., energy carriers, energy services, materials, money flows and emissions).

The key inputs to the TIMES model deals with all specific data that characterize the system under focus: energy demand, primary energy supply (availability of present and future sources), techno-economic factors (technology development and associated costs), environmental variables (e.g., GHG emission factors), and other policy parameters.

This research takes the standard TIMES modeling framework as a starting point to develop a novel model that focus on the agri-food system through a WEF nexus perspective and that can be merged into the general energy modeling framework, exploiting all conversion processes and the end-use sectors related to agriculture.

2.3. TIMES-WEF Model

The TIMES-WEF model represents an innovative application of the water-energy-food nexus approach into the ETSAP-TIMES framework where land use is chosen as an independent driving parameter to connect soil availability with input/output commodities. In particular, a land use-driven model allows evaluating directly the effects of climate change and energy-environmental policies in terms of use of resources (energy, water, and land use), agricultural productivity, highlighting the synergies among the different sectors. The land use demand on the whole time horizon represents therefore the “end-use demand” of the energy and material model to be fulfilled at the minimum feasible cost in compliance with the exogenous constraints on resources. The general objective is to ensure an optimal management of the territory, i.e., able to improve the use of endogenous resources, increase the resilience of the agri-food sector to climatic events and facilitate the implementation of agricultural, energy and environmental policies.

The analytical structure reported in the flowchart represented in Figure 1 was therefore designed to represent the agri-food system and characterize its data input.

TIMES-WEF flowchart, inspired by Chiodi, 2016 [ 26 ].

In this modeling approach, the Used Agriculture Area (UAA) and the Forestry Area (FA) represent the output commodities. More precisely, the UAA refers to the total area (hectares) used for agriculture, which includes arable land, permanent meadows, permanent crops, and vegetable gardens used by farms with reference to Eurostat data [ 49 ]. The FA, instead, is representative of the hectares of surface area covered by forests or the canopy of the forest or open woods.

The agricultural and forestry activities are modeled as end-use processes (dummy processes) with associated input and output commodities, operating costs, and other key parameters characterizing these practices. New elements such as water, fertilizers, pesticides, and CO 2 capture from forestry were included among the input commodities of a standard TIMES model (energy vectors and materials). Biomass residuals from agriculture and forestry, greenhouse gas emissions from both the combustion processes and agricultural activities are modeled as process outputs.

The characterization of agricultural activities was defined according to the standard classification of farming classes used in the main European and national databases. Specifically, 10 categories were considered: arable crops, horticulture, viticulture, olive growing, fruit growing, herbivores livestock, granivorous livestock, polyculture, mixed livestock and mixed (that include livestock and crops).

Each agricultural activity was represented by two processes in series: (i) The first process consumes water, energy (electricity, diesel, natural gas), pesticides and/or fertilizers and produces crops (expressed in ton) or cattle (expressed in livestock unit (LSU)); (ii) The second process converts the productions of the first process in hectares of used agricultural/livestock area through a yield parameter. The sum of the outputs of the second processes of all ten categories of agricultural activities provides the demand for end use, i.e., the Used Agriculture Area. The only exception is represented by mixed activities that were modeled as a single process. It has energy and water consumption as input and hectares used as output.

Forests play a multifunctional role, contributing to the protection of biodiversity and the environment (through carbon sequestration) and to the economy (through the production of biomass as an energy resource). They are also particularly affected by climate change (droughts, forest fires, etc.), which reduce their carbon sequestration power and bioenergy resource potential.

The detailed structure of the TIMES-WEF module is described by the flowchart in Figure 2 that shows the flows of commodities through processes, from resource mining to end-use demands (i.e., UAA and FA).

The TIMES-WEF Reference Energy and Materials System (REMS).

The analyzed time horizon covers a period of 50 years, from 2010, the model base year, to 2060 and was divided into time intervals of 5-years each (model’s time slices), considering 2030 as a milestone. This long-term time horizon makes it possible to evaluate the effects of the strategies for achieving the targets of the 2030 Agenda beyond the year 2030 and to trace the path toward the Energy Roadmap 2050 through a scenario analysis. Furthermore, from a modeling point of view, the choice of a time horizon that goes beyond 2030 derives from the need to harmonize the timing of the TIMES-WEF to the TIMES Basilicata energy system module [ 50 ] to allow its integration and increase the reliability of the solutions in a long-term perspective. Therefore, in the future stages, the integration of the TIMES-WEF agri-food model into the whole energy system model, will make possible to have a comprehensive perspective of the synergies and competition between the two sectors in a circular economy perspective.

The TIMES-WEF module data input is made up of a set of an Excel spreadsheet, structured around three type of file:

• ▪ The Base Year Templates: they contain the basic data about input commodities (energy, water, fertilizers), operating costs and output commodities (products and by-products-including straw, manure and emissions, and Used Agricultural Area) in order to characterize the Forestry and Agricultural Activities in the base year. They provide the statistical data for the model calibration.
• ▪ A “Technology Repository”: that is a virtual basket of alternative options for agricultural practices, described by technical (e.g., efficiencies, lifetime, emission factors) and economic parameters (e.g., investment and operation costs). These options can be implemented over the time horizon to replace the current processes to fulfill the exogenous constraints.
• ▪ The Scenario Files: a set of spreadsheets containing coherent demand projections, exogenous constraints on resources availability, and other parameters by scenario.

2.4. The Basilicata Region Case Study

To test the applicability and consistency of the integrated approach developed through the TIMES-WEF modeling module, the TIMES model was customized according to the Basilicata region data as a pilot case. Basilicata is a small region located in the South of Italy bordering with Campania on the West, Apulia on the North and East, and Calabria on the South. The region covers 10,073 square kilometers and has a population of 562,869 inhabitants [ 51 ] with a quite low regional population density (57.8 inhabitants per km 2 ).

Basilicata is a relevant case study mainly due to the urgent threat that the effects of climate crisis represent for its territory. In fact, 55% of the Region is at risk of desertification [ 52 ] endangering the future of the agricultural sector, which plays a crucial role in the local economy. Therefore, for this Region it is a priority to define policy mitigation and adaptation actions to increase the resilience of the territory and the agriculture sector, identifying sustainable pathways of local resources. Due to the wide morphological difference, mainly in elevation, the Region is characterized by a varied climate that ranges from the continental one in the internal areas to the Mediterranean one of the coastal areas. There are six distinct soil and climatic zonas (Ionian, Bradanica, Northern Apennines, North Western Apennines and South Western Apennines, Tyrrhenian) in which climate deeply influences the type of agriculture, in particular in the internal and non-irrigated areas [ 53 ].

The territory is mainly mountainous (47%) and hilly (45%) with a modest flat percentage (8%). The total agricultural area is 716,838 hectares accounting for about 70% of the regional surface area. The forest area according to the Regional Forest Charter is 355,409 hectares, characterized by variegated species in both environmental and vegetation terms which make the regional territory a mosaic landscape.

As happens in several Mediterranean countries under the pressure of contingent factors (climate change, changes in land use, over-exploitation of resources, etc.), large areas of Basilicata are particularly susceptible to land degradation [ 54 , 55 , 56 ]. This situation highlights the importance of a more sustainable management of resources and forests to improve the resilience of the territory and guarantee the functioning of a key sector such as agriculture [ 57 ]. Therefore, it is a priority for this region to define comprehensive mitigation and adaptation actions that increase the resilience of the territory and to identify sustainable pathways in the use of local resources aimed at improving the agricultural sector. Agriculture is, in fact, one of main activities of the Basilicata economy [ 58 ], along with industry (manufacturing, automotive, and especially oil extraction) and services, where the importance of tourism is increasing. Despite its small contribution compared to other sectors, with an added value of 3% in 2019, the agricultural sector has a significant weight in terms of exports and employment, registering for the latter, a positive trend in recent years reaching an increase of 7% in 2019, in contrast with the other sectors [ 59 ]. Therefore, the high level of specialization achieved in the agri-food sector (characterized by the production of a wide range of high quality food (most of which included in national list of Protected Designation of Origin (PDO), Protected Geographical Indication (PGI), and Controlled and Guaranteed Designation of Origin (DOCG, the Italian acronym) marks of traditional food and wine products) could be a potential strength to increase the competitiveness of the entire regional system. From this perspective, the 2014–2020 Rural Development Program played a key role by providing a valuable financial support (680 M€) to encourage innovation in this sector in order to improve its economic and sustainability environmental performance [ 60 ].

As regard energy resources, the Basilicata region hosts the Europe’s largest onshore oil and gas field, with an annual production of 48,550,554,911 BBOE of oil and 1,493,816,334 Smc of natural gas [ 61 ]. Furthermore, Basilicata achieved remarkable targets in the production of renewable electricity, reaching and encompassing the goals set by the Regional Environmental Energy Plan (PIEAR) for 2020, namely 981 MW of onshore wind (60% of the total renewable capacity), 359 MW of solar-photovoltaic (20%), 50 MW of biomass (15%) and 48 MW of hydroelectric (5%). The regional authorities, through various policy measures, have also encouraged the use of biomass to produce thermal energy. In 2017, 45% of the regional energy requirement (thermal and electric) was met by renewable sources, also in this case anticipating the objective of the national legislation (33%) expected by 2020 [ 62 ].

The information necessary to implement the TIMES-WEF module for the Basilicata case study was collected by elaborating the national and European statistical sources of data (FADN, RICA and ISTAT), the Regional Environmental Energy Plan–PIEAR and other local sources (e.g., irrigation water, fertilizers and agricultural diesel prices for which there are different values depending on the region) ( Table 2 ).

List of data source and type of information provided.

2.5. Scenarios

The scenario analysis has two main objectives in this research: (I) evaluating the effects of climate change in terms of land use variations and resources availability (water and land); and (II) assessing the consequences of EU, national and regional environmental and agricultural policies on the entire agri-food system of Basilicata region.

A reference Business-as-Usual (BaU) scenario was set to provide a benchmark for comparing the alternative scenarios solutions. It represents the development of the system under the policies in force, both energy (through Regional Environmental Energy Policy Plan-PIEAR), and agricultural. The reference scenario shows the “status quo” evolution of the Basilicata Region agricultural system in term of land use availability, resources, technologies, and policy in place. The land use demand by category is projected along the analyzed time horizon (2010–2060) using appropriate statistical techniques.

The selection of scenarios considers the phenomena in progress and the most urgent challenges to be faced in the Mediterranean Europe region for mitigation and adaptation to climate change by focusing on the agri-food sector.

In fact, this area is significantly affected by climate change, especially agriculture. The intensity and frequency of extreme weather event (in particular, heat waves, flooding, wildfires) accelerate the degradation of agricultural land and cause a substantial decrease in yields, with an estimated average loss of 3.24 ton/ha per year compared to 2010 (reference year) [ 71 ]. As for the consequences in terms of WEF Nexus, droughts will favor a greater demand for water, with an increase predicted by the Representative Concentration Pathways (RCP) 4.5 and 8.5 climate scenarios between 4% and 18% by the end of the century. This will inevitably reduce the availability of water for irrigation, undermining the suitability of the land for rain crop production [ 72 ].

Indeed, through more sustainable agricultural and forestry practices it is possible to reduce the environmental impact of this sector, strengthen the carbon capture capacity of soils and forests and protect biodiversity.

From this perspective, the agriculture and forestry sector has been included for the first time in the greenhouse gas emission reduction targets set by the European Union (EU) for 2030 [ 73 ] and has been at the center of two fundamental strategies of the Green Deal, “Farm to fork” and “Biodiversity 2030” [ 74 ].

As highlighted by Nikolakopoulou [ 75 ] “food related targets run throughout the Sustainable Development Goals (SDG) and they are often interconnected”. In particular, a sustainable agriculture system, in line with the SDGs, should be more resilient to climate risks, protect the environment and deliver healthy and affordable products to fulfill the demand of an increasing population. The new European Green Deal package of measures will have a strong impact on future planning, reinforcing the orientation already taken by the Common Agricultural Policy (CAP), aimed at favoring a progressive abandonment of intensive agriculture in favor of more sustainable cultivation techniques that preserve the soil quality and fertility by reducing the use of fertilizers and pesticides.

Taking into account the complex environmental and policy framework, two alternative classes of scenarios have been defined as follows:

• ▪ Climate Scenarios modeling the relationship between climate change and land use. They highlight, on the one hand, how agriculture is affected by climate change, and on the other hand, how crucial is the role of this sector within the greenhouse gas emissions reduction strategies. These scenarios are characterized by water and land availability as quantitative parameters.
• ▪ Policy Scenarios modeling the EU, national and local policies on energy, environment and agriculture in quantitative terms such as percentage of use of pesticides, fertilizers, area for organic agriculture and GHG emissions reduction targets.

The climate scenarios parameters have been selected coherently with the IPCC 2019 report [ 16 ]. Moreover, data of water availability and consumption was taken from the Inter-Sectoral Impact Model Intercomparison Project (ISI-MIP) [ 76 ], for the RCP 4.5 and RCP 8.5 climate scenarios.

The scenario analysis will identify sustainable development paths for the agricultural system taking into account the general objectives of the EU Green Deal and the “From Farm to Fork” strategy. In this framework, the contribution of the various alternatives to achieving the SDGs relating to energy, climate and sustainable production and consumption models will be assessed (namely SDG 7 “Clean and accessible energy”, SDG12: “Responsible consumption and production” and SDG 13 “Climate action” [ 77 ]).

3. Results and Discussions

3.1. data input construction.

Statistical data pre-processing is a fundamental step for the implementation of the model’s data input.

Following the specifications of the TIMES-WEF model, the agricultural system has been characterized on the basis of the agricultural area (driving parameter), dividing, for the reference year, the total agricultural area in the region (502,197 hectares) into the shares of the various types of farming that characterize the local agricultural system ( Figure 3 ).

Share of land use by type of farming in Basilicata region.

The estimations were based on the data of the Agriculture Census for 2010 [ 63 ], the latest available survey of the National Statistical Institute, integrated with the information of the annual sample survey of Agricultural Accounting Information Network for Italy (RICA) [ 65 ], included in the European Network of Farm Accounting Data (FADN) [ 78 ].

The FADN dataset represents the main information source of the European Commission on the Member States to assess the impact of the proposed changes to the CAP through the simulation of different scenarios on corporate sustainability (economic, environmental, social and innovations).

The RICA data were therefore preliminarily processed to represent the agricultural system of the Basilicata region in relation to the 10 categories considered in the model. Further elaborations were also required to model in detail the features of specific sub-categories.

After determining the land use demand by type of crop, the subsequent step was to estimate the energy consumption by type of farming in order to determine the energy balance. This was done considering the Basilicata Regional Environmental Energy Policy Plan (PIEAR) data for Agriculture ( Table 3 ), which was disaggregated through a weighting procedure based on RICA information on the annual energy expenditures for diesel, natural gas and electricity incurred by local farms.

Energy Consumption in Agriculture for the Basilicata Region (PIEAR data).

In addition, a further disaggregation was necessary for permanent crops, represented in the TIMES-WEF model by the following categories: viticulture, fruit and olive growing. The energy consumption of each category was estimated by a weighting procedure based on the used agricultural area. Table 4 summarizes the percentage breakdown of the estimated energy consumption for each type of farming.

Percentage breakdown of energy consumption.

The costs of the energy carriers for the reference year were estimated considering different sources of data. In particular, the electricity price (59 Euro/MWh, to 16.4 MEuro/PJ) was estimated on the basis of the National Energy Services Operator data [ 70 ], diesel price for agriculture (0.60 Euro/liter, 16.2 MEuro/PJ) using the local Chambers of Commerce data [ 66 ] and natural gas price (0.66 Euro/liter, 25.3 MEuro/PJ) using the Ministry of Economic Development data [ 69 ].

Similarly, the share of water consumption for the ten agricultural activities was calculated using the aggregate data provided by the National Institute of Statistics, assuming a sales price of 0.47 euro per cubic meter [ 68 ]. The estimated values by type of crop are shown in Table 5 .

Water consumption by type of farming.

The three most important fertilizers used in agricultural practices, potassium (K), phosphorus (P), and nitrogen (N), were also included in the model data input, estimating the tons of fertilizers used in the base year 2010 [ 65 ] starting from the hectares of agricultural area for each type of crop and the tons of product ( Table 6 and Table 7 ).

Consumption of fertilizers per hectare by type of crop.

Consumption of fertilizers per ton of product and by type of crop.

The average costs of fertilizers were estimated based on the Turin Chamber of Commerce data [ 67 ] assuming a constant value at national level. Specifically, 359 Euro/ton for Nitrogen, 328 Euro/ton for Phosphorus and 498 Euro/ton for Potassium were considered.

Once characterized the input commodities (diesel, natural gas, electricity, water, and fertilizers), it was necessary to characterize the production processes as fictitious technologies (dummy processes) estimating the operating costs, since the technical parameters are not of interest in this approach.

The agricultural production (expressed in tons) and the number of cattle (expressed in in livestock unit-LSU) for 2010 for each type of agricultural activity were estimated by the RICA database. Furthermore, the fixed and variable operating costs incurred by the farms were estimated considering their annual budgets ( Table 8 ).

Production and operating costs of agricultural activities.

Another aspect of the setup of the model’s data input concerned waste production that was estimated considering the statistical data relating to the production of straws and manure by crops and livestock type [ 79 ], harmonizing these data to the categories of the TIMES–WEF model ( Table 9 ).

Waste production by crops and livestock type.

In addition, the emissions of the main greenhouse gases (CO 2 , CH 4 and N 2 O) were estimated by considering the emission factors provided by the United Nations Framework Convention on Climate Change (UNFCCC) database [ 80 ]. In the base year calibration of the β version of the TIMES-WEF model, only the energy combustion processes were initially considered. The emissions from processes will be added subsequently.

3.2. Base Year Calibration

Once the data input for the base year has been implemented, the subsequent fundamental step concerns model calibration to statistics [ 81 ]. This is essential to validate the modeling approach and to refine the initial data. The main results of the TIMES-WEF calibration are shown in this section. Figure 4 shows the energy consumption by type of farming. Diesel oil is the most used fuel (1.74 PJ), (in particular from arable crops (0.8 PJ), polyculture (0.25 PJ) and herbivores livestock (0.19 PJ)) followed by electricity (0.23 PJ) and natural gas (0.04 PJ). Granivorous livestock, mixed livestock and mixed farming categories do not use natural gas.

Energy consumption for each type of farming. Base year (2010).

Figure 5 shows water consumption of the 10 agricultural activities. As expected, fruit growing and polyculture have the highest consumption, 47 Mm 3 and 34 Mm 3 respectively, as they are typical “water demanding” categories. Arable crops consumption is about 11 Mm 3 , while granivorous livestock consumption is negligible.

Water consumption by farm typology. Base year (2010).

As concern the use of fertilizers ( Figure 6 ), fruit growing shows the highest consumption of both Phosphorus (18.02 kton) and Nitrogen (8.66 kton), while viticulture shows the highest consumption of Potassium (14.53 kton). Arable crops, which represent an energy-intensive category covering the highest percentage of territory, show a high consumption of Phosphorous and moderate consumption of Nitrogen (0.41 kton) and Potassium (1.32 kton). On the other hand, olive growing shows an almost similar consumption for all three fertilizers (7.88 kton for Nitrogen, 8.21 kton for Phosphorus and 6.01 kton Potassium).

Fertilizers consumption for each type farm. Base year (2010).

As regards waste production ( Figure 7 ) arable crops provide the highest values of dry matter (442.46 kton). Herbivorous livestock (64.66 kton of manure), fruit growing (59.27 kton of dry matter), polyculture (56.37 kton of dry matter) and olive growing (46.59 kton of dry matter) mixed livestock and granivorous produce a very low amount of waste (respectively 0.22 kton and 0.06 kton of manure) also produce small quantities while waste production from horticulture is negligible.

Waste production by farm category. Base year (2010).

In the calibration runs, CO 2eq emissions from combustion processes are entirely determined by fossil fuel consumption for agricultural activities, with a total amount of CO 2eq of 131 ton for the base year, the highest contribution being provided by arable crops (61 ton of CO 2eq ) while the lowest by granivorous livestock (0.3 ton of CO 2eq ).

Figure 8 provides an overview of the distribution of the different commodities (energy consumption, water consumption, fertilizers consumption, waste production and CO 2eq emission production) by farming activity.

( a ) Consumption and emissions distribution by type of agricultural activity ( b ) Weight of agricultural activities in resource consumption and emission production.

In the two polar maps of Figure 8 , energy and fertilizer consumption were aggregated. Arable crops represent the category with the highest energy consumption (44%), associated CO 2eq emissions and waste production (64%). Fruit growing has the greatest weight in both water (37%) and fertilizer consumption (23%). Polyculture is the second category for water consumption (27%) and viticulture for fertilizer consumption (21%). Granivorous show the lowest energy consumption (0.3%), associated CO 2eq emissions (0.3%) and waste production (0.1%).

The calibration of the β version of the TIMES-WEF model highlights the differences between the various agricultural activities in terms of resources used and output, confirming the validity of the modeling approach and preparing a solid reference structure for the implementation of the scenarios.

4. Conclusions

This paper presents the fundamental steps for the implementation of a modeling platform that integrates the water-energy-food nexus approach into the energy system modeling framework, with the aim to investigate the impacts of climate change and EU policies on the agricultural system.

In fact, agriculture plays a crucial role in achieving sustainable development goals and climate targets. The rational use of soil, water and energy is essential to ensure the well-being of the population and adequate food production, while the adoption of sustainable agricultural practices can help mitigating the effects of climate crisis and supporting adaptation to extreme weather events.

The Nexus approach, increasingly used for an integrated vision of the main challenges of sustainability, allows highlighting the interdependencies between the three key variables Water-Energy-Food and to understand how they are influenced by climate change and the policies in place. The study goes beyond the state of the art [ 26 ] by designing, implementing, and validating a modeling approach based on energy system analysis, widely used in policy assessment.

A broad scientific debate is underway about the choice of modeling tools and indicators for defining policy strategies and measuring progress toward the achievement of sustainable goals. In fact, the selection of indicators can deeply influence the results of the monitoring [ 82 ] and the implementation of SDGs, requiring broader qualitative analyses [ 83 ].

In this context, the use of a modeling framework based on the IEA-ETSAP methodology, explicitly designed for long-term energy-environment analyses, to design the least-cost pathways for a sustainable development, allows ensuring transparency in the basic assumptions and a high detail in the identification of the possible strategies.

The integration of the nexus approach in a comprehensive partial equilibrium model based on the ETSAP-TIMES structure makes possible to set up a robust platform to identify the optimal allocation of energy and material resources in compliance with the EU strategic policy targets and to explore possible alternatives, measuring their effectiveness in terms of economic, energy and environmental indicators.

An innovative modeling approach based on land use as driving variable was adopted to develop the TIMES-WEF model whose database includes non-energy resources (water, fertilizers, pesticides) among the input commodities while food and biomass residuals represent the output commodities. The β version of the model was customized and calibrated on the agriculture system of the Basilicata Region, to enable its straightforward integration with the former TIMES Basilicata energy model [ 50 ] in a circular economy perspective.

Starting from the official European Union Classification [ 84 ], ten end-use categories representing the entire agri-food system were identified (arable crops, horticulture, viticulture, fruit growing, growing olive growing, polyculture, livestock herbivorous, granivorous livestock, mixed and mixed livestock) and the model database was set up accordingly.

The modeling approach was validated by calibrating the model β version and evaluating the congruence of the results to the statistics.

The next steps will concern, first, the optimization of the model under the Reference Scenario, representing a “status quo” development along the time horizon and, second, a scenario analysis to identify suited roadmaps for a sustainable development of the regional agricultural system.

The proposed innovative modeling framework applied to a regional energy and agriculture system can contribute to supporting decision makers in a complex governance of a system in which conflicting objectives must be faced. In particular, a thorough analysis of the water-energy-food system will allow enhancing the role of the agriculture and forestry sectors to achieve the national CO 2eq reduction and RES targets, evaluating the effectiveness of different mitigation strategies.

At the same time, the assessment of the CAP at the local scale will provide insights to coordinate the implementation of policies at local and national level with a long-term perspective.

Acknowledgments

The authors are grateful to anonymous reviewers for their helpful comments and suggestions that resulted in an improved quality of the manuscript and gratefully acknowledge the Portuguese Foundation for Science and Technology (FCT) for the support provided to CENSE through the strategic projects UIDB/04085/2020.

Author Contributions

Conceptualization, M.M.T., S.D.L., C.C., P.F., M.V., M.C., F.P., M.S. and S.R.; methodology, M.M.T., S.D.L., C.C., P.F., M.V., M.C., F.P. and M.S., software, M.M.T. and S.D.L.; validation, M.M.T. and S.D.L.; formal analysis, M.M.T., S.D.L., C.C.; data curation, M.M.T., S.D.L., P.F. and M.V.; writing—original draft preparation, M.M.T., S.D.L. and C.C.; writing—review and editing, M.M.T., S.D.L., C.C., P.F., M.V., M.C., F.P., M.S. and S.R.; supervision, C.C., P.F., M.V. and S.R.; project administration, S.R.; funding acquisition S.R. All authors have read and agreed to the published version of the manuscript.

This research was funded by the regulation “Dottorati Innovativi con Specializzazione in Tecnologie Abilitanti in Industria 4.0”, funded by the Basilicata Region-Industry 4.0 Strategy, period 2017–2020.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Article  

• Volume 15, issue 4
• ESD, 15, 1117–1135, 2024
• Peer review
• Related articles
• Tipping points in the Anthropocene

Tipping point detection and early warnings in climate, ecological, and human systems

Vasilis dakos, chris a. boulton, joshua e. buxton, jesse f. abrams, beatriz arellano-nava, david i. armstrong mckay, sebastian bathiany, lana blaschke, niklas boers, daniel dylewsky, carlos lópez-martínez, isobel parry, paul ritchie, bregje van der bolt, larissa van der laan, els weinans.

Tipping points characterize the situation when a system experiences abrupt, rapid, and sometimes irreversible changes in response to only a gradual change in environmental conditions. Given that such events are in most cases undesirable, numerous approaches have been proposed to identify if a system is approaching a tipping point. Such approaches have been termed early warning signals and represent a set of methods for identifying statistical changes in the underlying behaviour of a system across time or space that would be indicative of an approaching tipping point. Although the idea of early warnings for a class of tipping points is not new, in the last 2 decades, the topic has generated an enormous amount of interest, mainly theoretical. At the same time, the unprecedented amount of data originating from remote sensing systems, field measurements, surveys, and simulated data, coupled with innovative models and cutting-edge computing, has made possible the development of a multitude of tools and approaches for detecting tipping points in a variety of scientific fields. However, we miss a complete picture of where, how, and which early warnings have been used so far in real-world case studies. Here we review the literature of the last 20 years to show how the use of these indicators has spread from ecology and climate to many other disciplines. We document what metrics have been used; their success; and the field, system, and tipping points involved. We find that, despite acknowledged limitations and challenges, in the majority of the case studies we reviewed, the performance of most early warnings was positive in detecting tipping points. Overall, the generality of the approaches employed – the fact that most early warnings can in theory be observed in many dynamical systems – explains the continuous multitude and diversification in their application across scientific domains.

• Article (PDF, 3403 KB)
• Supplement (3019 KB)
• Article (3403 KB)
• Full-text XML

Dakos, V., Boulton, C. A., Buxton, J. E., Abrams, J. F., Arellano-Nava, B., Armstrong McKay, D. I., Bathiany, S., Blaschke, L., Boers, N., Dylewsky, D., López-Martínez, C., Parry, I., Ritchie, P., van der Bolt, B., van der Laan, L., Weinans, E., and Kéfi, S.: Tipping point detection and early warnings in climate, ecological, and human systems, Earth Syst. Dynam., 15, 1117–1135, https://doi.org/10.5194/esd-15-1117-2024, 2024.

Tipping points characterize a situation when a system experiences abrupt, rapid, and sometimes irreversible changes. Such shifts occur when a threshold is crossed and the system transitions from its current state to a contrasting one (van Nes et al., 2016). Given that tipping points are associated with abrupt, rapid, and sometimes irreversible changes, numerous approaches have been proposed to identify if a system is getting closer to such a point. These approaches are often referred to as early warning signals (EWSs), and they represent a powerful generic tool for anticipating tipping points in a variety of systems (Scheffer et al., 2009). The general mechanism behind EWSs is that, as a dynamical system approaches a tipping point, it becomes slower at recovering from small perturbations (Wissel, 1984), and this critical slowing down (CSD) of the system leaves signatures in the temporal and spatial dynamics of the system (Drake et al., 2020). EWSs rely on identifying exactly such changes in the underlying behaviour of a system across time or space prior to a tipping point.

Table 1 Available software tools for the estimation of early warnings with temporal and spatial datasets.

Download Print Version | Download XLSX

Early, after their introduction in the literature, it became clear that EWSs did not allow the anticipation of all types of tipping points in advance (Hastings and Wysham, 2010) and that they are not unique to tipping point responses but also occur when systems are undergoing smoother transitions (Kéfi et al., 2013). These realizations imply that some shifts (typically referred to as abrupt shifts or regime shifts) may require alternative or additional signals (Boettiger et al., 2013; Dakos et al., 2015). Thus, a rich research programme has been triggered in the theory behind tipping point anticipation and the development of tools (Table  1 ) for quantifying changes in dynamical patterns of system behaviours that could be used as early warnings preceding tipping points and abrupt shifts in general. Different terms have been used to describe the great variety of metrics proposed in the literature, like “early warning systems” (Lenton, 2013b), “observation-based early warning signals” (Boers, 2021), “statistical stability indicators” (Bathiany et al., 2016), “critical slowing down (CSD) indicators” (Tang et al., 2022), “leading indicators” (Carpenter et al., 2008), “resilience indicators” (Dakos et al., 2015), “generic indicators” (Scheffer et al., 2015), “dynamical indicators of resilience” (DiOR) (Scheffer et al., 2018), “indicators of transitions” (Clements and Ozgul, 2018), and “universal early warning signals” (Dylewsky et al., 2023). In the rest of the paper, we will use the term “early warnings” to refer to this whole family of indicators.

Whatever the term used, while early warnings are well grounded in theory, the challenge remains to apply them to real-world systems. A number of review and synthesis papers have summarized the theoretical aspects of early warnings and provided partial accounts of their empirical applications (Alberto et al., 2021; Bestelmeyer et al., 2011; Dakos and Kéfi, 2022; Lenton, 2011, 2013b; Litzow and Hunsicker, 2016; Nijp et al., 2019; Scheffer et al., 2012a, 2015). However, although the utility of early warnings has led to early warnings proliferating beyond ecology and climate and being applied across a variety of scientific domains, we miss a complete picture of where, how, and which early warnings have been used so far in real-world case studies.

Here, after summarizing the basics of the theory underlying early warnings and giving an overview of their taxonomy, we review the literature for the use of early warnings in empirical studies across all scientific fields. We document what metrics have been used; their success; and the field, system, and tipping points involved. We then classify this information in order to provide an overview of the progress, limitations, and opportunities in the empirical application of early warnings after 15 years of research on the topic.

The theory behind tipping point anticipation is mostly based on destabilizing stable fixed equilibrium points. In such cases, there are three ways that a tipping point may theoretically occur (Lenton, 2013a). A system may undergo a bifurcation when a parameter (or multiple parameters) in the system changes beyond a critical threshold and the stability of the state the system occupies is lost, thus causing the system to shift to an alternative state (bifurcation tipping or B-tipping). Noise-induced tipping can occur when a system is shifted outside its stable basin of attraction by some form of stochastic forcing (N-tipping). A third class, known as rate-induced tipping (R-tipping), occurs when a parameter rapidly changes and the system is no longer able to track its stable state (Ashwin et al., 2011). Tipping points also occur through phase transitions, a long-studied set of emergent phenomena in physics which resemble the characteristics of the B-tipping described above (Hagstrom and Levin, 2021; Sole et al., 1996).

The majority of the early warnings discussed below are primarily developed to detect cases where there is a gradual approach towards a bifurcation-tipping event causing a loss of system state stability. Rate-induced tipping could also show early warning (Ritchie and Sieber, 2016). Noise-induced tipping is likely to occur unpredictably; therefore early warnings are less expected. In a realistic scenario with constant stochasticity and conditions gradually changing, tipping is commonly a combination of a movement towards bifurcation and noise pushing the system before the bifurcation actually occurs. In such cases, noise-induced tipping becomes more likely, as it is easier for the system to leave its current basin of attraction when it is closer to the bifurcation, and this increase in the probability of tipping can be identified through particular early warnings (Sect.  2.2 ).

Although most of the theory behind early warning signals is related to saddle-node (or fold) bifurcations, other types of bifurcations have also been considered, like transcritical, pitchfork, or Hopf bifurcations (more general codimension-1 bifurcations; Kuznetsov, 1995). Such bifurcations are smooth (also called continuous) in contrast to the abrupt (i.e. discontinuous) fold bifurcations associated with tipping points, yet it has been shown that similar early warning signals can be applied for them (Boettiger et al., 2013; Kéfi et al., 2013). A full list of bifurcation types (discontinuous and continuous) and their relationship to CSD can be found in Thompson and Sieber (2011). In this paper, the early warnings considered are mostly developed in the context of the discontinuous fold bifurcation.

Table 2 A taxonomy of early warnings depending on whether or not the warning is based on critical slowing down (CSD). CSD-based early warnings are mostly associated with bifurcation tipping (B-tipping), while non-CSD-based ones are associated both with B-tipping and noise-induced tipping (N-tipping; see also Sect.  2.1 ). A second dichotomy is based on whether the warning is a statistical metric based on the dynamical patterns of the system or whether it is based on a process model that is as simple as possible. In parentheses is the type of data (temporal and/or spatial) used to estimate the early warning.

We hereafter present a representative (but not complete) overview of the early warnings mostly used, both theoretically and empirically. These signals can be classified in different ways depending, for instance, on the type of mechanism or tipping point (e.g. CSD-based, non-CSD-based), the type of data used (e.g. temporal, spatial, trait, abundance data), and the approach employed (e.g. analysing patterns, fitting models, network methods). In Table  2 , we suggest a taxonomy of early warnings based on the mechanism and the approach used. We then present their basics without going into the details. A full description and the methods to estimate them can be found elsewhere (Clements and Ozgul, 2018; Dakos et al., 2012; Génin et al., 2018; Kéfi et al., 2014; Lenton, 2011; Scheffer et al., 2015) and in dedicated software packages (Table  1 ).

2.1  Early warnings based on critical slowing down (CSD-based)

The early warnings most used are based on searching for evidence of “critical slowing down” (CSD) in the system. Essentially, as the system is forced towards a tipping point, the state it currently occupies starts to lose its stability, and the restoring feedbacks that “pull” the system back to that state after it is perturbed start to degrade. This causes the system to respond more sluggishly to these perturbations and thus slow down (Wissel, 1984). Figure  1 shows this concept visually using the “ball in the well” analogy. When the system is more stable, represented by the well with steeper sides, the recovery is faster as the ball (representing the state of the system) returns faster. A system close to tipping, represented by a shallower well, has a slower recovery as the ball takes longer to return. Eventually, the restoring feedbacks of the system may become so weak that the stability of the current state may be lost, and the system may transition to a new stable state. Mathematically, CSD occurs as the leading eigenvalue of the system approaches 0 from below. However, in reality, we do not have the equations that govern the system's dynamics, and, as such, we have to estimate the occurrence of CSD with methods that aim to infer CSD mostly from the patterns of the system dynamics or by fitting very simple and generic process-based models (Table  2 ).

Figure 1 Using the “ball in the well” analogy to compare a system that is far from tipping (left) and a system that is close to tipping (right). The system that is further away from tipping recovers faster from perturbations, with the steeper sides of the well describing the stronger restoring feedbacks of the system. Closer to tipping, the sides of the well are shallower, such that the system will take longer to return from the same perturbation because the restoring feedbacks are weaker.

Return rate, autocorrelation, and variance

Using statistical techniques makes it possible to detect CSD based on the dynamical patterns a system is generating. The most direct way to detect CSD is to consider the rate at which a system returns to its initial state following a perturbation (return rate or return time). A resilient system with strong restoring feedbacks will return to its initial state faster than one which is near to a tipping point (Wissel, 1984). However, this method requires the occurrence of a well-defined perturbation and clear knowledge of when the equilibrium state of the system has been reached, neither of which are always clearly defined in the real world.

Figure 2 A comparison of the lag-1 autocorrelation (AR(1)) for a system that is far from tipping (blue), getting close to tipping (purple), and close to tipping (red). As the time series approaches tipping (top row), there is no correlation between subsequent values of the time series in the blue part of the time series (far from tipping). However, closer to tipping, in the purple and then red regions of the time series, there are correlations and thus higher AR(1) values. In the time series itself there are clear deviations towards the end compared to the beginning, suggesting that CSD occurs as the tipping point approaches. The early warnings are calculated on a moving window (coloured regions in the bottom plot). Here, AR(1) is shown at the end of the window used to calculate it, with examples shown as coloured points to match those windows on the detrended time series.

As the system approaches a tipping point and its recovery slows down, each time step X ( t ) is more correlated to the previous time step X ( t −1) (as shown in Fig.  2 ). This can be measured with lag-1 autocorrelation, or AR(1), which tends towards 1 as a system experiences CSD prior to tipping to an alternate state. Visually, this can be viewed by observing a scatterplot of a time series of the system against the time series lagged by 1 time point (Fig.  2 ). When the system is far from tipping (top row of Fig.  2 ), there is no relationship between the current state and the state at the previous point in time (low AR(1)). As the system approaches the tipping point, CSD means that there is a strong correlation between the system state now and at the previous point in time (and thus a higher AR(1)). Larger deviations in the red section of the time series can be seen, further showing this slowing down and increase in AR(1).

Similarly, as the system struggles to return to its initial state as resilience is lost, the variance of the system is also expected to increase, as the system can sample more of the “state space” (all the possible states the system can be in) due to the shallower well. However, this is often recorded alongside an increase in AR(1) because other factors can lead to a change in variance, such as how the system is forced externally.

Spatial analogues of the temporal variance and temporal AR(1) exist, too, with a similar underlying theory to the one for the temporal ones: as a system approaches a tipping point and responds more sluggishly to external perturbations and samples more of the state space, it is expected that there will be a higher spatial autocorrelation (Dakos et al., 2010) and spatial variance (Guttal and Jayaprakash, 2009).

Just like AR(1) and variance, all other CSD-based early warnings aim at detecting characteristic changes in the dynamical patterns of the system either by directly estimating a statistical property (e.g. spectral reddening) or by fitting a statistical model (e.g. detrended fluctuation analysis) (Table  2 ). A parallel approach involves more complex methods to predict the movement towards tipping points that involve the use of simple process-based models. One such example is that of using a generalized model that integrates knowledge about the system into a model, which may allow us to estimate changes in the leading eigenvalue of the system once minimal model assumptions have been made (Lade and Gross, 2012).

2.2  Early warnings not based on critical slowing down (non-CSD-based)

CSD-based early warnings rely on the assumption that the system state shows only small deviations around the equilibrium state of the system. However, this assumption does not hold in the presence of strong stochasticity. In other cases, either CSD is hard to measure or more idiosyncratic metrics have been suggested to act as alternatives to CSD-based warnings. Below, we outline a few of the most representative non-CSD-based early warnings (Table  2 ).

2.2.1  Skewness

As the current equilibrium state of the system is losing resilience and the probability to shift to an alternative equilibrium increases, the temporal distribution of states of the system is expected to become increasingly skewed toward the alternative state. This can be quantified by the skewness of the system. The skewness may increase or decrease, depending on whether the alternative equilibrium is larger or smaller than the current equilibrium (Guttal and Jayaprakash, 2008). Similarly to the change in skewness observed between the two states with temporal data, it is also possible to observe this change in skewness in the spatial domain (Guttal and Jayaprakash, 2009).

2.2.2  Flickering

“Flickering” is the situation where strong stochasticity can “push” a system temporarily into the basin of attraction of the alternative state before returning to the current state with increasing likelihood as the system is approaching tipping (Dakos et al., 2013; Wang et al., 2012). Flickering can be measured either by a simple increase in variance (Dakos et al., 2013) or by more complex statistical approaches, e.g. quickest detection method (Carpenter et al., 2014) and heteroscedasticity (Seekell et al., 2012; Seekell and Dakos, 2015).

2.2.3  Potential analysis

Information about a system at multiple sampling points through time or multiple locations across space can allow us to reconstruct a “stability landscape” of the system – or potential, which gives an idea of the most frequent states of the systems observed in systems experiencing different environmental conditions and history (Livina et al., 2010). Multimodality in such a landscape for a given set of environmental conditions suggests that the system could exhibit alternative stable states for that range of conditions (Abis and Brovkin, 2019; Hirota et al., 2011; Scheffer et al., 2012b; Staver et al., 2011), although seasonality patterns should be accounted for to reduce misinterpretation of externally forced “states”.

2.2.4  Spatial patterns

A number of ecosystems have a clear spatial structure which is self-organized (e.g. drylands, peatlands, salt marshes, mussel beds; Rietkerk et al., 2008). Theoretical models have shown that the size and shape of the spatial patterns change in a consistent way along stress gradients, and, as such, they are good indicators of ecosystem degradation (von Hardenberg et al., 2001; Kéfi et al., 2007; Rietkerk et al., 2004). Probably one of the most studied examples is the case of dryland ecosystems, where changes in the shape of regular patterns (Rietkerk et al., 2004) and in the patch size distribution (Kéfi et al., 2007) could inform us about the stress experienced by the ecosystem (Dakos et al., 2011).

2.2.5  Fitting a threshold model

An alternative approach to pattern-based early warnings is based on fitting process-based models on the time series of a system prior to a tipping point. This approach mainly consists of fitting the simplest dynamical model with a tipping point (i.e. a saddle-node normal form) (Ditlevsen and Ditlevsen, 2023) and testing its likelihood compared to a model without a tipping point (Boettiger and Hastings, 2012b) or of fitting threshold models assuming simple autoregressive state-space models (Ives and Dakos, 2012; Laitinen et al., 2021).

2.2.6  Structural changes

A novel way to detect tipping points involves monitoring structural change properties (e.g. connectivity, node centrality) in network systems (i.e. a network of interacting components) like spatially connected sites, interacting actors, or species in a community (Cavaliere et al., 2016; Mayfield et al., 2020; Yin et al., 2016). Alternatively, the temporal correlation between components in multivariate systems has been used to construct an interaction network and analyse its structural properties (Tirabassi et al., 2014).

2.2.7  Trait changes

Another idiosyncratic approach involves monitoring changes in the statistical moments of fitness-related traits (e.g. body size) (Clements and Ozgul, 2018). Such trait changes have been found in populations under stress where there are changes in the traits of individuals (i.e. decreasing mean and increasing variance in body size) (Clements and Ozgul, 2016; Spanbauer et al., 2016). These trait-based and the above-mentioned structural-based signals are case-specific and idiosyncratic to the details of the system, as there is no universal mechanism that would generate an expected pattern related to the approach of tipping points.

We performed a (non-exhaustive) literature review on the empirical (not theoretical) use of warning signals. We first did a topic search (TS) that included title, abstract, and keywords in the Web of Science for the period from 1 January 2004 to 1 April 2023 with the following terms. TS  =  ((“tipping point ∗ ” or “tipping” or “catastrophic bifurcation ∗ ” or “catastrophic shift ∗ ” or “regime shift ∗ ” or “abrupt shift ∗ ” or “critical transition ∗ ”) and (“early warning ∗ ” or “early warning ∗ ” or “warning sign ∗ ” OR “resilience indicator ∗ ” or “leading indicator ∗ ” OR “precursor ∗ ”)). We selected the year 2004 as the starting date of our search, despite the fact that CSD was known in ecology much earlier (Wissel, 1984) and that signatures of catastrophic bifurcations were theoretically described for dynamical systems (Gilmore, 1981). Our choice was driven by the fact that 2004 is around the time of the first studies in climate (Held and Kleinen, 2004; Kleinen et al., 2003) and ecology (Carpenter and Brock, 2006) where the theoretical idea of using CSD as a warning signal emerged, while a few years later the first review on early warnings on critical transitions was published (Scheffer et al., 2009). Within this time period, our topic search returned 887 unique publications. For completeness, we also ran the same topic search before 2004 going back to 1960, and we retrieved 11 publications of which only 1 was related to bifurcations. Clearly, we might have missed relevant records with the TS we selected. For example, had we also included the term “phase transition ∗ ”, we would have retrieved 3916 records. We decided not to include this term, as it pertains to a specific and rich field of physics, but with our TS we are confident to have a rather complete overview of the tipping point (and related terms) literature.

We screened all 887 publications to select only the ones where there was an empirical application of early warnings (i.e. an indicator was measured on real data to signal the occurrence of a tipping point). This screening led to 229 papers that we classified as ones that included at least one empirical application of early warnings. For each paper, we collected the following information: “domain” (e.g. climate, ecology), “system” (e.g. Arctic sea ice, fisheries, mental depression), the “tipping point” described, data source (e.g. lab experiment, field survey, remote-sensed datasets, social data), “data type” (i.e. temporal, spatial, spatiotemporal), “indicator” (i.e. the specific warning signal(s) used), and “performance” (whether the performance of the early warning was reported in the paper as positive, negative, mixed (in the case of multiple signals used or multiple datasets analysed), or inconclusive). To facilitate the analysis, we regrouped the “data source” and “indicator” categories into broader groups (see Sect. S1 in the Supplement). We also created two extra categories: we classified systems under a specific “field”, and we introduced an “indicator type” based on whether the early warning was CSD-based or non-CSD based. We then excluded the running year 2023 and summarized results in terms of unique publications using simple statistics and alluvial plots in R (4.3.1).

Figure 3 Evolution of studies applying early warnings in empirical datasets. The total 229 papers we identified through our literature review between 2004 and 2022 were classified within five main scientific domains (ecology, health, climate, social sciences, and physical sciences). The dotted white line shows the cumulative number of papers.

3.1  Overall use of early warnings across disciplines

We were able to classify the total 229 papers published from 2004 to 2023 into five main domains: ecology, climate, health, social sciences, and physical sciences. We found that empirical papers first appeared in 2007 in the domains of ecology and climate but that the first papers in health, social sciences, and physical sciences were only published after 2010 and 2011 (Fig.  3 ). This change may be associated with the highly cited review by Scheffer et al. (2009) that introduced (and popularized) the terms “early warning signals” and “critical transitions”. Since then, the number of empirical studies has quickly increased but has remained dominated by ecology (43.6 % of the papers overall), followed by health (22.6 %), climate (14.6 %), social sciences (12 %), and physical sciences (7.6 %) – showing the diversification of the uses of early warning (Fig.  3 ).

The higher number of publications in the health domain compared to the climate domain is unexpected. We found a large number of studies in the medical field (Fig. S1 in the Supplement) that form a distinct group on the emergence of human diseases, such as cancer (Liu et al., 2020), which use a non-CSD context-specific early warning (“dynamic_network_biomarkers”; see also Sect.  3.3 ). Zooming into within each domain, we observed that the most ecological studies are on terrestrial and freshwater fields (Fig. S1), namely on drylands and in forest and lake ecosystems (Table S1 in the Supplement). The majority of climate studies are on past climate transitions and modern records (Fig. S1, Table S2), while the social studies are split between societal shifts (like in politics, social behaviour, and transport) and finance transitions (Fig. S1, Table S4). Lastly, studies on physical sciences appear more heterogeneous, including tipping points in materials, power systems, and even astronomy (Fig. S1, Table S5).

Figure 4 Alluvial plot connecting scientific domains, data sources, and data types. Colours indicate the data source used for the estimation of early warnings. The size of the boxes in each column represents the proportion of each category. The figure is “read” from the middle column ( b , “data source”) to either the right ( c , “data type”) or the left ( a , “scientific domain”). The thickness of the lines is proportional to the studies of a given data source that belong to a certain domain (from the “data source” column,  b , to the “domain” column,  a ) or is proportional to the specific data type used (from the “data source” column,  b , to the “data type” column,  c ). For example, for the “data source” field experiment (light green), all studies using field experiments belong to the “ecology” domain, while field experiments are split into three types of data (spatial, spatiotemporal, and temporal). field_exp: field experiment; lab_exp: lab experiment; palaeo: palaeo-reconstructed data; remote-sensed: data through remote sensing; social_data: financial data and from social media; survey: data from surveys (field, lab, social).

3.2  Multiple sources of data used

Across scientific domains, the vast majority of early warnings were analysed on temporal data (77.7 %), while the spatial data were used in only 8 % of all studies (Fig.  4 ) only pertaining to ecology (Fig. S2). Survey data made up the majority of the data sources (43.8 %; including field surveys, social survey data, data from weather stations or other monitoring devices, medical data from hospitalization records to electroencephalograms (EEGs)), followed by data from lab experiments (20.7 %), remote sensing (12 %), palaeo-reconstructions (10 %), and field experiments (7 %). This partitioning can mostly be explained by our classification, meaning that we have grouped together a heterogeneity of data sources (e.g. field surveys, historical climate data, social study surveys, hospitalization records; Sect. S1). However, it also reflects the availability of each data source (e.g. most survey and palaeo data were readily available and reanalysed in the context of tipping points) or the difficulty in their acquisition (e.g. field experiments are harder to execute compared to lab experiments). Looking at how data sources are used across domains, ecology is the only domain where all kinds of data sources have been used. What is also interesting to note is that two sources of data are increasingly used: survey data and remote-sensed data (Fig. S3). Specifically, the latter were the latest to be used (2011) but show a consistent rising pattern over the last years, mainly due to the fact that satellite products span a long enough time period by now ( ∼  20 years) to allow the estimation of early warnings.

A closer look at studies using remote-sensed products reveals a focus on the analysis of temporal early warning in land environments, mainly forests (e.g Boulton et al., 2022; Majumder et al., 2019; Saatchi et al., 2021) and drylands (e.g Veldhuis et al., 2022; Uden et al., 2019; Wu et al., 2023), but this also extends to the cryosphere, focusing on the analysis of the Arctic and the Antarctic ice sheets (AlMomani and Bollt, 2021; Carstensen and Weydmann, 2012). The spatial resolution of remote-sensed data has also been exploited for the identification of spatial early warning, especially regarding desertification (Berdugo et al., 2017) and vegetation analyses (Majumder et al., 2019). Overall, we found that the use of remote sensing products offers two distinct yet complementary approaches to detect early warning: high-level products, which correspond to physical variables, for instance sea surface temperature (SST) (Wu et al., 2015), or different types of indices like the normalized difference vegetation index (NDVI) (Liu et al., 2019) and low-level products or direct sensor observables.

Figure 5 Number of papers of the 21 early warnings used more than once in our literature review. Each bar is partitioned into the five scientific domains.

3.3  A growing list of early warnings

We recorded 65 different early warnings after reclassifying some into the same group (for example, variance, coefficient of variation, and standard deviation were reclassified as “variance”; Sect. S1). As expected, the majority were CSD-based warnings (74.9 %), while 25.1 % were non-CSD-based ones. Out of the 65 reclassified early warnings, only 21 were used more than once (the remaining 44 early warnings were used only once; Figs.  5 , S5). Variance and autocorrelation were the dominantly used early warnings across all domains, followed by skewness (Fig.  5 ). Besides these 3 early warnings, the remaining 18 were used selectively within particular domains. The most striking are “spatial variance” (only used in ecological studies) and “dynamic network biomarkers” (only used in health studies; see also Sect.  3.1 ). Within domains (Fig. S4), ecology is the domain with the highest heterogeneity in the early warnings (18 out of the 21 used more than once), followed by health (10) and climate (7).

Figure 6 Alluvial plot connecting scientific domains, the performance of the early warnings, and the type of early warning (CSD-based vs non-CSD-based). Colours indicate the performance. The size of the boxes in each column represents the proportion of each category. The figure is “read” from the middle column (“performance”,  b ) to either the right (“early warning type”,  c ) or the left (“scientific domain”,  a ). The thickness of the lines is proportional to the performance that belongs to a certain domain (from the “performance” column,  b , to the “domain” column,  a ) or is proportional to the type of early warning (from the “performance” column,  b , to the “early warning type” column,  c ). For example, for the “performance” mixed (blue), studies with mixed performance were done with both CSD-based and non-CSD-based warnings (“early warning type” column,  c ), while the CSD-based mixed were found in all domains and the non-CSD-based were split among climate, ecology, and social sciences (“scientific domain” column,  a ). “Positive” performance indicates there was a warning identified, “negative” indicates no warning was identified, and “mixed” indicates positive and negative performances when tested in multiple datasets or when testing more than one early warning in the same dataset. For “inconclusive”, the results could indicate neither a positive nor a negative warning.

Figure 7 Early warnings (used  >  1) and their performance. “Positive” performance indicates there was a warning identified, “negative” indicates no warning was identified, and “mixed” indicates positive and negative performances when tested in multiple datasets or when testing more than one early warning in the same dataset. For “inconclusive”, the results could indicate neither a positive nor a negative warning.

3.4  A positively skewed performance of early warnings

Of all studies, 67.8 % reported a positive performance of all 65 reclassified early warnings across all domains (Fig.  6 ). Only 3.4 % of the studies reported negative performances (i.e. no expected warning or opposite to expected warning). Studies in ecology reported the most negative results, followed by climate and health, with none reported for physical or social sciences. The performance in the rest of the studies was either mixed (i.e. positive or negative in studies which analysed multiple early warnings or datasets; 24.7 %) or inconclusive (i.e. a statistically weak result; 4.8 %). This is an impressively positively skewed result potentially reflecting the known bias in publishing significant results (Fanelli, 2012) or in post hoc analysis where a tipping point has already been documented and early warnings have been applied in hindsight (Boettiger and Hastings, 2012a). Interestingly, all the negative results included CSD-based warnings, while for non-CSD warnings only a small fraction reported inconclusive or mixed results (Fig.  6 ). This difference could be attributed to the fact that non-CSD warnings are at times idiosyncratically developed for the specific system under study compared to the more generic CSD-based warnings. Indeed, focusing on the 21 early warnings that were used more than once (Sect.  3.3 ), such system-specific indicators (such as the “dynamical_network_biomarkers”) always had a positive result (Fig.  7 ). Overall, the least-used warnings were associated with a positive performance, whereas the most-used ones (like variance, autocorrelation, skewness, and power spectrum) showed all types of performance. There was no early warning that had predominantly negative or mixed results except for kurtosis (Fig.  7 ). There was no particular difference in the performance across domains of the early warnings used more than once (Fig. S5).

The idea of early warnings based on CSD is relatively old. In the book Catastrophe Theory for Scientists and Engineers (1981), Gilmore already talked about “catastrophe flags” for indicators of CSD (Gilmore, 1981). After an early ecological paper (Wissel, 1984), the topic generated an enormous amount of interest, mainly theoretical, 20 years later, with the first empirical tests being on past climate tipping points (Dakos et al., 2008; Livina and Lenton, 2007). Our review of the literature of the last 2 decades shows how the use of these indicators has since spread to many other disciplines. Indeed, the generality of the approach – the fact that CSD can be observed in many dynamical systems, independent of the details of the underlying dynamical equations – created an opportunity for testing their validity on many systems and explains the enthusiasm they generated and the diversification of applications which followed.

4.1  Early warning applications: a success story?

Our literature overview suggests that the 65 early warning signals identified successfully detected a tipping point almost 70 % of the times they were used. This is an impressively positive result which should nonetheless be treated with caution. Firstly, in many cases, the empirical studies were conducted on systems that are either relatively simple or under controlled lab conditions, making them mostly proof-of-principle demonstrations (Dai et al., 2012; Veraart et al., 2012). Secondly, most empirical studies were restricted to hindsight application, meaning that an a priori knowledge of a tipping point may introduce bias towards detecting CSD indicators (Boettiger and Hastings, 2012a; Spears et al., 2017). Thirdly, the documented publication bias against negative or insignificant results (Fanelli, 2012; Franco et al., 2014) probably applies in the case of the early warning research given the attention this specific topic has attracted in recent years. One important aspect that we have not considered in the comparative analysis of the reviewed literature is the fact that each paper uses different statistical methods, different hypothesis-testing approaches (like surrogate data and Bayesian and frequentist p-values), and different significance levels to conclude on the identification of an early warning or not. To what extent such differences may even induce p-hacking is unclear, but it needs to be acknowledged in future work.

However, these considerations do not reduce the value and prospect of early warning research. For instance, one of their biggest values lies in the possible detection of an approaching tipping point. A number of studies have demonstrated the potential proximity of tipping points in modern climate data using early warnings (Boers, 2021; Boulton and Lenton, 2015; Ditlevsen and Ditlevsen, 2023). It has also become increasingly clear that early warnings can be very useful for comparing the resilience of similar ecosystems across space (e.g. (Forzieri et al., 2022; Lenton et al., 2022; Verbesselt et al., 2016) to provide an approximate estimate of resilience that can help prioritize management. In this way, with “resilience maps”, rather than speculating about the proximity to a potential threshold, we can rank situations at a given moment and place.

4.2  Challenges

There are a number of conceptual, operational, and methodological challenges that can blur the detection of early warnings on real data (Dakos et al., 2015). For these reasons, some studies have highlighted their failure at detecting early warnings on data (Burthe et al., 2016), while others have raised caution about their uninformed use without knowing more about the system's drivers and underlying mechanisms (Boettiger et al., 2013). Our review can be used as a starting point in trying to understand when and why early warnings can fail or not by looking at how they have been applied within the domains we have identified, as their performance seems to be idiosyncratic to the data type and case study used. In what follows, we discuss more generally some of the most important challenges related to the use of early warnings.

4.2.1  Fast changes, slow responses, stochasticity, multiple drivers, and limited data challenge early warning performance

The detection of early warnings relies on the assumption that the system is approaching a transition gradually. A system should be externally forced on a slow timescale towards the tipping point while experiencing perturbations on a shorter timescale such that CSD-based early warning signals can be estimated. In theory, it is generally assumed that the short-term noise is independent and identically distributed with a mean of zero. This is unlikely to be the case with climate systems experiencing extreme weather events, for example, which are likely becoming more prevalent with the changing climate. There have been three “1-in-100-year” droughts in the Amazon rainforest since 2005 (Erfanian et al., 2017; Lewis et al., 2011) which clearly alter the signals observed. For cases like these, it is worth measuring early warnings on the drivers themselves. If these show early warnings, then it is likely that signals observed in the system itself are being driven by changes in forcing rather than by a gradual movement towards tipping. However, early warnings of the drivers as a false-positive check make sense only in the case where the drivers are independent from the system variable. For instance, in the case of the Amazon, early warnings of rainfall can be seen as indicators of the Amazon tipping itself because of the strong moisture recycling feedback present rather than an external factor inducing early warnings on Amazon vegetation dynamics.

Things get even more complicated when more than one driver is acting on the system. In most cases, the assumption is that there is a single driver with a monotonic directional change towards the tipping point or that there are multiple drivers which all have the same effect and directionality. However, it has been shown experimentally that, in the presence of multiple drivers, contradictory early warnings may arise even if both drivers would produce similar patterns in early warnings acting in isolation (Dai et al., 2015).

When monitoring a system, longer time series are desirable to detect the upcoming tipping point. For instance, the best-case studies found in this literature review from remote-sensed products, which have been available since ∼  1972, have approximately 50-year-long time series. However, due to sensor degradation and upgrades, it can be challenging to get a long time series from a single sensor, and products are often created from combined data sources. This can interfere with most of the early warnings if this merging changes the signal-to-noise ratio (SNR) across time (Smith et al., 2023). For example, newer sensors will measure with a greater radiometric accuracy, increasing the SNR and in turn “erroneously” increasing the AR(1) as far as an early warning is concerned. This increase in SNR will also decrease variance, thus allowing the user to check for anticorrelation between AR(1) and variance to see if the early warnings are being influenced or not.

As well as questions around data availability and noise behaviour, the inherent timescale of the system being studied can hinder our ability to detect tipping points. While tipping is by definition a “fast” process, for slow-moving systems like the thermohaline circulation (AMOC), this tipping event occurs over decades; therefore, it could be difficult to detect that the tipping point has been passed using early warnings. Another example of this is the Amazon rainforest, where, at least in modelled vegetation, there is a slow response of the forest based on the climate change that it has been subjected to (Jones et al., 2009). It could take decades for dieback to occur even under a constant climate such that a tipping point could be passed long before it is actually realized (Hughes et al., 2013). This “committed response” has been explored in a number of GCM experiments (Boulton et al., 2017; Jones et al., 2009), but it is unclear how early warnings would be affected by this (van Der Bolt et al., 2021).

4.2.2  Non-specificity of early warnings

The generic and universal character of most (but in particular CSD-based) early warnings comes at a price of these warnings not being specific to abrupt and irreversible tipping points. Instead they can also be used to detect smooth and reversible transitions (Kéfi et al., 2013). This limitation suggests that we need additional indicators, in particular system-specific indicators (Boettiger et al., 2013). In the case of spatially structured ecosystems such as drylands, studies have shown that temporal early warnings could fail (Dakos et al., 2011), in which case the use of the changes in the patterns themselves could provide a good alternative (Kéfi et al., 2007; Rietkerk et al., 2004). In the same way, specific indicators have been developed in health sciences for the monitoring of disease emergence (Table S3).

System-specific early warnings may also be a better prospect, where understanding processes in the system can help us to monitor its resilience in novel ways (Boulton et al., 2013). However, the original idea behind the development of early warnings was based on the premise that this knowledge is missing or insufficient; thus a pattern-based approach could be more informative (Scheffer et al., 2009). Therefore, the challenge is to strike the right level of system-specific warnings and to combine them with the generic ones. For instance, trait-based (Clements and Ozgul, 2016) and function-based (Hu et al., 2022) warnings have been recently suggested as complementary to the existing generic warning signals. A first step towards that direction could be to map the 65 classified early warnings we reviewed on a gradient of generic to system-specific indicators.

4.2.3  Multivariate (high-dimensional) systems

Most early warnings are well-tailored for unidimensional systems, meaning systems described by a single observable (e.g. vegetation cover). However, real dynamical systems are typically high-dimensional, and the quantification of early warnings in those multivariate systems presents challenges. For instance, two different variables may give conflicting information or obscure a clear signal (Boerlijst et al., 2013; Weinans et al., 2021). In theory, one expects that the variables directly involved in the interactions to cause a tipping point are the best to monitor (Carpenter et al., 2014). However, it is challenging to know from which variable(s) to measure early warnings in a multivariate system (Dakos, 2018).

Two main approaches in the analysis of multivariate systems have recently been developed. One relies on conceiving the system as a network, where the nodes are the variables and whose dynamics are followed through time, and evaluating changes in the structure of the network. For instance, as the system moves towards a tipping point, changes in degree distributions of such a network are representative for an approaching tipping point (Lu et al., 2021). Recent research explores a complementary approach where causal links are calculated instead of correlation links and where the strength of the causal link works as the indicator of resilience (Nowack et al., 2020).

Alternatively, dimension-reduction techniques can capture overall network dynamics into a representative statistic. For instance, principal component analysis (PCA; often referred to as empirical orthogonal function (EOF) in climate science) can be used to get directions of change (Held and Kleinen, 2004; Weinans et al., 2019). Data can be projected onto the leading principal component, effectively yielding a univariate time series on which the univariate early warning can be applied (Bathiany et al., 2013; Boulton and Lenton, 2015; Held and Kleinen, 2004). This analysis does not make any a priori assumptions about the interactions between the different network nodes and is therefore quite flexible in its use. However, it requires large amounts of high-quality data to yield accurate results. The underlying assumption is that, as the system approaches the tipping point, the dynamics become more correlated, leading to a high explained variance of a PCA and clear directionality in the dynamics (Lever et al., 2020).

4.2.4  Tipping cascades

A more peculiar challenge in the application of early warnings is their ability to detect cascading tipping points, where a tipping point in one system has a knock-on effect on another system, causing that to also tip (Klose et al., 2020; van de Leemput et al., 2018; Saade et al., 2023). Unless these systems are linked in such a way that early warnings can be observed in both systems, the cascade is likely to present as a shock to the second system such that it would be unpredictable whilst monitoring it in isolation (Bathiany et al., 2013). For systems where tipping in one system causes the connected system partially towards a tipping point (known as a “two-phase transition”), a stepwise jump in early warnings in the second system can be detected. For coupled systems where the tipping in the second system happens instantaneously (a “joint cascade”) or soon after the tipping in the first system (a “domino cascade”), early warnings are unlikely to be detectable (Klose et al., 2021).

4.3  Opportunities

These challenges associated with the use of early warnings are also accompanied by a number of opportunities to improve their detection in real data. Below we outline a few of the most promising ones.

4.3.1  Composite metrics

Although there exists a multitude of early warnings (CSD-based and non-CSD-based; generic and system-specific; and on spatial, structural, and temporal data), few studies have compared in a systematic way how these warnings behave one against the other or across different systems (Dakos et al., 2011; Veldhuis et al., 2022). Apart from the CSD-based warnings where their relationships are mathematically known (Kuehn, 2012), we simply do not know what similar information early warnings provide. Understanding the interrelationships between all types of the early warnings most used will be crucial to improve their use for detecting tipping points. Composite metrics – where multiple early warnings are combined (Drake and Griffen, 2010), abundance-based with trait-based warnings are compared (Clements and Ozgul, 2016), or machine learning has been used to train models of multiple warnings as predictors (Brett and Rohani, 2020) – have been suggested to improve the significance and detectability of approaching tipping points. Given the increasing capacity to monitor the multivariate aspects of most systems (discussed in Sect.  4.2.3 ) and the increasing availability of such data (see Sect.  4.3.2 ), we are not far from estimating multiple early warnings on multiple dimensions of a system. The next step is to develop meaningful ways to best combine them for detecting tipping points.

4.3.2  Increasing data availability: open databases and remote-sensed data

Over the last decade, data from long-term databases and remote sensing have grown to become the primary sources for capturing temporal and spatial early warnings for tipping points. Especially for remote sensing data, this coincides with the expansion of freely available Earth observation datasets combined with access to cloud-based systems which provide the computational power to process this increase in data (Gorelick et al., 2017). A primary focus has been on the temporal analysis of optical imagery from satellites such as the Moderate Resolution Imaging Spectroradiometer (MODIS) (Liu et al., 2019; Majumder et al., 2019) or from the AVHRR (Lenton et al., 2022). Additionally, the vegetation optical depth (VOD) derived from microwave passive radiometers (Moesinger et al., 2020) has been employed to analyse early warnings, with temporal records since the late 1970s (Boulton et al., 2022; Smith et al., 2023). Overall, the continued growth of remotely sensed datasets is likely to drive further temporal early warning research, while the emergence of new satellite sensors with enhanced spatial resolutions (in the order of metres) will also enable an improved analysis of spatial early warning at large scales. However, such development requires a profound understanding of the acquisition systems to effectively control and account for parameters that may impact the extraction of early warnings.

4.3.3  New approaches: machine learning

The success of neural networks for time series classification problems has inspired the development of machine learning (ML) techniques for early warning detection. There is a natural synergy to this approach in that the same CSD phenomena manifest across a wide range of systems approaching tipping points, so the notoriously data-intensive task of training a neural network can be accomplished using plentiful synthetic data and still produce a result which can plausibly be applied to empirical data.

Deep learning models which combine convolutional layers have been shown to outperform methods using statistical CSD-based warnings (e.g. variance, AR(1)) in a variety of both real and simulated case studies (Bury et al., 2021; Deb et al., 2022). Furthermore, these models have exhibited success in inferring the type of upcoming bifurcation from observed pre-transition dynamics and have performed well in extensions to phase transitions on spatiotemporal lattices (Dylewsky et al., 2023). Other ML techniques can also tell us something about how far systems are from tipping. For example, random forest models could be used to determine the factors that determine autocorrelation in forest areas on a global scale and thus how close to tipping these forest areas could be based on driving variables (Forzieri et al., 2022), or they can help us determine the factors the influence the occurrence of tipping points (Berdugo et al., 2022). However, one should always bear in mind that ML will be as good as their training sets. Testing these approaches on existing datasets will help understand their potential along with testing them in cases of noise- or rate-tipping. Taking into consideration their limitations (Lapeyrolerie and Boettiger, 2022), combining ML techniques with “traditional” early warnings could become promising for monitoring systems that may be approaching tipping points.

The unprecedented amount of data originating from remote sensing systems, field measurements, surveys, and simulated data, coupled with innovative models and cutting-edge computing, has made possible the development of a multitude of tools and approaches for detecting tipping points in a variety of scientific fields. Early warnings can tell us that “something” important may be about to happen, but they do not tell us what precisely that “something” may be and when exactly it will happen (Dakos et al., 2015). The next step is to test the real potential of early warnings as preventive and management tools in anticipating natural and human-induced changes to come.

Literature review data can be found in Sect. S3 in the Supplement.

The supplement related to this article is available online at:  https://doi.org/10.5194/esd-15-1117-2024-supplement .

VD, CAB, JEB, and SK developed research and drafted the paper. VD led the literature review analysed by VD, CAB, JEB, JFA, DIAM, DD, CLM, BvdB, LvdL, EW, and SK. VD and SK ran the analyses. All authors commented on the final text.

The contact author has declared that none of the authors has any competing interests.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors.

This article is part of the special issue “Tipping points in the Anthropocene”. It is a result of the “Tipping Points: From Climate Crisis to Positive Transformation” international conference hosted by the Global Systems Institute (GSI) and University of Exeter (12–14 September 2022), as well as the associated creation of a Tipping Points Research Alliance by GSI and the Potsdam Institute for Climate Research, Exeter, Great Britain, 12–14 September 2022.

Chris A. Boulton, Joshua E. Buxton, and David I. Armstrong McKay were supported by the Bezos Earth Fund via the Global Tipping Points Report project. Carlos López-Martínez was supported by INTERACT project PID2020-114623RB-C32 funded by MCIN/AEI/10.13039/501100011033.

This paper was edited by Jonathan Donges and reviewed by Juan Rocha and two anonymous referees.

Abis, B. and Brovkin, V.: Alternative tree-cover states of the boreal ecosystem: A conceptual model, Global Ecol. Biogeogr., 28, 612–627, https://doi.org/10.1111/geb.12880 , 2019. 

Alberto, J., Stelzer, A., Padric, J., Adrian, R., and Willem, B.: Early warning signals of regime shifts for aquatic systems: Can experiments help to bridge the gap between theory and real-world application?, Ecol. Complex., 47, 100944, https://doi.org/10.1016/j.ecocom.2021.100944 , 2021. 

AlMomani, A. A. and Bollt, E.: An early warning sign of critical transition in the Antarctic ice sheet – a data-driven tool for a spatiotemporal tipping point, Nonlin. Processes Geophys., 28, 153–166, https://doi.org/10.5194/npg-28-153-2021 , 2021. 

Ashwin, P., Wieczorek, S., Vitolo, R., and Cox, P.: Tipping points in open systems: bifurcation, noise-induced and rate-dependent examples in the climate system, Philos. T. R. Soc. A, 13, 20, https://doi.org/10.1098/rsta.2011.0306 , 2011. 

Bathiany, S., Claussen, M., and Fraedrich, K.: Detecting hotspots of atmosphere–vegetation interaction via slowing down – Part 1: A stochastic approach, Earth Syst. Dynam., 4, 63–78, https://doi.org/10.5194/esd-4-63-2013 , 2013. 

Bathiany, S., van der Bolt, B., Williamson, M. S., Lenton, T. M., Scheffer, M., van Nes, E. H., and Notz, D.: Statistical indicators of Arctic sea-ice stability – prospects and limitations, The Cryosphere, 10, 1631–1645, https://doi.org/10.5194/tc-10-1631-2016 , 2016. 

Berdugo, M., Kéfi, S., Soliveres, S., and Maestre, F. T.: Plant spatial patterns identify alternative ecosystem multifunctionality states in global drylands, Nat. Ecol. Evol., 1, 0003, https://doi.org/10.1038/s41559-016-0003 , 2017. 

Berdugo, M., Gaitán, J. J., Delgado-Baquerizo, M., Crowther, T. W., and Dakos, V.: Prevalence and drivers of abrupt vegetation shifts in global drylands, P. Natl. Acad. Sci. USA, 119, 1–10, https://doi.org/10.1073/pnas.2123393119 , 2022. 

Bestelmeyer, B. T., A. M. Ellison, Fraser, W. R., Gorman, K. B., Holbrook, S. J., Laney, C. M., Ohman, M. D., Peters, D. P. C., Pillsbury, F. C., Rassweiler, A., Schmitt, R. J., and Sharm, S.: Analysis of abrupt transitions in ecological systems, Ecosphere, 2, 129, https://doi.org/10.1890/ES11-00216.1 , 2011. 

Boerlijst, M. C., Oudman, T., and de Roos, A. M.: Catastrophic Collapse Can Occur without Early Warning: Examples of Silent Catastrophes in Structured Ecological Models, PLOS ONE, 8, e62033, https://doi.org/10.1371/journal.pone.0062033 , 2013. 

Boers, N.: Observation-based early-warning signals for a collapse of the Atlantic Meridional Overturning Circulation, Nat. Clim. Change, 11, 680–688, https://doi.org/10.1038/s41558-021-01097-4 , 2021. 

Boettiger, C. and Hastings, A.: Early warning signals and the prosecutor's fallacy, P. R. Soc. B, 279, 4734–4739, https://doi.org/10.1098/rspb.2012.2085 , 2012a. 

Boettiger, C. and Hastings, A.: Quantifying limits to detection of early warning for critical transitions, J. R. Soc. Interface, 9, 2527–39, https://doi.org/10.1098/rsif.2012.0125 , 2012b. 

Boettiger, C., Ross, N., and Hastings, A.: Early warning signals: the charted and uncharted territories, Theor. Ecol., 6, 255–264, https://doi.org/10.1007/s12080-013-0192-6 , 2013. 

Boulton, C. A. and Lenton, T. M.: Slowing down of North Pacific climate variability and its implications for abrupt ecosystem change, P. Natl. Acad. Sci. USA, 112, 11496–11501, https://doi.org/10.1073/pnas.1501781112 , 2015. 

Boulton, C. a., Good, P., and Lenton, T. M.: Early warning signals of simulated Amazon rainforest dieback, Theor. Ecol., 6, 373–384, https://doi.org/10.1007/s12080-013-0191-7 , 2013. 

Boulton, C. A., Booth, B. B. B., and Good, P.: Exploring uncertainty of Amazon dieback in a perturbed parameter Earth system ensemble, Glob. Change Biol., 23, 5032–5044, https://doi.org/10.1111/gcb.13733 , 2017. 

Boulton, C. A., Lenton, T. M., and Boers, N.: Pronounced loss of Amazon rainforest resilience since the early 2000s, Nat. Clim. Change, 12, 271–278, https://doi.org/10.1038/s41558-022-01287-8 , 2022. 

Brett, T. S. and Rohani, P.: Dynamical footprints enable detection of disease emergence, PLOS Biol., 18, e3000697, https://doi.org/10.1371/journal.pbio.3000697 , 2020. 

Burthe, S. J., Henrys, P. A., Mackay, E. B., Spears, B. M., Campbell, R., Carvalho, L., Dudley, B., Gunn, I. D. M., Johns, D. G., Maberly, S. C., May, L., Newell, M. A., Wanless, S., Winfield, I. J., Thackeray, S. J., and Daunt, F.: Do early warning indicators consistently predict nonlinear change in long-term ecological data?, 53, 666–676, https://doi.org/10.1111/1365-2664.12519 , 2016. 

Bury, T. M.: ewstools: A Python package for early warning signals of bifurcations in time series data, J. Open Source Softw., 8, 8–11, https://doi.org/10.21105/joss.05038 , 2023. 

Bury, T. M., Sujith, R. I., Pavithran, I., Scheffer, M., Lenton, T. M., Anand, M., and Bauch, C. T.: Deep learning for early warning signals of regime shifts, P. Natl. Acad. Sci. USA, 450, e2106140118, https://doi.org/10.1073/pnas.2106140118 , 2021. 

Carpenter, S. R. and Brock, W. A.: Rising variance: a leading indicator of ecological transition, Ecol. Lett., 9, 311–318, https://doi.org/10.1111/j.1461-0248.2005.00877.x , 2006. 

Carpenter, S. R., Brock, W. A., Cole, J. J., Kitchell, J. F., and Pace, M. L.: Leading indicators of trophic cascades, Ecol. Lett., 11, 128–38, https://doi.org/10.1111/j.1461-0248.2007.01131.x , 2008. 

Carpenter, S. R., Brock, W. A., Cole, J. J., and Pace, M. L.: A new approach for rapid detection of nearby thresholds in ecosystem time series, Oikos, 123, 290–297, https://doi.org/10.1111/j.1600-0706.2013.00539.x , 2014. 

Carstensen, J. and Weydmann, A.: Tipping Points in the Arctic: Eyeballing or Statistical Significance?, AMBIO, 41, 34–43, https://doi.org/10.1007/s13280-011-0223-8 , 2012. 

Cavaliere, M., Yang, G., Danos, V., and Dakos, V.: Detecting the Collapse of Cooperation in Evolving Networks, Sci. Rep.-UK, 6, 30845, https://doi.org/10.1038/srep30845 , 2016. 

Clements, C. F. and Ozgul, A.: Including trait-based early warning signals helps predict population collapse, Nat. Commun., 7, 1–8, https://doi.org/10.1038/ncomms10984 , 2016. 

Clements, C. F. and Ozgul, A.: Indicators of transitions in biological systems, Ecol. Lett., 21, 905–919, https://doi.org/10.1111/ele.12948 , 2018. 

Dai, L., Vorselen, D., Korolev, K. S., and Gore, J.: Generic Indicators for Loss of Resilience Before a Tipping Point Leading to Population Collapse, Science, 336, 1175–1177, https://doi.org/10.1126/science.1219805 , 2012. 

Dai, L., Korolev, K. S., and Gore, J.: Relation between stability and resilience determines the performance of early warning signa ls under different environmental drivers, P. Natl. Acad. Sci. USA, 112, 201418415, https://doi.org/10.1073/pnas.1418415112 , 2015. 

Dakos, V.: Identifying best-indicator species for abrupt transitions in multispecies communities, Ecol. Indic., 94, 494–502, https://doi.org/10.1016/j.ecolind.2017.10.024 , 2018. 

Dakos, V. and Kéfi, S.: Ecological resilience: what to measure and how, Environ. Res. Lett., 17, 043003, https://doi.org/10.1088/1748-9326/ac5767 , 2022. 

Dakos, V., Scheffer, M., van Nes, E. H., Brovkin, V., Petoukhov, V., and Held, H.: Slowing down as an early warning signal for abrupt climate change., P. Natl. Acad. Sci. USA, 105, 14308–12, https://doi.org/10.1073/pnas.0802430105 , 2008. 

Dakos, V., van Nes, E. H., Donangelo, R., Fort, H., and Scheffer, M.: Spatial correlation as leading indicator of catastrophic shifts, Theor. Ecol., 3, 163–174, https://doi.org/10.1007/s12080-009-0060-6 , 2010. 

Dakos, V., Kéfi, S., Rietkerk, M., van Nes, E. H., and Scheffer, M.: Slowing down in spatially patterned ecosystems at the brink of collapse, Am. Nat., 177, E153–166, https://doi.org/10.1086/659945 , 2011. 

Dakos, V., Carpenter, S. R., Brock, W. A., Ellison, A. M., Guttal, V., Ives, A. R., Kéfi, S., Livina, V., Seekell, D. A., van Nes, E. H., and Scheffer, M.: Methods for Detecting Early Warnings of Critical Transitions in Time Series Illustrated Using Simulated Ecological Data, PLoS ONE, 7, e41010, https://doi.org/10.1371/journal.pone.0041010 , 2012. 

Dakos, V., van Nes, E. H., and Scheffer, M.: Flickering as an early warning signal, Theor. Ecol., 6, 309–317, https://doi.org/10.1007/s12080-013-0186-4 , 2013. 

Dakos, V., Carpenter, S. R., Nes, E. H. V., and Scheffer, M.: Resilience indicators: prospects and limitations for early warnings of regime shifts, Philos. T. R. Soc. B, 370, 20130263, https://doi.org/10.1098/rstb.2013.0263 , 2015. 

Deb, S., Sidheekh, S., Clements, C. F., Krishnan, N. C., and Dutta, P. S.: Machine learning methods trained on simple models can predict critical transitions in complex natural systems, Roy. Soc. Open Sci., 9, 211475, https://doi.org/10.1098/rsos.211475 , 2022. 

Ditlevsen, P. and Ditlevsen, S.: Warning of a forthcoming collapse of the Atlantic meridional overturning circulation, Nat. Commun., 14, 4254, https://doi.org/10.1038/s41467-023-39810-w , 2023. 

Drake, J. M. and Griffen, B. D.: Early warning signals of extinction in deteriorating environments, Nature, 467, 456–459, https://doi.org/10.1038/nature09389 , 2010. 

Drake, J. M., O'Regan, S. M., Dakos, V., Kéfi, S., Rohani, P., Drake, J. M., O'Regan, S. M., Dakos, V., Kéfi, S., and Rohani, P.: Alternative stable states, tipping points, and early warning signals of ecological transitions, in: Theoretical Ecology: Concepts and Applications, edited by: McCann, K. S. and Gellner, G., Oxford University Press, 263–284, https://doi.org/10.1093/oso/9780198824282.003.0015 , 2020. 

Dylewsky, D., Lenton, T. M., Scheffer, M., Bury, T. M., Fletcher, C. G., Anand, M., and Bauch, C. T.: Universal early warning signals of phase transitions in climate systems, J. R. Soc. Interface, 20, 20220562, https://doi.org/10.1098/rsif.2022.0562 , 2023. 

Erfanian, A., Wang, G., and Fomenko, L.: Unprecedented drought over tropical South America in 2016: significantly under-predicted by tropical SST, Sci. Rep.-UK, 7, 5811, https://doi.org/10.1038/s41598-017-05373-2 , 2017. 

Fanelli, D.: Negative results are disappearing from most disciplines and countries, Scientometrics, 90, 891–904, https://doi.org/10.1007/s11192-011-0494-7 , 2012. 

Forzieri, G., Dakos, V., McDowell, N. G., Ramdane, A., and Cescatti, A.: Emerging signals of declining forest resilience under climate change, Nature, 608, 534–539, https://doi.org/10.1038/s41586-022-04959-9 , 2022. 

Franco, A., Malhotra, N., and Simonovits, G.: Publication bias in the social sciences: Unlocking the file drawer, Science, 345, 1502–1505, https://doi.org/10.1126/science.1255484 , 2014. 

Génin, A., Majumder, S., Sankaran, S., Danet, A., Guttal, V., Schneider, F. D., and Kéfi, S.: Monitoring ecosystem degradation using spatial data and the R package spatialwarnings, Methods Ecol. Evol., 9, 2067–2075, https://doi.org/10.1111/2041-210X.13058 , 2018. 

Gilmore, R.: Catastrophe theory for scientists and engineers, John Wiley & Sons, ISBN 0-486-67539-4, 1981. 

Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., and Moore, R.: Google Earth Engine: Planetary-scale geospatial analysis for everyone, Remote Sens. Environ., 202, 18–27, https://doi.org/10.1016/j.rse.2017.06.031 , 2017. 

Guttal, V. and Jayaprakash, C.: Changing skewness: an early warning signal of regime shifts in ecosystems, Ecol. Lett., 11, 450–460, https://doi.org/10.1111/j.1461-0248.2008.01160.x , 2008. 

Guttal, V. and Jayaprakash, C.: Spatial variance and spatial skewness: leading indicators of regime shifts in spatial ecological systems, Theor. Ecol., 2, 3–12, https://doi.org/10.1007/s12080-008-0033-1 , 2009. 

Hagstrom, G. and Levin, S.: Phase Transitions and the Theory of Early Warning Indicators for Critical Transitions, arXiv [preprint], https://doi.org/10.48550/arXiv.2110.12287 , 23 October 2021. 

Hastings, A. and Wysham, D. B.: Regime shifts in ecological systems can occur with no warning, Ecol. Lett., 13, 464–472, https://doi.org/10.1111/j.1461-0248.2010.01439.x , 2010. 

Held, H. and Kleinen, T.: Detection of climate system bifurcations by degenerate fingerprinting, Geophys. Res. Lett., 31, 1–4, https://doi.org/10.1029/2004GL020972 , 2004. 

Hirota, M., Holmgren, M., van Nes, E. H., and Scheffer, M.: Global resilience of tropical forest and savanna to critical transitions, Science, 334, 232–235, 2011. 

Hu, Z., Dakos, V., and Rietkerk, M.: Using functional indicators to detect state changes in terrestrial ecosystems, Trends Ecol. Evol., 37, 1036–1045, https://doi.org/10.1016/j.tree.2022.07.011 , 2022. 

Hughes, T. P., Linares, C., Dakos, V., van de Leemput, I. A., and van Nes, E. H.: Living dangerously on borrowed time during slow, unrecognized regime shifts, Trends Ecol. Evol., 28, 149–55, https://doi.org/10.1016/j.tree.2012.08.022 , 2013. 

Ives, A. R. and Dakos, V.: Detecting dynamical changes in nonlinear time series using locally linear state-space models, Ecosphere, 3, 1–15, https://doi.org/10.1890/ES11-00347.1 , 2012. 

Jones, C., Lowe, J., Liddicoat, S., and Betts, R.: Committed terrestrial ecosystem changes due to climate change, Nat. Geosci., 2, 484–487, https://doi.org/10.1038/ngeo555 , 2009. 

Kéfi, S., Rietkerk, M., Alados, C. L., Pueyo, Y., Papanastasis, V. P., ElAich, A., de Ruiter, P. C., Alados, L., and Ruite r, P. C. D.: Spatial vegetation patterns and imminent desertification in Mediterranean arid ecosystems, Nature, 449, 213–217, https://doi.org/10.1038/nature06111 , 2007. 

Kéfi, S., Dakos, V., Scheffer, M., Van Nes, E. H., and Rietkerk, M.: Early warning signals also precede non-catastrophic transitions, Oikos, 122, 641–648, 2013. 

Kéfi, S., Guttal, V., Brock, W. A., Carpenter, S. R., Ellison, A. M., Livina, V. N., Seekell, D. A., Scheffer, M., van Nes, E. H., and Dakos, V.: Early Warning Signals of Ecological Transitions: Methods for Spatial Patterns, PLOS ONE, 9, e92097, https://doi.org/10.1371/journal.pone.0092097 , 2014. 

Kleinen, T., Held, H., Petschel-held, G., and Held, Á. H.: The potential role of spectral properties in detecting thresholds in the Earth system: application to the thermohaline circulation, Ocean Dynam., 53, 53–63, https://doi.org/10.1007/s10236-002-0023-6 , 2003. 

Klose, A. K., Karle, V., Winkelmann, R., and Donges, J. F.: Emergence of cascading dynamics in interacting tipping elements of ecology and climate: Cascading dynamics in tipping elements, Roy. Soc. Open Sci., 7, 200599, https://doi.org/10.1098/rsos.200599 , 2020. 

Klose, A. K., Wunderling, N., Winkelmann, R., and Donges, J. F.: What do we mean, “tipping cascade”?, Environ. Res. Lett., 16, 1–11, https://doi.org/10.1088/1748-9326/ac3955 , 2021. 

Kuehn, C.: A Mathematical Framework for Critical Transitions: Normal Forms, Variance and Applications, J. Nonlinear Sci., 23, 457–510, https://doi.org/10.1007/s00332-012-9158-x , 2012. 

Kuznetsov, Y. A.: Elements of Applied Bifurcation Theory, Springer-Verlag, New York, https://doi.org/10.1007/978-1-4757-3978-7 , 591 pp., 1995. 

Lade, S. J. and Gross, T.: Early warning signals for critical transitions: A generalized modeling approach, PLOS Comput. Biol., 8, e1002360, https://doi.org/10.1371/journal.pcbi.1002360 , 2012. 

Laitinen, V., Dakos, V., and Lahti, L.: Probabilistic early warning signals, Ecol. Evol., 1–14, https://doi.org/10.1002/ece3.8123 , 2021. 

Lapeyrolerie, M. and Boettiger, C.: Limits to ecological forecasting: Estimating uncertainty for critical transitions with deep learning, Meth. Ecol. Evol., 14, 785–798, https://doi.org/10.1111/2041-210X.14013 , 2022. 

Lenton, T. M.: Early warning of climate tipping points, Nat. Clim. Change, 1, 201–209, 2011. 

Lenton, T. M.: Environmental Tipping Points, 38, 1–29, https://doi.org/10.1146/annurev-environ-102511-084654 , 2013a. 

Lenton, T. M.: What early warning systems are there for environmental shocks?, Environ. Sci. Policy, 27, S60–S75, https://doi.org/10.1016/j.envsci.2012.06.011 , 2013b. 

Lenton, T. M., Buxton, J. E., Armstrong McKay, D. I., Abrams, J. F., Boulton, C. A., Lees, K., Powell, T. W. R., Boers, N., Cunliffe, A. M., and Dakos, V.: A resilience sensing system for the biosphere, Philos. T. R. Soc. B, 377, 20210383, https://doi.org/10.1098/rstb.2021.0383 , 2022. 

Lever, J. J., van de Leemput, I. A., Weinans, E., Quax, R., Dakos, V., van Nes, E. H., Bascompte, J., and Scheffer, M.: Foreseeing the future of mutualistic communities beyond collapse, Ecol. Lett., 23, 2–15, https://doi.org/10.1111/ele.13401 , 2020. 

Lewis, S. L., Brando, P. M., Phillips, O. L., Van Der Heijden, G. M. F., and Nepstad, D.: The 2010 Amazon Drought, Science, 331, 554–554, https://doi.org/10.1126/science.1200807 , 2011. 

Litzow, M. A. and Hunsicker, M. E.: Early warning signals, nonlinearity, and signs of hysteresis in real ecosystems, Ecosphere, 7, e01614, https://doi.org/10.1002/ecs2.1614 , 2016. 

Liu, L., Shao, Z., Lv, J., Xu, F., Ren, S., Jin, Q., Yang, J., Ma, W., Xie, H., Zhang, D., and Chen, X.: Identification of Early Warning Signals at the Critical Transition Point of Colorectal Cancer Based on Dynamic Network Analysis, Front. Bioeng. Biotechnol., 8, 530, https://doi.org/10.3389/fbioe.2020.00530 , 2020. 

Liu, Y., Kumar, M., Katul, G. G., and Porporato, A.: Reduced resilience as an early warning signal of forest mortality, Nat. Clim. Change, 9, 880–885, https://doi.org/10.1038/s41558-019-0583-9 , 2019. 

Livina, V. N. and Lenton, T. M.: A modified method for detecting incipient bifurcations in a dynamical system, Geophys. Res. Lett., 34, 1–5, https://doi.org/10.1029/2006GL028672 , 2007. 

Livina, V. N., Kwasniok, F., and Lenton, T. M.: Potential analysis reveals changing number of climate states during the last 60  kyr , Clim. Past, 6, 77–82, https://doi.org/10.5194/cp-6-77-2010 , 2010. 

Lu, Z., Yuan, N., Yang, Q., Ma, Z., and Kurths, J.: Early Warning of the Pacific Decadal Oscillation Phase Transition Using Complex Network Analysis, Geophys. Res. Lett., 48, e2020GL091674, https://doi.org/10.1029/2020GL091674 , 2021. 

Majumder, S., Tamma, K., Ramaswamy, S., and Guttal, V.: Inferring critical thresholds of ecosystem transitions from spatial data, Ecology, 100, e02722, https://doi.org/10.1002/ecy.2722 , 2019. 

Mayfield, R. J., Langdon, P. G., Doncaster, C. P., Dearing, J. A., Wang, R., Nazarova, L. B., Medeiros, A. S., and Brooks, S. J.: Metrics of structural change as indicators of chironomid community stability in high latitude lakes, Quaternary Sci. Rev., 249, 106594, https://doi.org/10.1016/j.quascirev.2020.106594 , 2020. 

Moesinger, L., Dorigo, W., de Jeu, R., van der Schalie, R., Scanlon, T., Teubner, I., and Forkel, M.: The global long-term microwave Vegetation Optical Depth Climate Archive (VODCA), Earth Syst. Sci. Data, 12, 177–196, https://doi.org/10.5194/essd-12-177-2020 , 2020. 

Nijp, J. J., Temme, A. J. A. M., van Voorn, G. A. K., Kooistra, L., Hengeveld, G. M., Soons, M. B., Teuling, A. J., and Wallinga, J.: Spatial early warning signals for impending regime shifts: A practical framework for application in real-world landscapes, Glob. Change Biol., 25, 1905–1921, https://doi.org/10.1111/gcb.14591 , 2019. 

Nowack, P., Runge, J., Eyring, V., and Haigh, J. D.: Causal networks for climate model evaluation and constrained projections, Nat. Commun., 11, 1415, https://doi.org/10.1038/s41467-020-15195-y , 2020. 

O'Brien, D. A., Deb, S., Sidheekh, S., Krishnan, N. C., Sharathi Dutta, P., and Clements, C. F.: EWSmethods: an R package to forecast tipping points at the community level using early warning signals, resilience measures, and machine learning models, Ecography, 2023, e06674, https://doi.org/10.1111/ecog.06674 , 2023. 

Rietkerk, M., Dekker, S. C., de Ruiter, P. C., and van de Koppel, J.: Self-Organized Patchiness and Catastrophic Shifts in Ecosystems, Science, 305, 1926–1929, https://doi.org/10.1126/science.1101867 , 2004. 

Rietkerk, M., van de Koppel, J., and Koppel, J. V. D.: Regular pattern formation in real ecosystems, Trends Ecol. Evol., 23, 169–175, https://doi.org/10.1016/j.tree.2007.10.013 , 2008. 

Ritchie, P. and Sieber, J.: Early-warning indicators for rate-induced tipping, Chaos, 26, 093116, https://doi.org/10.1063/1.4963012 , 2016. 

Saade, C., Fronhofer, E. A., Pichon, B., and Kéfi, S.: Landscape Structure Affects Metapopulation-Scale Tipping Points, Am. Nat., 202, E17–E30, https://doi.org/10.1086/724550 , 2023. 

Saatchi, S., Longo, M., Xu, L., Yang, Y., Abe, H., André, M., Aukema, J. E., Carvalhais, N., Cadillo-Quiroz, H., Cerbu, G. A., Chernela, J. M., Covey, K., Sánchez-Clavijo, L. M., Cubillos, I. V., Davies, S. J., De Sy, V., De Vleeschouwer, F., Duque, A., Sybille Durieux, A. M., De Avila Fernandes, K., Fernandez, L. E., Gammino, V., Garrity, D. P., Gibbs, D. A., Gibbon, L., Gowae, G. Y., Hansen, M., Lee Harris, N., Healey, S. P., Hilton, R. G., Johnson, C. M., Kankeu, R. S., Laporte-Goetz, N. T., Lee, H., Lovejoy, T., Lowman, M., Lumbuenamo, R., Malhi, Y., Albert Martinez, J.-M. M., Nobre, C., Pellegrini, A., Radachowsky, J., Román, F., Russell, D., Sheil, D., Smith, T. B., Spencer, R. G. M., Stolle, F., Tata, H. L., Torres, D. del C., Tshimanga, R. M., Vargas, R., Venter, M., West, J., Widayati, A., Wilson, S. N., Brumby, S., and Elmore, A. C.: Detecting vulnerability of humid tropical forests to multiple stressors, One Earth, 4, 988–1003, https://doi.org/10.1016/j.oneear.2021.06.002 , 2021. 

Scheffer, M., Bascompte, J., Brock, W. A., Brovkin, V., Carpenter, S. R., Dakos, V., Held, H., van Nes, E. H., Rietkerk, M., and Sugihara, G.: Early-warning signals for critical transitions, Nature, 461, 53–59, https://doi.org/10.1038/nature08227 , 2009. 

Scheffer, M., Carpenter, S. R., Lenton, T. M., Bascompte, J., Brock, W., Dakos, V., van de Koppel, J., van de Leemput, I. A., Levin, S. A., van Nes, E. H., Pascual, M., and Vandermeer, J.: Anticipating Critical Transitions, Science, 338, 344–348, https://doi.org/10.1126/science.1225244 , 2012a. 

Scheffer, M., Hirota, M., Holmgren, M., Van Nes, E. H., and Chapin, F. S.: Thresholds for boreal biome transitions., P. Natl. Acad. Sci. USA, 109, 21384–9, https://doi.org/10.1073/pnas.1219844110 , 2012b. 

Scheffer, M., Carpenter, S. R., Dakos, V., and van Nes, E. H.: Generic Indicators of Ecological Resilience: Inferring the Chance of a Critical Transition, Annu. Rev. Ecol. Evol. S., 46, 145–167, https://doi.org/10.1146/annurev-ecolsys-112414-054242 , 2015. 

Scheffer, M., Bolhuis, J. E., Borsboom, D., Buchman, T. G., Gijzel, S. M. W., Goulson, D., Kammenga, J. E., Kemp, B., van de Leemput, I. A., Levin, S., Martin, C. M., Melis, R. J. F., van Nes, E. H., Romero, L. M., and Olde Rikkert, M. G. M.: Quantifying resilience of humans and other animals, P. Natl. Acad. Sci. USA, 115, 201810630, https://doi.org/10.1073/pnas.1810630115 , 2018. 

Seekell, D. A. and Dakos, V.: Heteroskedasticity as a leading indicator of desertification in spatially explicit data, Ecol. Evol., 5, 2185–2192, https://doi.org/10.1002/ece3.1510 , 2015. 

Seekell, D. A., Carpenter, S. R., Cline, T. J., and Pace, M. L.: Conditional Heteroskedasticity Forecasts Regime Shift in a Whole-Ecosystem Experiment, Ecosystems, 15, 741–747, https://doi.org/10.1007/s10021-012-9542-2 , 2012. 

Smith, T., Zotta, R.-M., Boulton, C. A., Lenton, T. M., Dorigo, W., and Boers, N.: Reliability of resilience estimation based on multi-instrument time series, Earth Syst. Dynam., 14, 173–183, https://doi.org/10.5194/esd-14-173-2023 , 2023. 

Sole, R. V., Manrubia, S. C., Luque, B., Delgado, J., Bascompte, J., Solé, R. V., Manrubia, S. C., Luque, B., Delgado, J., and Bascompte, J.: Phase transitions and complex systems, Complexity, 1, 13–26, 1996. 

Spanbauer, T. L., Allen, C. R., Angeler, D. G., Eason, T., Fritz, S. C., Garmestani, A. S., Nash, K. L., Stone, J. R., Stow, C. A., and Sundstrom, S. M.: Body size distributions signal a regime shift in a lake ecosystem, P. R. Soc. B, 283, https://doi.org/10.1098/rspb.2016.0249 , 2016. 

Spears, B. M., Futter, M. N., Jeppesen, E., Huser, B. J., Ives, S., Davidson, T. A., Adrian, R., Angeler, D. G., Burthe, S. J., Carvalho, L., Daunt, F., Gsell, A. S., Hessen, D. O., Janssen, A. B. G., Mackay, E. B., May, L., Moorhouse, H., Olsen, S., Søndergaard, M., Woods, H., and Thackeray, S. J.: Ecological resilience in lakes and the conjunction fallacy, Nat. Ecol. Evol., 1, 1616–1624, https://doi.org/10.1038/s41559-017-0333-1 , 2017. 

Staver, A. C., Archibald, S., and Levin, S. A.: The Global Extent and Determinants of Savanna and Forest as Alternative Biome States, Science, 334, 230–232, https://doi.org/10.1126/science.1210465 , 2011. 

Tang, Y., Zhu, X., He, C., Hu, J., and Fan, J.: Critical slowing down theory provides early warning signals for sandstone failure, Front. Earth Sci., 10, 934498, https://doi.org/10.3389/feart.2022.934498 , 2022. 

Thompson, J. M. T. and Sieber, J.: Predicting climate tipping as a noisy bifurcation: a review, Int. J. Bifurcat Chaos, 21, 399–423, https://doi.org/10.1142/s0218127411028519 , 2011. 

Tirabassi, G., Viebahn, J., Dakos, V., Dijkstra, H. A., Masoller, C., Rietkerk, M., and Dekker, S. C.: Interaction network based early-warning indicators of vegetation transitions, Ecol. Complex., 19, 148–157, https://doi.org/10.1016/j.ecocom.2014.06.004 , 2014. 

Uden, D. R., Twidwell, D., Allen, C. R., Jones, M. O., Naugle, D. E., Maestas, J. D., Allred, B. W., and Wallace, K. J.: Spatial Imaging and Screening for Regime Shifts, Front. Ecol. Evol., 7, 1–16, https://doi.org/10.3389/fevo.2019.00407 , 2019. 

van de Leemput, I. A., Dakos, V., Scheffer, M., and van Nes, E. H.: Slow Recovery from Local Disturbances as an Indicator for Loss of Ecosystem Resilience, Ecosystems, 21, 141–152, https://doi.org/10.1007/s10021-017-0154-8 , 2018. 

van Der Bolt, B., van Nes, E. H., and Scheffer, M.: No warning for slow transitions, J. R. Soc. Interface, 18, 20200935, https://doi.org/10.1098/rsif.2020.0935 , 2021. 

van Nes, E. H., Arani, B. M. S., Staal, A., van der Bolt, B., Flores, B. M., Bathiany, S., and Scheffer, M.: What Do You Mean, `Tipping Point'?, Trends Ecol. Evol., 31, 902–904, https://doi.org/10.1016/j.tree.2016.09.011 , 2016. 

Veldhuis, M. P., Martinez-Garcia, R., Deblauwe, V., and Dakos, V.: Remotely-sensed slowing down in spatially patterned dryland ecosystems, Ecography, 2022, e06139, https://doi.org/10.1111/ecog.06139 , 2022. 

Veraart, A. J. A. J., Faassen, E. J. E. J., Dakos, V., Van Nes, E. H. E. H., Lürling, M., Scheffer, M., and Lurling, M.: Recovery rates reflect distance to a tipping point in a living system, Nature, 481, 357–359, https://doi.org/10.1038/nature10723 , 2012. 

Verbesselt, J., Umlauf, N., Hirota, M., Holmgren, M., Van Nes, E. H., Herold, M., Zeileis, A., and Scheffer, M.: Remotely sen sed resilience of tropical forests, Nat. Clim. Change, 6, 1028–1031, https://doi.org/10.1038/nclimate3108 , 2016. 

von Hardenberg, J., Meron, E., Shachak, M., Zarmi, Y., and Hardenberg, J. V.: Diversity of vegetation patterns and desertification, Phys. Rev. Lett., 8719, 198101, https://doi.org/10.1103/PhysRevLett.87.198101 , 2001. 

Wang, R., Dearing, J. A., Langdon, P. G., Zhang, E., Yang, X., Dakos, V., and Scheffer, M.: Flickering gives early warning signals of a critical transition to a eutrophic lake state, Nature, 492, 419–422, https://doi.org/10.1038/nature11655 , 2012. 

Weinans, E., Lever, J. J., Bathiany, S., Quax, R., Bascompte, J., Van Nes, E. H., Scheffer, M., and Van De Leemput, I. A.: Finding the direction of lowest resilience in multivariate complex systems, J. R. Soc. Interface, 16, https://doi.org/10.1098/rsif.2019.0629 , 2019. 

Weinans, E., Quax, R., van Nes, E. H., and Leemput, I. A. van de: Evaluating the performance of multivariate indicators of resilience loss, Sci. Rep.-UK, 11, 9148, https://doi.org/10.1038/s41598-021-87839-y , 2021.  

Wissel, C.: A universal law of the characteristic return time near thresholds, Oecologia, 65, 101–107, 1984. 

Wu, H., Hou, W., Yan, P.-C., Zhang, Z.-S., and Wang, K.: A study of the early warning signals of abrupt change in the Pacific decadal oscillation, Chinese Phys. B, 24, 089201, https://doi.org/10.1088/1674-1056/24/8/089201 , 2015. 

Wu, S., Liu, L., Li, D., Zhang, W., Liu, K., Shen, J., and Zhang, L.: Global desert expansion during the 21st century: Patterns, predictors and signals, Land Degrad. Dev., 34, 377–388, https://doi.org/10.1002/ldr.4466 , 2023. 

Yin, Z., Dekker, S. C., Rietkerk, M., van den Hurk, B. J. J. M., and Dijkstra, H. A.: Network based early warning indicators of vegetation changes in a land–atmosphere model, Ecol. Complex., 26, 68–78, https://doi.org/10.1016/j.ecocom.2016.02.004 , 2016. 

• Introduction
• The basics of early warnings
• Overview of early warnings empirical research in the last 20 years
• Conclusions
• Data availability
• Author contributions
• Competing interests
• Special issue statement
• Financial support
• Review statement

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Removal of atmospheric methane by increasing hydroxyl radicals via a water vapor enhancement strategy.

1. Introduction

1.1. importance of methane mitigation for climate, 1.2. relationship between methane, atmospheric oh, and water vapor, 1.2.1. relationship between methane and atmospheric oh, 1.2.2. relationship between atmospheric oh and water vapor, 2. effects of oh on methane concentration, 3. greenhouse effects of water vapor, 4. comparison of radiative forcing due to methane and water vapor variations, 5. discussion, 6. conclusions, author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest.

• Abernethy, S.; O’Connor, F.M.; Jones, C.D.; Jackson, R.B. Methane removal and the proportional reductions in surface temperature and ozone. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2021 , 379 , 20210104. [ Google Scholar ] [ CrossRef ]
• Shahabadi, M.B.; Huang, Y. Logarithmic radiative effect of water vapor and spectral kernels. J. Geophys. Res. Atmos. 2014 , 119 , 6000–6008. [ Google Scholar ] [ CrossRef ]
• Collins, W.J.; Webber, C.P.; Cox, P.M.; Huntingford, C.; Lowe, J.; Sitch, S.; E Chadburn, S.; Comyn-Platt, E.; Harper, A.B.; Hayman, G.; et al. Increased importance of methane reduction for a 1.5 degree target. Environ. Res. Lett. 2018 , 13 , 054003. [ Google Scholar ] [ CrossRef ]
• Collins, W.J.; Frame, D.J.; Fuglestvedt, J.S.; Shine, K.P. Stable climate metrics for emissions of short and long-lived species—Combining steps and pulses. Environ. Res. Lett. 2020 , 15 , 024018. [ Google Scholar ] [ CrossRef ]
• Dang, R.; Liao, H. Radiative Forcing and Health Impact of Aerosols and Ozone in China as the Consequence of Clean Air Actions over 2012–2017. Geophys. Res. Lett. 2019 , 46 , 12511–12519. [ Google Scholar ] [ CrossRef ]
• Dean, J.F.; Middelburg, J.J.; Röckmann, T.; Aerts, R.; Blauw, L.G.; Egger, M.; Jetten, M.S.M.; de Jong, A.E.E.; Meisel, O.H.; Rasigraf, O.; et al. Methane Feedbacks to the Global Climate System in a Warmer World. Rev. Geophys. 2018 , 56 , 207–250. [ Google Scholar ] [ CrossRef ]
• Dentener, F.; Stevenson, D.; Cofala, J.; Mechler, R.; Amann, M.; Bergamaschi, P.; Raes, F.; Derwent, R. The impact of air pollutant and methane emission controls on tropospheric ozone and radiative forcing: CTM calculations for the period 1990–2030. Atmos. Chem. Phys. 2005 , 5 , 1731–1755. [ Google Scholar ] [ CrossRef ]
• Dessler, A.E.; Zhang, Z.; Yang, P. Water-vapor climate feedback inferred from climate fluctuations, 2003–2008. Geophys. Res. Lett. 2008 , 35 , L20704. [ Google Scholar ] [ CrossRef ]
• Etminan, M.; Myhre, G.; Highwood, E.J.; Shine, K.P. Radiative forcing of carbon dioxide, methane, and nitrous oxide: A significant revision of the methane radiative forcing. Geophys. Res. Lett. 2016 , 43 , 12614–12623. [ Google Scholar ] [ CrossRef ]
• Feldl, N.; Bordoni, S.; Merlis, T.M. Coupled High-Latitude Climate Feedbacks and Their Impact on Atmospheric Heat Transport. J. Clim. 2017 , 30 , 189–201. [ Google Scholar ] [ CrossRef ]
• Gedney, N.; Huntingford, C.; Comyn-Platt, E.; Wiltshire, A. Significant feedbacks of wetland methane release on climate change and the causes of their uncertainty. Environ. Res. Lett. 2019 , 14 , 084027. [ Google Scholar ] [ CrossRef ]
• Gordon, N.D.; Jonko, A.K.; Forster, P.M.; Shell, K.M. An observationally based constraint on the water-vapor feedback. J. Geophys. Res. Atmos. 2013 , 118 , 12435–12443. [ Google Scholar ] [ CrossRef ]
• Haines, A.; Amann, M.; Borgford-Parnell, N.; Leonard, S.; Kuylenstierna, J.; Shindell, D. Short-lived climate pollutant mitigation and the Sustainable Development Goals. Nat. Clim. Chang. 2017 , 7 , 863–869. [ Google Scholar ] [ CrossRef ]
• Harmsen, M.; Fricko, O.; Hilaire, J.; van Vuuren, D.P.; Drouet, L.; Durand-Lasserve, O.; Fujimori, S.; Keramidas, K.; Klimont, Z.; Luderer, G.; et al. Taking some heat off the NDCs? The limited potential of additional short-lived climate forcers’ mitigation. Clim. Chang. 2020 , 163 , 1443–1461. [ Google Scholar ] [ CrossRef ]
• Heimann, I.; Griffiths, P.T.; Warwick, N.J.; Abraham, N.L.; Archibald, A.T.; Pyle, J.A. Methane Emissions in a Chemistry-Climate Model: Feedbacks and Climate Response. J. Adv. Model. Earth Syst. 2020 , 12 , e2019MS002019. [ Google Scholar ] [ CrossRef ]
• Heinze, C.; Eyring, V.; Friedlingstein, P.; Jones, C.; Balkanski, Y.; Collins, W.; Fichefet, T.; Gao, S.; Hall, A.; Ivanova, D.; et al. ESD Reviews: Climate feedbacks in the Earth system and prospects for their evaluation. Earth Syst. Dyn. 2019 , 10 , 379–452. [ Google Scholar ] [ CrossRef ]
• Held, I.M.; Soden, B.J. Water Vapor Feedback and Global Warming. Annu. Rev. Energy Environ. 2000 , 25 , 441–475. [ Google Scholar ] [ CrossRef ]
• Holmes, C.D.; Prather, M.J.; Søvde, O.A.; Myhre, G. Future methane, hydroxyl, and their uncertainties: Key climate and emission parameters for future predictions. Atmos. Chem. Phys. 2013 , 13 , 285–302. [ Google Scholar ] [ CrossRef ]
• Holmes, C.D. Methane Feedback on Atmospheric Chemistry: Methods, Models, and Mechanisms. J. Adv. Model. Earth Syst. 2018 , 10 , 1087–1099. [ Google Scholar ] [ CrossRef ]
• Huang, Y.; Ramaswamy, V.; Soden, B. An investigation of the sensitivity of the clear-sky outgoing longwave radiation to atmospheric temperature and water vapor. J. Geophys. Res. Atmos. 2007 , 112 , D05104. [ Google Scholar ] [ CrossRef ]
• Huang, Y.; Xia, Y.; Tan, X. On the pattern of CO 2 radiative forcing and poleward energy transport. J. Geophys. Res. Atmos. 2017 , 122 , 10578–10593. [ Google Scholar ] [ CrossRef ]
• IPCC. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change ; Stocker, T.F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P.M., Eds.; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2013. [ Google Scholar ]
• IPCC. Summary for Policymakers. In Global Warming of 1.5 °C. An IPCC Special Report on the Impacts of Global Warming of 1.5 °C above Pre-Industrial Levels and Related Global Greenhouse Gas Emission Pathways, in the Context of Strengthening the Global Response to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate Poverty ; Masson-Delmotte, V., Zhai, P., Pörtner, H.-O., Roberts, D., Skea, J., Shukla, P.R., Pirani, A., Moufouma-Okia, W., Péan, C., Pidcock, R., et al., Eds.; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2018; pp. 3–24. [ Google Scholar ]
• IPCC. Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change ; Masson-Delmotte, V., Zhai, P., Pirani, A., Connors, S.L., Péan, C., Berger, S., Caud, N., Chen, Y., Goldfarb, L., Gomis, M.I., et al., Eds.; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2021; p. 2391. [ Google Scholar ]
• Isaksen, I.S.A.; Gauss, M.; Myhre, G.; Anthony, K.M.W.; Ruppel, C. Strong atmospheric chemistry feedback to climate warming from Arctic methane emissions. Glob. Biogeochem. Cycles 2011 , 25 , GB2002. [ Google Scholar ] [ CrossRef ]
• Burkholder, J.B.; Sander, S.P.; Abbatt, J.; Barker, J.R.; Cappa, C.; Crounse, J.D.; Dibble, T.S.; Huie, R.E.; Kolb, C.E.; Kurylo, M.J.; et al. Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies, Evaluation No. 19 ; Jet Propulsion Laboratory: Pasadena, CA, USA, 2019. [ Google Scholar ]
• Jackson, R.B.; Solomon, E.I.; Canadell, J.G.; Cargnello, M.; Field, C.B. Methane removal and atmospheric restoration. Nat. Sustain. 2019 , 2 , 436–438. [ Google Scholar ] [ CrossRef ]
• Koffi, E.N.; Bergamaschi, P.; Alkama, R.; Cescatti, A. An observation-constrained assessment of the climate sensitivity and future trajectories of wetland methane emissions. Sci. Adv. 2020 , 6 , eaay4444. [ Google Scholar ] [ CrossRef ]
• Kramer, R.J.; Matus, A.V.; Soden, B.J.; L’Ecuyer, T.S. Observation-Based Radiative Kernels From CloudSat/CALIPSO. J. Geophys. Res. Atmos. 2019 , 124 , 5431–5444. [ Google Scholar ] [ CrossRef ]
• Lan, X.; Thoning, K.W.; Dlugokencky, E.J. Trends in Globally-Averaged CH4, N2O, and SF6 Determined from NOAA Global Monitoring Laboratory Measurements (No. Version 2023-04). 2023. Available online: https://gml.noaa.gov/ccgg/trends_doi.html (accessed on 1 June 2023).
• Lelieveld, J.; Gromov, S.; Pozzer, A.; Taraborrelli, D. Global tropospheric hydroxyl distribution, budget and reactivity. Atmos. Chem. Phys. 2016 , 16 , 12477–12493. [ Google Scholar ] [ CrossRef ]
• Li, M.; Karu, E.; Brenninkmeijer, C.; Fischer, H.; Lelieveld, J.; Williams, J. Tropospheric OH and stratospheric OH and Cl concentrations determined from CH4, CH3Cl, and SF6 measurements. NPJ Clim. Atmospheric Sci. 2018 , 1 , 29. [ Google Scholar ] [ CrossRef ]
• Liu, R.; Su, H.; Liou, K.-N.; Jiang, J.H.; Gu, Y.; Liu, S.C.; Shiu, C. An Assessment of Tropospheric Water Vapor Feedback Using Radiative Kernels. J. Geophys. Res. Atmos. 2018 , 123 , 1499–1509. [ Google Scholar ] [ CrossRef ]
• López-Comí, L.; Morgenstern, O.; Zeng, G.; Masters, S.L.; Querel, R.R.; Nedoluha, G.E. Assessing the sensitivity of the hydroxyl radical to model biases in composition and temperature using a single-column photochemical model for Lauder, New Zealand. Atmos. Chem. Phys. 2016 , 16 , 14599–14619. [ Google Scholar ] [ CrossRef ]
• Lund, M.T.; Aamaas, B.; Stjern, C.W.; Klimont, Z.; Berntsen, T.K.; Samset, B.H. A continued role of short-lived climate forcers under the Shared Socioeconomic Pathways. Earth Syst. Dyn. 2020 , 11 , 977–993. [ Google Scholar ] [ CrossRef ]
• Maycock, A.C.; Smith, C.J.; Rap, A.; Rutherford, O. On the Structure of Instantaneous Radiative Forcing Kernels for Greenhouse Gases. J. Atmospheric Sci. 2021 , 78 , 949–965. [ Google Scholar ] [ CrossRef ]
• Meinshausen, M.; Nicholls, Z.R.J.; Lewis, J.; Gidden, M.J.; Vogel, E.; Freund, M.; Beyerle, U.; Gessner, C.; Nauels, A.; Bauer, N.; et al. The shared socio-economic pathway (SSP) greenhouse gas concentrations and their extensions to 2500. Geosci. Model Dev. 2020 , 13 , 3571–3605. [ Google Scholar ] [ CrossRef ]
• Miner, K.R.; Turetsky, M.R.; Malina, E.; Bartsch, A.; Tamminen, J.; McGuire, A.D.; Fix, A.; Sweeney, C.; Elder, C.D.; Miller, C.E. Permafrost carbon emissions in a changing Arctic. Nat. Rev. Earth Environ. 2022 , 3 , 55–67. [ Google Scholar ] [ CrossRef ]
• Monks, S.A.; Arnold, S.R.; Emmons, L.K.; Law, K.S.; Turquety, S.; Duncan, B.N.; Flemming, J.; Huijnen, V.; Tilmes, S.; Langner, J.; et al. Multi-model study of chemical and physical controls on transport of anthropogenic and biomass burning pollution to the Arctic. Atmos. Chem. Phys. 2015 , 15 , 3575–3603. [ Google Scholar ] [ CrossRef ]
• Montzka, S.A.; Spivakovsky, C.M.; Butler, J.H.; Elkins, J.W.; Lock, L.T.; Mondeel, D.J. New Observational Constraints for Atmospheric Hydroxyl on Global and Hemispheric Scales. Science 2000 , 288 , 500–503. [ Google Scholar ] [ CrossRef ]
• Naik, V.; Voulgarakis, A.; Fiore, A.M.; Horowitz, L.W.; Lamarque, J.-F.; Lin, M.; Prather, M.J.; Young, P.J.; Bergmann, D.; Cameron-Smith, P.J.; et al. Preindustrial to present-day changes in tropospheric hydroxyl radical and methane lifetime from the Atmospheric Chemistry and Climate Model Intercomparison Project (ACCMIP). Atmos. Chem. Phys. 2013 , 13 , 5277–5298. [ Google Scholar ] [ CrossRef ]
• Naus, S.; Montzka, S.A.; Pandey, S.; Basu, S.; Dlugokencky, E.J.; Krol, M. Constraints and biases in a tropospheric two-box model of OH. Atmos. Chem. Phys. 2019 , 19 , 407–424. [ Google Scholar ] [ CrossRef ]
• Nguyen, N.H.; Turner, A.J.; Yin, Y.; Prather, M.J.; Frankenberg, C. Effects of Chemical Feedbacks on Decadal Methane Emissions Estimates. Geophys. Res. Lett. 2020 , 47 , e2019GL085706. [ Google Scholar ] [ CrossRef ]
• Nicely, J.M.; Salawitch, R.J.; Canty, T.; Anderson, D.C.; Arnold, S.R.; Chipperfield, M.P.; Emmons, L.K.; Flemming, J.; Huijnen, V.; Kinnison, D.E.; et al. Quantifying the causes of differences in tropospheric OH within global models. J. Geophys. Res. Atmos. 2017 , 122 , 1983–2007. [ Google Scholar ] [ CrossRef ]
• Nicely, J.M.; Canty, T.P.; Manyin, M.; Oman, L.D.; Salawitch, R.J.; Steenrod, S.D.; Strahan, S.E.; Strode, S.A. Changes in Global Tropospheric OH Expected as a Result of Climate Change Over the Last Several Decades. J. Geophys. Res. Atmos. 2018 , 123 , 10774–10795. [ Google Scholar ] [ CrossRef ]
• Nisbet, E.G.; Fisher, R.E.; Lowry, D.; France, J.L.; Allen, G.; Bakkaloglu, S.; Broderick, T.J.; Cain, M.; Coleman, M.; Fernandez, J.; et al. Methane Mitigation: Methods to Reduce Emissions, on the Path to the Paris Agreement. Rev. Geophys. 2020 , 58 , e2019RG000675. [ Google Scholar ] [ CrossRef ]
• Ocko, I.B.; Sun, T.; Shindell, D.; Oppenheimer, M.; Hristov, A.N.; Pacala, S.W.; Mauzerall, D.L.; Xu, Y.; Hamburg, S.P. Acting rapidly to deploy readily available methane mitigation measures by sector can immediately slow global warming. Environ. Res. Lett. 2021 , 16 , 054042. [ Google Scholar ] [ CrossRef ]
• Patra, P.K.; Krol, M.C.; Montzka, S.A.; Arnold, T.; Atlas, E.L.; Lintner, B.R.; Stephens, B.B.; Xiang, B.; Elkins, J.W.; Fraser, P.J.; et al. Observational evidence for interhemispheric hydroxyl-radical parity. Nature 2014 , 513 , 219–223. [ Google Scholar ] [ CrossRef ]
• Pendergrass, A.G.; Conley, A.; Vitt, F.M. Surface and top-of-atmosphere radiative feedback kernels for CESM-CAM5. Earth Syst. Sci. Data 2018 , 10 , 317–324. [ Google Scholar ] [ CrossRef ]
• Peng, S.; Lin, X.; Thompson, R.L.; Xi, Y.; Liu, G.; Hauglustaine, D.; Lan, X.; Poulter, B.; Ramonet, M.; Saunois, M.; et al. Wetland emission and atmospheric sink changes explain methane growth in 2020. Nature 2022 , 612 , 477–482. [ Google Scholar ] [ CrossRef ]
• Prather, M.J. Time scales in atmospheric chemistry: Theory, GWPs for CH 4 and CO, and runaway growth. Geophys. Res. Lett. 1996 , 23 , 2597–2600. [ Google Scholar ] [ CrossRef ]
• Prather, M.J. Lifetimes and time scales in atmospheric chemistry. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2007 , 365 , 1705–1726. [ Google Scholar ] [ CrossRef ]
• Prather, M.J.; Holmes, C.D.; Hsu, J. Reactive greenhouse gas scenarios: Systematic exploration of uncertainties and the role of atmospheric chemistry. Geophys. Res. Lett. 2012 , 39 , L09803. [ Google Scholar ] [ CrossRef ]
• Saunois, M.; Stavert, A.R.; Poulter, B.; Bousquet, P.; Canadell, J.G.; Jackson, R.B.; Raymond, P.A.; Dlugokencky, E.J.; Houweling, S.; Patra, P.K.; et al. The Global Methane Budget 2000–2017. Earth Syst. Sci. Data 2020 , 12 , 1561–1623. [ Google Scholar ] [ CrossRef ]
• Shell, K.M.; Kiehl, J.T.; Shields, C.A. Using the Radiative Kernel Technique to Calculate Climate Feedbacks in NCAR’s Community Atmospheric Model. J. Clim. 2008 , 21 , 2269–2282. [ Google Scholar ] [ CrossRef ]
• Sherwood, S.C.; Dixit, V.; Salomez, C. The global warming potential of near-surface emitted water vapour. Environ. Res. Lett. 2018 , 13 , 104006. [ Google Scholar ] [ CrossRef ]
• Smith, C.J.; Kramer, R.J.; Myhre, G.; Forster, P.M.; Soden, B.J.; Andrews, T.; Boucher, O.; Faluvegi, G.; Fläschner, D.; Hodnebrog, Ø.; et al. Understanding Rapid Adjustments to Diverse Forcing Agents. Geophys. Res. Lett. 2018 , 45 , 12023–12031. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Soden, B.J.; Held, I.M.; Colman, R.; Shell, K.M.; Kiehl, J.T.; Shields, C.A. Quantifying Climate Feedbacks Using Radiative Kernels. J. Clim. 2008 , 21 , 3504–3520. [ Google Scholar ] [ CrossRef ]
• Stevenson, D.S.; Zhao, A.; Naik, V.; O’Connor, F.M.; Tilmes, S.; Zeng, G.; Murray, L.T.; Collins, W.J.; Griffiths, P.T.; Shim, S.; et al. Trends in global tropospheric hydroxyl radical and methane lifetime since 1850 from AerChemMIP. Atmos. Chem. Phys. 2020 , 20 , 12905–12920. [ Google Scholar ] [ CrossRef ]
• Sun, T.; Ocko, I.B.; Hamburg, S.P. The value of early methane mitigation in preserving Arctic summer sea ice. Environ. Res. Lett. 2022 , 17 , 044001. [ Google Scholar ] [ CrossRef ]
• Thorsen, T.J.; Kato, S.; Loeb, N.G.; Rose, F.G. Observation-Based Decomposition of Radiative Perturbations and Radiative Kernels. J. Clim. 2018 , 31 , 10039–10058. [ Google Scholar ] [ CrossRef ]
• Tian, H.Q.; Xu, R.T.; Canadell, J.G.; Thompson, R.L.; Winiwarter, W.; Suntharalingam, P.; Davidson, E.A.; Ciais, P.; Jackson, R.B.; Janssens-Maenhout, G.; et al. A comprehensive quantification of global nitrous oxide sources and sinks. Nature 2020 , 586 , 248–256. [ Google Scholar ] [ CrossRef ]
• Turetsky, M.R.; Abbott, B.W.; Jones, M.C.; Anthony, K.W.; Olefeldt, D.; Schuur, E.A.G.; Grosse, G.; Kuhry, P.; Hugelius, G.; Koven, C.; et al. Carbon release through abrupt permafrost thaw. Nat. Geosci. 2020 , 13 , 138–143. [ Google Scholar ] [ CrossRef ]
• Turner, A.J.; Frankenberg, C.; Wennberg, P.O.; Jacob, D.J. Ambiguity in the causes for decadal trends in atmospheric methane and hydroxyl. Proc. Natl. Acad. Sci. USA 2017 , 114 , 5367–5372. [ Google Scholar ] [ CrossRef ]
• Turner, A.J.; Frankenberg, C.; Kort, E.A. Interpreting contemporary trends in atmospheric methane. Proc. Natl. Acad. Sci. USA 2019 , 116 , 2805–2813. [ Google Scholar ] [ CrossRef ]
• Anthony, K.W.; von Deimling, T.S.; Nitze, I.; Frolking, S.; Emond, A.; Daanen, R.; Anthony, P.; Lindgren, P.; Jones, B.; Grosse, G. 21st-century modeled permafrost carbon emissions accelerated by abrupt thaw beneath lakes. Nat. Commun. 2018 , 9 , 3262. [ Google Scholar ] [ CrossRef ]
• Wang, X.; Jacob, D.J.; Eastham, S.D.; Sulprizio, M.P.; Zhu, L.; Chen, Q.; Alexander, B.; Sherwen, T.; Evans, M.J.; Lee, B.H.; et al. The role of chlorine in global tropospheric chemistry. Atmos. Chem. Phys. 2019 , 19 , 3981–4003. [ Google Scholar ] [ CrossRef ]
• Wang, Y.; Ming, T.; Li, W.; Yuan, Q.; de Richter, R.; Davies, P.; Caillol, S. Atmospheric removal of methane by enhancing the natural hydroxyl radical sink. Greenh. Gases Sci. Technol. 2022 , 12 , 784–795. [ Google Scholar ] [ CrossRef ]
• Wuebbles, D.J.; Tamaresis, J.S. The Role of Methane in the Global Environment. In Atmospheric Methane: Sources, Sinks, and Role in Global Change ; Khalil, M.A.K., Ed.; Springer: Berlin/Heidelberg, Germany, 1993; pp. 469–513. [ Google Scholar ]
• Xiong, H.; Ming, T.; Wu, Y.; Li, W.; Mu, L.; de Richter, R.; Yan, S.; Yuan, Y.; Peng, C. Numerical analysis of a negative emission technology of methane to mitigate climate change. Sol. Energy 2023 , 255 , 416–424. [ Google Scholar ] [ CrossRef ]
• Yin, Y.; Chevallier, F.; Ciais, P.; Broquet, G.; Fortems-Cheiney, A.; Pison, I.; Saunois, M. Decadal trends in global CO emissions as seen by MOPITT. Atmos. Chem. Phys. 2015 , 15 , 13433–13451. [ Google Scholar ] [ CrossRef ]
• Yin, Y.; Chevallier, F.; Ciais, P.; Bousquet, P.; Saunois, M.; Zheng, B.; Worden, J.; Bloom, A.A.; Parker, R.J.; Jacob, D.J.; et al. Accelerating methane growth rate from 2010 to 2017: Leading contributions from the tropics and East Asia. Atmos. Chem. Phys. 2021 , 21 , 12631–12647. [ Google Scholar ] [ CrossRef ]
• Yuan, W.; Zheng, Y.; Piao, S.; Ciais, P.; Lombardozzi, D.; Wang, Y.; Ryu, Y.; Chen, G.; Dong, W.; Hu, Z.; et al. Increased atmospheric vapor pressure deficit reduces global vegetation growth. Sci. Adv. 2019 , 5 , eaax1396. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Zelinka, M.D.; Hartmann, D.L. Climate Feedbacks and Their Implications for Poleward Energy Flux Changes in a Warming Climate. J. Clim. 2012 , 25 , 608–624. [ Google Scholar ] [ CrossRef ]
• Zhang, Z.; Zimmermann, N.E.; Stenke, A.; Li, X.; Hodson, E.L.; Zhu, G.; Huang, C.; Poulter, B. Emerging role of wetland methane emissions in driving 21st century climate change. Proc. Natl. Acad. Sci. USA 2017 , 114 , 9647–9652. [ Google Scholar ] [ CrossRef ]
• Zhao, Y.; Saunois, M.; Bousquet, P.; Lin, X.; Berchet, A.; Hegglin, M.I.; Canadell, J.G.; Jackson, R.B.; Dlugokencky, E.J.; Langenfelds, R.L.; et al. Influences of hydroxyl radicals (OH) on top-down estimates of the global and regional methane budgets. Atmos. Chem. Phys. 2020 , 20 , 9525–9546. [ Google Scholar ] [ CrossRef ]
• Zhao, Y.; Saunois, M.; Bousquet, P.; Lin, X.; Berchet, A.; Hegglin, M.I.; Canadell, J.G.; Jackson, R.B.; Deushi, M.; Jöckel, P.; et al. On the role of trend and variability in the hydroxyl radical (OH) in the global methane budget. Atmos. Chem. Phys. 2020 , 20 , 13011–13022. [ Google Scholar ] [ CrossRef ]

Click here to enlarge figure

Share and Cite

Liu, Y.; Yao, X.; Zhou, L.; Ming, T.; Li, W.; de Richter, R. Removal of Atmospheric Methane by Increasing Hydroxyl Radicals via a Water Vapor Enhancement Strategy. Atmosphere 2024 , 15 , 1046. https://doi.org/10.3390/atmos15091046

Liu Y, Yao X, Zhou L, Ming T, Li W, de Richter R. Removal of Atmospheric Methane by Increasing Hydroxyl Radicals via a Water Vapor Enhancement Strategy. Atmosphere . 2024; 15(9):1046. https://doi.org/10.3390/atmos15091046

Liu, Yang, Xiaokun Yao, Li Zhou, Tingzhen Ming, Wei Li, and Renaud de Richter. 2024. "Removal of Atmospheric Methane by Increasing Hydroxyl Radicals via a Water Vapor Enhancement Strategy" Atmosphere 15, no. 9: 1046. https://doi.org/10.3390/atmos15091046

Article Metrics

Further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 22 July 2024

Neural general circulation models for weather and climate

• Dmitrii Kochkov   ORCID: orcid.org/0000-0003-3846-4911 1   na1 ,
• Janni Yuval   ORCID: orcid.org/0000-0001-7519-0118 1   na1 ,
• Ian Langmore 1   na1 ,
• Peter Norgaard 1   na1 ,
• Jamie Smith 1   na1 ,
• Griffin Mooers 1 ,
• Milan Klöwer 2 ,
• James Lottes 1 ,
• Stephan Rasp 1 ,
• Peter Düben   ORCID: orcid.org/0000-0002-4610-3326 3 ,
• Sam Hatfield 3 ,
• Peter Battaglia 4 ,
• Alvaro Sanchez-Gonzalez 4 ,
• Matthew Willson   ORCID: orcid.org/0000-0002-8730-1927 4 ,
• Michael P. Brenner 1 , 5 &
• Stephan Hoyer   ORCID: orcid.org/0000-0002-5207-0380 1   na1  

Nature volume  632 ,  pages 1060–1066 ( 2024 ) Cite this article

54k Accesses

3 Citations

678 Altmetric

Metrics details

• Atmospheric dynamics
• Climate and Earth system modelling
• Computational science

General circulation models (GCMs) are the foundation of weather and climate prediction 1 , 2 . GCMs are physics-based simulators that combine a numerical solver for large-scale dynamics with tuned representations for small-scale processes such as cloud formation. Recently, machine-learning models trained on reanalysis data have achieved comparable or better skill than GCMs for deterministic weather forecasting 3 , 4 . However, these models have not demonstrated improved ensemble forecasts, or shown sufficient stability for long-term weather and climate simulations. Here we present a GCM that combines a differentiable solver for atmospheric dynamics with machine-learning components and show that it can generate forecasts of deterministic weather, ensemble weather and climate on par with the best machine-learning and physics-based methods. NeuralGCM is competitive with machine-learning models for one- to ten-day forecasts, and with the European Centre for Medium-Range Weather Forecasts ensemble prediction for one- to fifteen-day forecasts. With prescribed sea surface temperature, NeuralGCM can accurately track climate metrics for multiple decades, and climate forecasts with 140-kilometre resolution show emergent phenomena such as realistic frequency and trajectories of tropical cyclones. For both weather and climate, our approach offers orders of magnitude computational savings over conventional GCMs, although our model does not extrapolate to substantially different future climates. Our results show that end-to-end deep learning is compatible with tasks performed by conventional GCMs and can enhance the large-scale physical simulations that are essential for understanding and predicting the Earth system.

Similar content being viewed by others

Accurate medium-range global weather forecasting with 3D neural networks

Deep learning for twelve hour precipitation forecasts

Skilful predictions of the Asian summer monsoon one year ahead

Solving the equations for Earth’s atmosphere with general circulation models (GCMs) is the basis of weather and climate prediction 1 , 2 . Over the past 70 years, GCMs have been steadily improved with better numerical methods and more detailed physical models, while exploiting faster computers to run at higher resolution. Inside GCMs, the unresolved physical processes such as clouds, radiation and precipitation are represented by semi-empirical parameterizations. Tuning GCMs to match historical data remains a manual process 5 , and GCMs retain many persistent errors and biases 6 , 7 , 8 . The difficulty of reducing uncertainty in long-term climate projections 9 and estimating distributions of extreme weather events 10 presents major challenges for climate mitigation and adaptation 11 .

Recent advances in machine learning have presented an alternative for weather forecasting 3 , 4 , 12 , 13 . These models rely solely on machine-learning techniques, using roughly 40 years of historical data from the European Center for Medium-Range Weather Forecasts (ECMWF) reanalysis v5 (ERA5) 14 for model training and forecast initialization. Machine-learning methods have been remarkably successful, demonstrating state-of-the-art deterministic forecasts for 1- to 10-day weather prediction at a fraction of the computational cost of traditional models 3 , 4 . Machine-learning atmospheric models also require considerably less code, for example GraphCast 3 has 5,417 lines versus 376,578 lines for the National Oceanic and Atmospheric Administration’s FV3 atmospheric model 15 (see Supplementary Information section  A for details).

Nevertheless, machine-learning approaches have noteworthy limitations compared with GCMs. Existing machine-learning models have focused on deterministic prediction, and surpass deterministic numerical weather prediction in terms of the aggregate metrics for which they are trained 3 , 4 . However, they do not produce calibrated uncertainty estimates 4 , which is essential for useful weather forecasts 1 . Deterministic machine-learning models using a mean-squared-error loss are rewarded for averaging over uncertainty, producing unrealistically blurry predictions when optimized for multi-day forecasts 3 , 13 . Unlike physical models, machine-learning models misrepresent derived (diagnostic) variables such as geostrophic wind 16 . Furthermore, although there has been some success in using machine-learning approaches on longer timescales 17 , 18 , these models have not demonstrated the ability to outperform existing GCMs.

Hybrid models that combine GCMs with machine learning are appealing because they build on the interpretability, extensibility and successful track record of traditional atmospheric models 19 , 20 . In the hybrid model approach, a machine-learning component replaces or corrects the traditional physical parameterizations of a GCM. Until now, the machine-learning component in such models has been trained ‘offline’, by learning parameterizations independently of their interaction with dynamics. These components are then inserted into an existing GCM. The lack of coupling between machine-learning components and the governing equations during training potentially causes serious problems, such as instability and climate drift 21 . So far, hybrid models have mostly been limited to idealized scenarios such as aquaplanets 22 , 23 . Under realistic conditions, machine-learning corrections have reduced some biases of very coarse GCMs 24 , 25 , 26 , but performance remains considerably worse than state-of-the-art models.

Here we present NeuralGCM, a fully differentiable hybrid GCM of Earth’s atmosphere. NeuralGCM is trained on forecasting up to 5-day weather trajectories sampled from ERA5. Differentiability enables end-to-end ‘online training’ 27 , with machine-learning components optimized in the context of interactions with the governing equations for large-scale dynamics, which we find enables accurate and stable forecasts. NeuralGCM produces physically consistent forecasts with accuracy comparable to best-in-class models across a range of timescales, from 1- to 15-day weather to decadal climate prediction.

Neural GCMs

A schematic of NeuralGCM is shown in Fig. 1 . The two key components of NeuralGCM are a differentiable dynamical core for solving the discretized governing dynamical equations and a learned physics module that parameterizes physical processes with a neural network, described in full detail in Methods , Supplementary Information sections  B and C , and Supplementary Table 1 . The dynamical core simulates large-scale fluid motion and thermodynamics under the influence of gravity and the Coriolis force. The learned physics module (Supplementary Fig. 1 ) predicts the effect of unresolved processes, such as cloud formation, radiative transport, precipitation and subgrid-scale dynamics, on the simulated fields using a neural network.

a , Overall model structure, showing how forcings F t , noise z t (for stochastic models) and inputs y t are encoded into the model state x t . The model state is fed into the dynamical core, and alongside forcings and noise into the learned physics module. This produces tendencies (rates of change) used by an implicit–explicit ordinary differential equation (ODE) solver to advance the state in time. The new model state x t +1 can then be fed back into another time step, or decoded into model predictions. b , The learned physics module, which feeds data for individual columns of the atmosphere into a neural network used to produce physics tendencies in that vertical column.

The differentiable dynamical core in NeuralGCM allows an end-to-end training approach, whereby we advance the model multiple time steps before employing stochastic gradient descent to minimize discrepancies between model predictions and reanalysis (Supplementary Information section  G.2 ). We gradually increase the rollout length from 6 hours to 5 days (Supplementary Information section  G and Supplementary Table 5 ), which we found to be critical because our models are not accurate for multi-day prediction or stable for long rollouts early in training (Supplementary Information section  H.6.2 and Supplementary Fig. 23 ). The extended back-propagation through hundreds of simulation steps enables our neural networks to take into account interactions between the learned physics and the dynamical core. We train deterministic and stochastic NeuralGCM models, each of which uses a distinct training protocol, described in full detail in Methods and Supplementary Table 4 .

We train a range of NeuralGCM models at horizontal resolutions with grid spacing of 2.8°, 1.4° and 0.7° (Supplementary Fig. 7 ). We evaluate the performance of NeuralGCM at a range of timescales appropriate for weather forecasting and climate simulation. For weather, we compare against the best-in-class conventional physics-based weather models, ECMWF’s high-resolution model (ECMWF-HRES) and ensemble prediction system (ECMWF-ENS), and two of the recent machine-learning-based approaches, GraphCast 3 and Pangu 4 . For climate, we compare against a global cloud-resolving model and Atmospheric Model Intercomparison Project (AMIP) runs.

Medium-range weather forecasting

Our evaluation set-up focuses on quantifying accuracy and physical consistency, following WeatherBench2 12 . We regrid all forecasts to a 1.5° grid using conservative regridding, and average over all 732 forecasts made at noon and midnight UTC in the year 2020, which was held-out from training data for all machine-learning models. NeuralGCM, GraphCast and Pangu compare with ERA5 as the ground truth, whereas ECMWF-ENS and ECMWF-HRES compare with the ECMWF operational analysis (that is, HRES at 0-hour lead time), to avoid penalizing the operational forecasts for different biases than ERA5.

Model accuracy

We use ECMWF’s ensemble (ENS) model as a reference baseline as it achieves the best performance across the majority of lead times 12 . We assess accuracy using (1) root-mean-squared error (RMSE), (2) root-mean-squared bias (RMSB), (3) continuous ranked probability score (CRPS) and (4) spread-skill ratio, with the results shown in Fig. 2 . We provide more in-depth evaluations including scorecards, metrics for additional variables and levels and maps in Extended Data Figs. 1 and 2 , Supplementary Information section  H and Supplementary Figs. 9 – 22 .

a , c , RMSE ( a ) and RMSB ( c ) for ECMWF-ENS, ECMWF-HRES, NeuralGCM-0.7°, NeuralGCM-ENS, GraphCast 3 and Pangu 4 on headline WeatherBench2 variables, as a percentage of the error of ECMWF-ENS. Deterministic and stochastic models are shown in solid and dashed lines respectively. e , g , CRPS relative to ECMWF-ENS ( e ) and spread-skill ratio for the ENS and NeuralGCM-ENS models ( g ). b , d , f , h , Spatial distributions of RMSE ( b ), bias ( d ), CRPS ( f ) and spread-skill ratio ( h ) for NeuralGCM-ENS and ECMWF-ENS models for 10-day forecasts of specific humidity at 700 hPa. Spatial plots of RMSE and CRPS show skill relative to a probabilistic climatology 12 with an ensemble member for each of the years 1990–2019. The grey areas indicate regions where climatological surface pressure on average is below 700 hPa.

Deterministic models that produce a single weather forecast for given initial conditions can be compared effectively using RMSE skill at short lead times. For the first 1–3 days, depending on the atmospheric variable, RMSE is minimized by forecasts that accurately track the evolution of weather patterns. At this timescale we find that NeuralGCM-0.7° and GraphCast achieve best results, with slight variations across different variables (Fig. 2a ). At longer lead times, RMSE rapidly increases owing to chaotic divergence of nearby weather trajectories, making RMSE less informative for deterministic models. RMSB calculates persistent errors over time, which provides an indication of how models would perform at much longer lead times. Here NeuralGCM models also compare favourably against previous approaches (Fig. 2c ), with notably much less bias for specific humidity in the tropics (Fig. 2d ).

Ensembles are essential for capturing intrinsic uncertainty of weather forecasts, especially at longer lead times. Beyond about 7 days, the ensemble means of ECMWF-ENS and NeuralGCM-ENS forecasts have considerably lower RMSE than the deterministic models, indicating that these models better capture the average of possible weather. A better metric for ensemble models is CRPS, which is a proper scoring rule that is sensitive to full marginal probability distributions 28 . Our stochastic model (NeuralGCM-ENS) running at 1.4° resolution has lower error compared with ECMWF-ENS across almost all variables, lead times and vertical levels for ensemble-mean RMSE, RSMB and CRPS (Fig. 2a,c,e and Supplementary Information section  H ), with similar spatial patterns of skill (Fig. 2b,f ). Like ECMWF-ENS, NeuralGCM-ENS has a spread-skill ratio of approximately one (Fig. 2d ), which is a necessary condition for calibrated forecasts 29 .

An important characteristic of forecasts is their resemblance to realistic weather patterns. Figure 3 shows a case study that illustrates the performance of NeuralGCM on three types of important weather phenomenon: tropical cyclones, atmospheric rivers and the Intertropical Convergence Zone. Figure 3a shows that all the machine-learning models make significantly blurrier forecasts than the source data ERA5 and physics-based ECMWF-HRES forecast, but NeuralCGM-0.7° outperforms the pure machine-learning models, despite its coarser resolution (0.7° versus 0.25° for GraphCast and Pangu). Blurry forecasts correspond to physically inconsistent atmospheric conditions and misrepresent extreme weather. Similar trends hold for other derived variables of meteorological interest (Supplementary Information section  H.2 ). Ensemble-mean predictions, from both NeuralGCM and ECMWF, are closer to ERA5 in an average sense, and thus are inherently smooth at long lead times. In contrast, as shown in Fig. 3 and in Supplementary Information section  H.3 , individual realizations from the ECMWF and NeuralGCM ensembles remain sharp, even at long lead times. Like ECMWF-ENS, NeuralGCM-ENS produces a statistically representative range of future weather scenarios for each weather phenomenon, despite its eight-times-coarser resolution.

All forecasts are initialized at 2020-08-22T12z, chosen to highlight Hurricane Laura, the most damaging Atlantic hurricane of 2020. a , Specific humidity at 700 hPa for 1-day, 5-day and 10-day forecasts over North America and the Northeast Pacific Ocean from ERA5 14 , ECMWF-HRES, NeuralGCM-0.7°, ECMWF-ENS (mean), NeuralGCM-ENS (mean), GraphCast 3 and Pangu 4 . b , Forecasts from individual ensemble members from ECMWF-ENS and NeuralGCM-ENS over regions of interest, including predicted tracks of Hurricane Laura from each of the 50 ensemble members (Supplementary Information section  I.2 ). The track from ERA5 is plotted in black.

We can quantify the blurriness of different forecast models via their power spectra. Supplementary Figs. 17 and 18 show that the power spectra of NeuralCGM-0.7° is consistently closer to ERA5 than the other machine-learning forecast methods, but is still blurrier than ECMWF’s physical forecasts. The spectra of NeuralGCM forecasts is also roughly constant over the forecast period, in stark contrast to GraphCast, which worsens with lead time. The spectrum of NeuralGCM becomes more accurate with increased resolution (Supplementary Fig. 22 ), which suggests the potential for further improvements of NeuralGCM models trained at higher resolutions.

Water budget

In NeuralGCM, advection is handled by the dynamical core, while the machine-learning parameterization models local processes within vertical columns of the atmosphere. Thus, unlike pure machine-learning methods, local sources and sinks can be isolated from tendencies owing to horizontal transport and other resolved dynamics (Supplementary Fig. 3 ). This makes our results more interpretable and facilitates the diagnosis of the water budget. Specifically, we diagnose precipitation minus evaporation (Supplementary Information section  H.5 ) rather than directly predicting these as in machine-learning-based approaches 3 . For short weather forecasts, the mean of precipitation minus evaporation has a realistic spatial distribution that is very close to ERA5 data (Extended Data Fig. 4c–e ). The precipitation-minus-evaporation rate distribution of NeuralGCM-0.7° closely matches the ERA5 distribution in the extratropics (Extended Data Fig. 4b ), although it underestimates extreme events in the tropics (Extended Data Fig. 4a ). It is noted that the current version of NeuralGCM directly predicts tendencies for an atmospheric column, and thus cannot distinguish between precipitation and evaporation.

Geostrophic wind balance

We examined the extent to which NeuralGCM, GraphCast and ECMWF-HRES capture the geostrophic wind balance, the near-equilibrium between the dominant forces that drive large-scale dynamics in the mid-latitudes 30 . A recent study 16 highlighted that Pangu misrepresents the vertical structure of the geostrophic and ageostrophic winds and noted a deterioration at longer lead times. Similarly, we observe that GraphCast shows an error that worsens with lead time. In contrast, NeuralGCM more accurately depicts the vertical structure of the geostrophic and ageostrophic winds, as well as their ratio, compared with GraphCast across various rollouts, when compared against ERA5 data (Extended Data Fig. 3 ). However, ECMWF-HRES still shows a slightly closer alignment to ERA5 data than NeuralGCM does. Within NeuralGCM, the representation of the geostrophic wind’s vertical structure only slightly degrades in the initial few days, showing no noticeable changes thereafter, particularly beyond day 5.

Generalizing to unseen data

Physically consistent weather models should still perform well for weather conditions for which they were not trained. We expect that NeuralGCM may generalize better than machine-learning-only atmospheric models, because NeuralGCM employs neural networks that act locally in space, on individual vertical columns of the atmosphere. To explore this hypothesis, we compare versions of NeuralCGM-0.7° and GraphCast trained to 2017 on 5 years of weather forecasts beyond the training period (2018–2022) in Supplementary Fig. 36 . Unlike GraphCast, NeuralGCM does not show a clear trend of increasing error when initialized further into the future from the training data. To extend this test beyond 5 years, we trained a NeuralGCM-2.8° model using only data before 2000, and tested its skill for over 21 unseen years (Supplementary Fig. 35 ).

Climate simulations

Although our deterministic NeuralGCM models are trained to predict weather up to 3 days ahead, they are generally capable of simulating the atmosphere far beyond medium-range weather timescales. For extended climate simulations, we prescribe historical sea surface temperature (SST) and sea-ice concentration. These simulations feature many emergent phenomena of the atmosphere on timescales from months to decades.

For climate simulations with NeuralGCM, we use 2.8° and 1.4° deterministic models, which are relatively inexpensive to train (Supplementary Information section  G.7 ) and allow us to explore a larger parameter space to find stable models. Previous studies found that running extended simulations with hybrid models is challenging due to numerical instabilities and climate drift 21 . To quantify stability in our selected models, we run multiple initial conditions and report how many of them finish without instability.

Seasonal cycle and emergent phenomena

To assess the capability of NeuralGCM to simulate various aspects of the seasonal cycle, we run 2-year simulations with NeuralGCM-1.4°. for 37 different initial conditions spaced every 10 days for the year 2019. Out of these 37 initial conditions, 35 successfully complete the full 2 years without instability; for case studies of instability, see Supplementary Information section  H.7 , and Supplementary Figs. 26 and 27 . We compare results from NeuralGCM-1.4° for 2020 with ERA5 data and with outputs from the X-SHiELD global cloud-resolving model, which is coupled to an ocean model nudged towards reanalysis 31 . This X-SHiELD run has been used as a target for training machine-learning climate models 24 . For comparison, we evaluate models after regridding predictions to 1.4° resolution. This comparison slightly favours NeuralGCM because NeuralGCM was tuned to match ERA5, but the discrepancy between ERA5 and the actual atmosphere is small relative to model error.

Figure 4a shows the temporal variation of the global mean temperature to 2020, as captured by 35 simulations from NeuralGCM, in comparison with the ERA5 reanalysis and standard climatology benchmarks. The seasonality and variability of the global mean temperature from NeuralGCM are quantitatively similar to those observed in ERA5. The ensemble-mean temperature RMSE for NeuralGCM stands at 0.16 K when benchmarked against ERA5, which is a significant improvement over the climatology’s RMSE of 0.45 K. We find that NeuralGCM accurately simulates the seasonal cycle, as evidenced by metrics such as the annual cycle of the global precipitable water (Supplementary Fig. 30a ) and global total kinetic energy (Supplementary Fig. 30b ). Furthermore, the model captures essential atmospheric dynamics, including the Hadley circulation and the zonal-mean zonal wind (Supplementary Fig. 28 ), as well as the spatial patterns of eddy kinetic energy in different seasons (Supplementary Fig. 31 ), and the distinctive seasonal behaviours of monsoon circulation (Supplementary Fig. 29 ; additional details are provided in Supplementary Information section  I.1 ).

a , Global mean temperature for ERA5 14 (orange), 1990–2019 climatology (black) and NeuralGCM-1.4° (blue) for 2020 using 35 simulations initialized every 10 days during 2019 (thick line, ensemble mean; thin lines, different initial conditions). b , Yearly global mean temperature for ERA5 (orange), mean over 22 CMIP6 AMIP experiments 34 (violet; model details are in Supplementary Information section  I.3 ) and NeuralGCM-2.8° for 22 AMIP-like simulations with prescribed SST initialized every 10 days during 1980 (thick line, ensemble mean; thin lines, different initial conditions). c , The RMSB of the 850-hPa temperature averaged between 1981 and 2014 for 22 NeuralGCM-2.8° AMIP runs (labelled NGCM), 22 CMIP6 AMIP experiments (labelled AMIP) and debiased 22 CMIP6 AMIP experiments (labelled AMIP*; bias was removed by removing the 850-hPa global temperature bias). In the box plots, the red line represents the median. The box delineates the first to third quartiles; the whiskers extend to 1.5 times the interquartile range (Q1 − 1.5IQR and Q3 + 1.5IQR), and outliers are shown as individual dots. d , Vertical profiles of tropical (20° S–20° N) temperature trends for 1981–2014. Orange, ERA5; black dots, Radiosonde Observation Correction using Reanalyses (RAOBCORE) 41 ; blue dots, mean trends for NeuralGCM; purple dots, mean trends from CMIP6 AMIP runs (grey and black whiskers, 25th and 75th percentiles for NeuralGCM and CMIP6 AMIP runs, respectively). e – g , Tropical cyclone tracks for ERA5 ( e ), NeuralGCM-1.4° ( f ) and X-SHiELD 31 ( g ). h – k , Mean precipitable water for ERA5 ( h ) and the precipitable water bias in NeuralGCM-1.4° ( i ), initialized 90 days before mid-January 2020 similarly to X-SHiELD, X-SHiELD ( j ) and climatology ( k ; averaged between 1990 and 2019). In d – i , quantities are calculated between mid-January 2020 and mid-January 2021 and all models were regridded to a 256 × 128 Gaussian grid before computation and tracking.

Next, we compare the annual biases of a single NeuralGCM realization with a single realization of X-SHiELD (the only one available), both initiated in mid-October 2019. We consider 19 January 2020 to 17 January 2021, the time frame for which X-SHiELD data are available. Global cloud-resolving models, such as X-SHiELD, are considered state of the art, especially for simulating the hydrological cycle, owing to their resolution being capable of resolving deep convection 32 . The annual bias in precipitable water for NeuralGCM (RMSE of 1.09 mm) is substantially smaller than the biases of both X-SHiELD (RMSE of 1.74 mm) and climatology (RMSE of 1.36 mm; Fig. 4i–k ). Moreover, NeuralGCM shows a lower temperature bias in the upper and lower troposphere than X-SHiELD (Extended Data Fig. 6 ). We also indirectly compare precipitation bias in X-SHiELD with precipitation-minus-evaporation bias in NeuralGCM-1.4°, which shows slightly larger bias and grid-scale artefacts for NeuralGCM (Extended Data Fig. 5 ).

Finally, to assess the capability of NeuralGCM to generate tropical cyclones in an annual model integration, we use the tropical cyclone tracker TempestExtremes 33 , as described in Supplementary Information section   I.2 , Supplementary Fig. 34 and Supplementary Table 6 . Figure 4e–g shows that NeuralGCM, even at a coarse resolution of 1.4°, produces realistic trajectories and counts of tropical cyclone (83 versus 86 in ERA5 for the corresponding period), whereas X-SHiELD, when regridded to 1.4° resolution, substantially underestimates the tropical cyclone count (40). Additional statistical analyses of tropical cyclones can be found in Extended Data Figs. 7 and 8 .

Decadal simulations

To assess the capability of NeuralGCM to simulate historical temperature trends, we conduct AMIP-like simulations over a duration of 40 years with NeuralGCM-2.8°. Out of 37 different runs with initial conditions spaced every 10 days during the year 1980, 22 simulations were stable for the entire 40-year period, and our analysis focuses on these results. We compare with 22 simulations run with prescribed SST from the Coupled Model Intercomparison Project Phase 6 (CMIP6) 34 , listed in Supplementary Information section  I.3 .

We find that all 40-year simulations of NeuralGCM, as well as the mean of the 22 AMIP runs, accurately capture the global warming trends observed in ERA5 data (Fig. 4b ). There is a strong correlation in the year-to-year temperature trends with ERA5 data, suggesting that NeuralGCM effectively captures the impact of SST forcing on climate. When comparing spatial biases averaged over 1981–2014, we find that all 22 NeuralGCM-2.8° runs have smaller bias than the CMIP6 AMIP runs, and this result remains even when removing the global temperature bias in CMIP6 AMIP runs (Fig. 4c and Supplementary Figs. 32 and 33 ).

Next, we investigated the vertical structure of tropical warming trends, which climate models tend to overestimate in the upper troposphere 35 . As shown in Fig. 4d , the trends, calculated by linear regression, of NeuralGCM are closer to ERA5 than those of AMIP runs. In particular, the bias in the upper troposphere is reduced. However, NeuralGCM does show a wider spread in its predictions than the AMIP runs, even at levels near the surface where temperatures are typically more constrained by prescribed SST.

Lastly, we evaluated NeuralGCM’s capability to generalize to unseen warmer climates by conducting AMIP simulations with increased SST (Supplementary Information section  I.4.2 ). We find that NeuralGCM shows some of the robust features of climate warming response to modest SST increases (+1 K and +2 K); however, for more substantial SST increases (+4 K), NeuralGCM’s response diverges from expectations (Supplementary Fig. 37 ). In addition, AMIP simulations with increased SST show climate drift, underscoring NeuralGCM’s limitations in this context (Supplementary Fig. 38 ).

NeuralGCM is a differentiable hybrid atmospheric model that combines the strengths of traditional GCMs with machine learning for weather forecasting and climate simulation. To our knowledge, NeuralGCM is the first machine-learning-based model to make accurate ensemble weather forecasts, with better CRPS than state-of-the-art physics-based models. It is also, to our knowledge, the first hybrid model that achieves comparable spatial bias to global cloud-resolving models, can simulate realistic tropical cyclone tracks and can run AMIP-like simulations with realistic historical temperature trends. Overall, NeuralGCM demonstrates that incorporating machine learning is a viable alternative to building increasingly detailed physical models 32 for improving GCMs.

Compared with traditional GCMs with similar skill, NeuralGCM is computationally efficient and low complexity. NeuralGCM runs at 8- to 40-times-coarser horizontal resolution than ECMWF’s Integrated Forecasting System and global cloud-resolving models, which enables 3 to 5 orders of magnitude savings in computational resources. For example, NeuralGCM-1.4° simulates 70,000 simulation days in 24 hours using a single tensor-processing-unit versus 19 simulated days on 13,824 central-processing-unit cores with X-SHiELD (Extended Data Table 1 ). This can be leveraged for previously impractical tasks such as large ensemble forecasting. NeuralGCM’s dynamical core uses global spectral methods 36 , and learned physics is parameterized with fully connected neural networks acting on single vertical columns. Substantial headroom exists to pursue higher accuracy using advanced numerical methods and machine-learning architectures.

Our results provide strong evidence for the disputed hypothesis 37 , 38 , 39 that learning to predict short-term weather is an effective way to tune parameterizations for climate. NeuralGCM models trained on 72-hour forecasts are capable of realistic multi-year simulation. When provided with historical SSTs, they capture essential atmospheric dynamics such as seasonal circulation, monsoons and tropical cyclones. However, we will probably need alternative training strategies 38 , 39 to learn important processes for climate with subtle impacts on weather timescales, such as a cloud feedback.

The NeuralGCM approach is compatible with incorporating either more physics or more machine learning, as required for operational weather forecasts and climate simulations. For weather forecasting, we expect that end-to-end learning 40 with observational data will allow for better and more relevant predictions, including key variables such as precipitation. Such models could include neural networks acting as corrections to traditional data assimilation and model diagnostics. For climate projection, NeuralGCM will need to be reformulated to enable coupling with other Earth-system components (for example, ocean and land), and integrating data on the atmospheric chemical composition (for example, greenhouse gases and aerosols). There are also research challenges common to current machine-learning-based climate models 19 , including the capability to simulate unprecedented climates (that is, generalization), adhering to physical constraints, and resolving numerical instabilities and climate drift. NeuralGCM’s flexibility to incorporate physics-based models (for example, radiation) offers a promising avenue to address these challenges.

Models based on physical laws and empirical relationships are ubiquitous in science. We believe the differentiable hybrid modelling approach of NeuralGCM has the potential to transform simulation for a wide range of applications, such as materials discovery, protein folding and multiphysics engineering design.

Differentiable atmospheric model

NeuralGCM combines components of the numerical solver and flexible neural network parameterizations. Simulation in time is carried out in a coordinate system suitable for solving the dynamical equations of the atmosphere, describing large-scale fluid motion and thermodynamics under the influence of gravity and the Coriolis force.

Our differentiable dynamical core is implemented in JAX, a library for high-performance code in Python that supports automatic differentiation 42 . The dynamical core solves the hydrostatic primitive equations with moisture, using a horizontal pseudo-spectral discretization and vertical sigma coordinates 36 , 43 . We evolve seven prognostic variables: vorticity and divergence of horizontal wind, temperature, surface pressure, and three water species (specific humidity, and specific ice and liquid cloud water content).

Our learned physics module uses the single-column approach of GCMs 2 , whereby information from only a single atmospheric column is used to predict the impact of unresolved processes occurring within that column. These effects are predicted using a fully connected neural network with residual connections, with weights shared across all atmospheric columns (Supplementary Information section  C.4 ).

The inputs to the neural network include the prognostic variables in the atmospheric column, total incident solar radiation, sea-ice concentration and SST (Supplementary Information section  C.1 ). We also provide horizontal gradients of the prognostic variables, which we found improves performance 44 . All inputs are standardized to have zero mean and unit variance using statistics precomputed during model initialization. The outputs are the prognostic variable tendencies scaled by the fixed unconditional standard deviation of the target field (Supplementary Information section  C.5 ).

To interface between ERA5 14 data stored in pressure coordinates and the sigma coordinate system of our dynamical core, we introduce encoder and decoder components (Supplementary Information section  D ). These components perform linear interpolation between pressure levels and sigma coordinate levels. We additionally introduce learned corrections to both encoder and decoder steps (Supplementary Figs. 4–6 ), using the same column-based neural network architecture as the learned physics module. Importantly, the encoder enables us to eliminate the gravity waves from initialization shock 45 , which otherwise contaminate forecasts.

Figure 1a shows the sequence of steps that NeuralGCM takes to make a forecast. First, it encodes ERA5 data at t  =  t 0 on pressure levels to initial conditions on sigma coordinates. To perform a time step, the dynamical core and learned physics (Fig. 1b ) then compute tendencies, which are integrated in time using an implicit–explicit ordinary differential equation solver 46 (Supplementary Information section  E and Supplementary Table 2 ). This is repeated to advance the model from t  =  t 0 to t  =  t final . Finally, the decoder converts predictions back to pressure levels.

The time-step size of the ODE solver (Supplementary Table 3 ) is limited by the Courant–Friedrichs–Lewy condition on dynamics, and can be small relative to the timescale of atmospheric change. Evaluating learned physics is approximately 1.5 times as expensive as a time step of the dynamical core. Accordingly, following the typical practice for GCMs, we hold learned physics tendencies constant for multiple ODE time steps to reduce computational expense, typically corresponding to 30 minutes of simulation time.

Deterministic and stochastic models

We train deterministic NeuralGCM models using a combination of three loss functions (Supplementary Information section  G.4 ) to encourage accuracy and sharpness while penalizing bias. During the main training phase, all losses are defined in a spherical harmonics basis. We use a standard mean squared error loss for prompting accuracy, modified to progressively filter out contributions from higher total wavenumbers at longer lead times (Supplementary Fig. 8 ). This filtering approach tackles the ‘double penalty problem’ 47 as it prevents the model from being penalized for predicting high-wavenumber features in incorrect locations at later times, especially beyond the predictability horizon. A second loss term encourages the spectrum to match the training data using squared loss on the total wavenumber spectrum of prognostic variables. These first two losses are evaluated on both sigma and pressure levels. Finally, a third loss term discourages bias by adding mean squared error on the batch-averaged mean amplitude of each spherical harmonic coefficient. For analysis of the impact that various loss functions have, refer to Supplementary Information section  H.6.1 , and Supplementary Figs. 23 and 24 . The combined action of the three training losses allow the resulting models trained on 3-day rollouts to remain stable during years-to-decades-long climate simulations. Before final evaluations, we perform additional fine-tuning of just the decoder component on short rollouts of 24 hours (Supplementary Information section  G.5 ).

Stochastic NeuralGCM models incorporate inherent randomness in the form of additional random fields passed as inputs to neural network components. Our stochastic loss is based on the CRPS 28 , 48 , 49 . CRPS consists of mean absolute error that encourages accuracy, balanced by a similar term that encourages ensemble spread. For each variable we use a sum of CRPS in grid space and CRPS in the spherical harmonic basis below a maximum cut-off wavenumber (Supplementary Information section  G.6 ). We compute CRPS on rollout lengths from 6 hours to 5 days. As illustrated in Fig. 1 , we inject noise to the learned encoder and the learned physics module by sampling from Gaussian random fields with learned spatial and temporal correlation (Supplementary Information section  C.2 and Supplementary Fig. 2 ). For training, we generate two ensemble members per forecast, which suffices for an unbiased estimate of CRPS.

Data availability

For training and evaluating the NeuralGCM models, we used the publicly available ERA5 dataset 14 , originally downloaded from https://cds.climate.copernicus.eu/ and available via Google Cloud Storage in Zarr format at gs://gcp-public-data-arco-era5/ar/full_37-1h-0p25deg-chunk-1.zarr-v3. To compare NeuralGCM with operational and data-driven weather models, we used forecast datasets distributed as part of WeatherBench2 12 at https://weatherbench2.readthedocs.io/en/latest/data-guide.html , to which we have added NeuralGCM forecasts for 2020. To compare NeuralGCM with atmospheric models in climate settings, we used CMIP6 data available at https://catalog.pangeo.io/browse/master/climate/ , as well as X-SHiELD 24 outputs available on Google Cloud storage in a ‘requester pays’ bucket at gs://ai2cm-public-requester-pays/C3072-to-C384-res-diagnostics. The Radiosonde Observation Correction using Reanalyses (RAOBCORE) V1.9 that was used as reference tropical temperature trends was downloaded from https://webdata.wolke.img.univie.ac.at/haimberger/v1.9/ . Base maps use freely available data from https://www.naturalearthdata.com/downloads/ .

Code availability

The NeuralGCM code base is separated into two open source projects: Dinosaur and NeuralGCM, both publicly available on GitHub at https://github.com/google-research/dinosaur (ref. 50 ) and https://github.com/google-research/neuralgcm (ref. 51 ). The Dinosaur package implements a differentiable dynamical core used by NeuralGCM, whereas the NeuralGCM package provides machine-learning models and checkpoints of trained models. Evaluation code for NeuralGCM weather forecasts is included in WeatherBench2 12 , available at https://github.com/google-research/weatherbench2 (ref. 52 ).

Bauer, P., Thorpe, A. & Brunet, G. The quiet revolution of numerical weather prediction. Nature 525 , 47–55 (2015).

Article   ADS   CAS   PubMed   Google Scholar  

Balaji, V. et al. Are general circulation models obsolete? Proc. Natl Acad. Sci. USA 119 , e2202075119 (2022).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Lam, R. et al. Learning skillful medium-range global weather forecasting. Science 382 , 1416–1421 (2023).

Article   ADS   MathSciNet   CAS   PubMed   Google Scholar  

Bi, K. et al. Accurate medium-range global weather forecasting with 3D neural networks. Nature 619 , 533–538 (2023).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Hourdin, F. et al. The art and science of climate model tuning. Bull. Am. Meteorol. Soc. 98 , 589–602 (2017).

Article   ADS   Google Scholar  

Bony, S. & Dufresne, J.-L. Marine boundary layer clouds at the heart of tropical cloud feedback uncertainties in climate models. Geophys. Res. Lett. 32 , L20806 (2005).

Webb, M. J., Lambert, F. H. & Gregory, J. M. Origins of differences in climate sensitivity, forcing and feedback in climate models. Clim. Dyn. 40 , 677–707 (2013).

Article   Google Scholar  

Sherwood, S. C., Bony, S. & Dufresne, J.-L. Spread in model climate sensitivity traced to atmospheric convective mixing. Nature 505 , 37–42 (2014).

Article   ADS   PubMed   Google Scholar  

Palmer, T. & Stevens, B. The scientific challenge of understanding and estimating climate change. Proc. Natl Acad. Sci. USA 116 , 24390–24395 (2019).

Fischer, E. M., Beyerle, U. & Knutti, R. Robust spatially aggregated projections of climate extremes. Nat. Clim. Change 3 , 1033–1038 (2013).

Field, C. B. Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation: Special Report of the Intergovernmental Panel on Climate Change (Cambridge Univ. Press, 2012).

Rasp, S. et al. WeatherBench 2: A benchmark for the next generation of data-driven global weather models. J. Adv. Model. Earth Syst. 16 , e2023MS004019 (2024).

Keisler, R. Forecasting global weather with graph neural networks. Preprint at https://arxiv.org/abs/2202.07575 (2022).

Hersbach, H. et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 146 , 1999–2049 (2020).

Zhou, L. et al. Toward convective-scale prediction within the next generation global prediction system. Bull. Am. Meteorol. Soc. 100 , 1225–1243 (2019).

Bonavita, M. On some limitations of current machine learning weather prediction models. Geophys. Res. Lett. 51 , e2023GL107377 (2024).

Weyn, J. A., Durran, D. R. & Caruana, R. Improving data-driven global weather prediction using deep convolutional neural networks on a cubed sphere. J. Adv. Model. Earth Syst. 12 , e2020MS002109 (2020).

Watt-Meyer, O. et al. ACE: a fast, skillful learned global atmospheric model for climate prediction. Preprint at https://arxiv.org/abs/2310.02074 (2023).

Bretherton, C. S. Old dog, new trick: reservoir computing advances machine learning for climate modeling. Geophys. Res. Lett. 50 , e2023GL104174 (2023).

Reichstein, M. et al. Deep learning and process understanding for data-driven Earth system science. Nature 566 , 195–204 (2019).

Brenowitz, N. D. & Bretherton, C. S. Spatially extended tests of a neural network parametrization trained by coarse-graining. J. Adv. Model. Earth Syst. 11 , 2728–2744 (2019).

Rasp, S., Pritchard, M. S. & Gentine, P. Deep learning to represent subgrid processes in climate models. Proc. Natl Acad. Sci. USA 115 , 9684–9689 (2018).

Yuval, J. & O’Gorman, P. A. Stable machine-learning parameterization of subgrid processes for climate modeling at a range of resolutions. Nat. Commun. 11 , 3295 (2020).

Kwa, A. et al. Machine-learned climate model corrections from a global storm-resolving model: performance across the annual cycle. J. Adv. Model. Earth Syst. 15 , e2022MS003400 (2023).

Arcomano, T., Szunyogh, I., Wikner, A., Hunt, B. R. & Ott, E. A hybrid atmospheric model incorporating machine learning can capture dynamical processes not captured by its physics-based component. Geophys. Res. Lett. 50 , e2022GL102649 (2023).

Han, Y., Zhang, G. J. & Wang, Y. An ensemble of neural networks for moist physics processes, its generalizability and stable integration. J. Adv. Model. Earth Syst. 15 , e2022MS003508 (2023).

Gelbrecht, M., White, A., Bathiany, S. & Boers, N. Differentiable programming for Earth system modeling. Geosci. Model Dev. 16 , 3123–3135 (2023).

Gneiting, T. & Raftery, A. E. Strictly proper scoring rules, prediction, and estimation. J. Am. Stat. Assoc. 102 , 359–378 (2007).

Article   MathSciNet   CAS   Google Scholar  

Fortin, V., Abaza, M., Anctil, F. & Turcotte, R. Why should ensemble spread match the RMSE of the ensemble mean? J. Hydrometeorol. 15 , 1708–1713 (2014).

Holton, J. R. An introduction to Dynamic Meteorology 5th edn (Elsevier, 2004).

Cheng, K.-Y. et al. Impact of warmer sea surface temperature on the global pattern of intense convection: insights from a global storm resolving model. Geophys. Res. Lett. 49 , e2022GL099796 (2022).

Stevens, B. et al. DYAMOND: the dynamics of the atmospheric general circulation modeled on non-hydrostatic domains. Prog. Earth Planet. Sci. 6 , 61 (2019).

Ullrich, P. A. et al. TempestExtremes v2.1: a community framework for feature detection, tracking, and analysis in large datasets. Geosc. Model Dev. 14 , 5023–5048 (2021).

Eyring, V. et al. Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9 , 1937–1958 (2016).

Mitchell, D. M., Lo, Y. E., Seviour, W. J., Haimberger, L. & Polvani, L. M. The vertical profile of recent tropical temperature trends: persistent model biases in the context of internal variability. Environ. Res. Lett. 15 , 1040b4 (2020).

Bourke, W. A multi-level spectral model. I. Formulation and hemispheric integrations. Mon. Weather Rev. 102 , 687–701 (1974).

Ruiz, J. J., Pulido, M. & Miyoshi, T. Estimating model parameters with ensemble-based data assimilation: a review. J. Meteorol. Soc. Jpn Ser. II 91 , 79–99 (2013).

Schneider, T., Lan, S., Stuart, A. & Teixeira, J. Earth system modeling 2.0: a blueprint for models that learn from observations and targeted high-resolution simulations. Geophys. Res. Lett. 44 , 12–396 (2017).

Schneider, T., Leung, L. R. & Wills, R. C. J. Opinion: Optimizing climate models with process knowledge, resolution, and artificial intelligence. Atmos. Chem. Phys. 24 , 7041–7062 (2024).

Sutskever, I., Vinyals, O. & Le, Q. V. Sequence to sequence learning with neural networks. Adv. Neural Inf. Process. Syst. 27 , 3104–3112 (2014).

Haimberger, L., Tavolato, C. & Sperka, S. Toward elimination of the warm bias in historic radiosonde temperature records—some new results from a comprehensive intercomparison of upper-air data. J. Clim. 21 , 4587–4606 (2008).

Bradbury, J. et al. JAX: composable transformations of Python+NumPy programs. GitHub http://github.com/google/jax (2018).

Durran, D. R. Numerical Methods for Fluid Dynamics: With Applications to Geophysics Vol. 32, 2nd edn (Springer, 2010).

Wang, P., Yuval, J. & O’Gorman, P. A. Non-local parameterization of atmospheric subgrid processes with neural networks. J. Adv. Model. Earth Syst. 14 , e2022MS002984 (2022).

Daley, R. Normal mode initialization. Rev. Geophys. 19 , 450–468 (1981).

Whitaker, J. S. & Kar, S. K. Implicit–explicit Runge–Kutta methods for fast–slow wave problems. Mon. Weather Rev. 141 , 3426–3434 (2013).

Gilleland, E., Ahijevych, D., Brown, B. G., Casati, B. & Ebert, E. E. Intercomparison of spatial forecast verification methods. Weather Forecast. 24 , 1416–1430 (2009).

Rasp, S. & Lerch, S. Neural networks for postprocessing ensemble weather forecasts. Month. Weather Rev. 146 , 3885–3900 (2018).

Pacchiardi, L., Adewoyin, R., Dueben, P. & Dutta, R. Probabilistic forecasting with generative networks via scoring rule minimization. J. Mach. Learn. Res. 25 , 1–64 (2024).

Smith, J. A., Kochkov, D., Norgaard, P., Yuval, J. & Hoyer, S. google-research/dinosaur: 1.0.0. Zenodo https://doi.org/10.5281/zenodo.11376145 (2024).

Kochkov, D. et al. google-research/neuralgcm: 1.0.0. Zenodo https://doi.org/10.5281/zenodo.11376143 (2024).

Rasp, S. et al. google-research/weatherbench2: v0.2.0. Zenodo https://doi.org/10.5281/zenodo.11376271 (2023).

Download references

Acknowledgements

We thank A. Kwa, A. Merose and K. Shah for assistance with data acquisition and handling; L. Zepeda-Núñez for feedback on the paper; and J. Anderson, C. Van Arsdale, R. Chemke, G. Dresdner, J. Gilmer, J. Hickey, N. Lutsko, G. Nearing, A. Paszke, J. Platt, S. Ponda, M. Pritchard, D. Rothenberg, F. Sha, T. Schneider and O. Voicu for discussions.

Author information

These authors contributed equally: Dmitrii Kochkov, Janni Yuval, Ian Langmore, Peter Norgaard, Jamie Smith, Stephan Hoyer

Authors and Affiliations

Google Research, Mountain View, CA, USA

Dmitrii Kochkov, Janni Yuval, Ian Langmore, Peter Norgaard, Jamie Smith, Griffin Mooers, James Lottes, Stephan Rasp, Michael P. Brenner & Stephan Hoyer

Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA

Milan Klöwer

European Centre for Medium-Range Weather Forecasts, Reading, UK

Peter Düben & Sam Hatfield

Google DeepMind, London, UK

Peter Battaglia, Alvaro Sanchez-Gonzalez & Matthew Willson

School of Engineering and Applied Sciences, Harvard University, Cambridge, MA, USA

Michael P. Brenner

You can also search for this author in PubMed   Google Scholar

Contributions

D.K., J.Y., I.L., P.N., J.S. and S. Hoyer contributed equally to this work. D.K., J.Y., I.L., P.N., J.S., G.M., J.L. and S. Hoyer wrote the code. D.K., J.Y., I.L., P.N., G.M. and S. Hoyer trained models and analysed the data. M.P.B. and S. Hoyer managed and oversaw the research project. M.K., S.R., P.D., S. Hatfield, P.B. and M.P.B. contributed technical advice and ideas. M.W. ran experiments with GraphCast for comparison with NeuralGCM. A.S.-G. assisted with data preparation. D.K., J.Y., I.L., P.N. and S. Hoyer wrote the paper. All authors gave feedback and contributed to editing the paper.

Corresponding authors

Correspondence to Dmitrii Kochkov , Janni Yuval or Stephan Hoyer .

Ethics declarations

Competing interests.

D.K., J.Y., I.L., P.N., J.S., J.L., S.R., P.B., A.S.-G., M.W., M.P.B. and S. Hoyer are employees of Google. S. Hoyer, D.K., I.L., J.Y., G.M., P.N., J.S. and M.B. have filed international patent application PCT/US2023/035420 in the name of Google LLC, currently pending, relating to neural general circulation models.

Peer review

Peer review information.

Nature thanks Karthik Kashinath and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended data fig. 1 maps of bias for neuralgcm-ens and ecmwf-ens forecasts..

Bias is averaged over all forecasts initialized in 2020.

Extended Data Fig. 2 Maps of spread-skill ratio for NeuralGCM-ENS and ECMWF-ENS forecasts.

Spread-skill ratio is averaged over all forecasts initialized in 2020.

Extended Data Fig. 3 Geostrophic balance in NeuralGCM, GraphCast 3 and ECMWF-HRES.

Vertical profiles of the extratropical intensity (averaged between latitude 30°–70° in both hemispheres) and over all forecasts initialized in 2020 of (a,d,g) geostrophic wind, (b,e,h) ageostrophic wind and (c,f,i) the ratio of the intensity of ageostrophic wind over geostrophic wind for ERA5 (black continuous line in all panels), (a,b,c) NeuralGCM-0.7°, (d,e,f) GraphCast and (g,h,i) ECMWF-HRES at lead times of 1 day, 5 days and 10 days.

Extended Data Fig. 4 Precipitation minus evaporation calculated from the third day of weather forecasts.

(a) Tropical (latitudes −20° to 20°) precipitation minus evaporation (P minus E) rate distribution, (b) Extratropical (latitudes 30° to 70° in both hemispheres) P minus E, (c) mean P minus E for 2020 ERA5 14 and (d) NeuralGCM-0.7° (calculated from the third day of forecasts and averaged over all forecasts initialized in 2020), (e) the bias between NeuralGCM-0.7° and ERA5, (f-g) Snapshot of daily precipitation minus evaporation for 2020-01-04 for (f) NeuralGCM-0.7° (forecast initialized on 2020-01-02) and (g) ERA5.

Extended Data Fig. 5 Indirect comparison between precipitation bias in X-SHiELD and precipitation minus evaporation bias in NeuralGCM-1.4°.

Mean precipitation calculated between 2020-01-19 and 2021-01-17 for (a) ERA5 14 (c) X-SHiELD 31 and the biases in (e) X-SHiELD and (g) climatology (ERA5 data averaged over 1990-2019). Mean precipitation minus evaporation calculated between 2020-01-19 and 2021-01-17 for (b) ERA5 (d) NeuralGCM-1.4° (initialized in October 18th 2019) and the biases in (f) NeuralGCM-1.4° and (h) climatology (data averaged over 1990–2019).

Extended Data Fig. 6 Yearly temperature bias for NeuralGCM and X-SHiELD 31 .

Mean temperature between 2020-01-19 to 2020-01-17 for (a) ERA5 at 200hPa and (b) 850hPa. (c,d) the bias in the temperature for NeuralGCM-1.4°, (e,f) the bias in X-SHiELD and (g,h) the bias in climatology (calculated from 1990–2019). NeuralGCM-1.4° was initialized in 18th of October (similar to X-SHiELD).

Extended Data Fig. 7 Tropical Cyclone densities and annual regional counts.

(a) Tropical Cyclone (TC) density from ERA5 14 data spanning 1987–2020. (b) TC density from NeuralGCM-1.4° for 2020, generated using 34 different initial conditions all initialized in 2019. (c) Box plot depicting the annual number of TCs across different regions, based on ERA5 data (1987–2020), NeuralGCM-1.4° for 2020 (34 initial conditions), and orange markers show ERA5 for 2020. In the box plots, the red line represents the median; the box delineates the first to third quartiles; the whiskers extend to 1.5 times the interquartile range (Q1 − 1.5IQR and Q3 + 1.5IQR), and outliers are shown as individual dots. Each year is defined from January 19th to January 17th of the following year, aligning with data availability from X-SHiELD. For NeuralGCM simulations, the 3 initial conditions starting in January 2019 exclude data for January 17th, 2021, as these runs spanned only two years.

Extended Data Fig. 8 Tropical Cyclone maximum wind distribution in NeuralGCM vs. ERA5 14 .

Number of Tropical Cyclones (TCs) as a function of maximum wind speed at 850hPa across different regions, based on ERA5 data (1987–2020; in orange), and NeuralGCM-1.4° for 2020 (34 initial conditions; in blue). Each year is defined from January 19th to January 17th of the following year, aligning with data availability from X-SHiELD. For NeuralGCM simulations, the 3 initial conditions starting in January 2019 exclude data for January 17th, 2021, as these runs spanned only two years.

Supplementary information

Supplementary information.

Supplementary Information (38 figures, 6 tables): (A) Lines of code in atmospheric models; (B) Dynamical core of NeuralGCM; (C) Learned physics of NeuralGCM; (D) Encoder and decoder of NeuralGCM; (E) Time integration; (F) Evaluation metrics; (G) Training; (H) Additional weather evaluations; (I) Additional climate evaluations.

Peer Review File

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Kochkov, D., Yuval, J., Langmore, I. et al. Neural general circulation models for weather and climate. Nature 632 , 1060–1066 (2024). https://doi.org/10.1038/s41586-024-07744-y

Download citation

Received : 13 November 2023

Accepted : 15 June 2024

Published : 22 July 2024

Issue Date : 29 August 2024

DOI : https://doi.org/10.1038/s41586-024-07744-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Google ai predicts long-term climate trends and weather — in minutes.

• Helena Kudiabor

Nature (2024)

Weather and climate predicted accurately — without using a supercomputer

• Oliver Watt-Meyer

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: AI and Robotics newsletter — what matters in AI and robotics research, free to your inbox weekly.