Solar Radiation Modification: an additional tool to fight global warming?
As highlighted again by the most recent IPCC report [IPCC AR6], climate change is an unprecedented threat. With every day of continued emissions and with every tenth of a degree of additional warming, more harm is done to people and ecosystems, and the risk of hitting tipping points is growing. Climate change affects people in all regions of the world, and is impeding progress towards all UN sustainable development goals, including the eradication of poverty, inequality, and injustice.
Eliminating emissions and removing carbon dioxide from the atmosphere (together “mitigation”), and reducing damage through adaptation to climate risks and impacts, are currently accepted tools for addressing global warming and its consequences. But despite significant efforts and progress, there is no guarantee that mitigation strategies will be sufficiently effective to keep global warming below 1.5 or even 2ºC (see below). This realisation has led to increasing interest in additional tools to reduce global warming and its impacts, such as solar radiation modification (SRM).
SRM aims to limit warming by reflecting a fraction of the incoming solar radiation, for example by introducing a thin aerosol layer in the higher atmosphere or by brightening clouds. Modelling studies available to date suggest that SRM could deliver rapid cooling to help limit peak global warming to 1.5ºC until emission reductions and carbon dioxide removal reduce greenhouse gas concentrations to tolerable levels. Models also indicate that using SRM to partially offset the assumed future warming could limit change across relevant climate variables, not just temperature, in most world regions.
SRM might provide an auxiliary tool to help reduce climate risk, limit suffering, lessen ecosystem degradation and improve the chances of sustainable development, but SRM is far from perfect. It does not address the root cause of global warming and is expected to come with its own set of societal and environmental risks and problems. Because of this, SRM is highly controversial. Here we argue that nonetheless, thorough and critical research on SRM is a safer path than willfully neglecting it.
Three concerns about climate mitigation call for investigating SRM
Mitigation is humanity’s safest and most powerful means of fighting global warming, and the only way to address its root cause. SRM can at best complement mitigation and adaptation. However, even if humanity implements ambitious efforts in mitigation and adaptation, this may not suffice to prevent severe climate impacts and suffering. We have the following three concerns:
Removal of CO2 from the atmosphere may not be achieved at sufficient scale and speed. Recent IPCC scenarios compatible with the Paris Agreement goal of limiting warming to 1.5ºC rely on future technologies to remove CO2 from the atmosphere, achieving global net-negative emissions by 2060. As the necessary techniques are still under development, it is uncertain whether they can be deployed in time and with sufficient intensity.
The climate may react more strongly to greenhouse gases than expected. There is persisting uncertainty about the amount of warming resulting from greenhouse gases (“climate sensitivity”). Even the most ambitious IPCC scenario mentioned above will lead to more than 1.5ºC warming if climate sensitivity is on the high end of current estimates.
Limiting global warming to 1.5ºC may not be enough to prevent serious damage. There is no certainty that severe impacts can be avoided even at this moderate warming level. Adaptation may not suffice to prevent losses and suffering, especially if ecosystems or parts of the climate system cross tipping points. But every tenth of a degree of warming we prevent, reduces the probability of disastrous outcomes.
All three concerns have considerable probability, hence the world must be prepared for the possibility that one or more of them prove true. If so, even ambitious decarbonisation can not prevent devastating climate impacts, which would disproportionately affect developing countries, induce injustice, and instigate unprecedented governance challenges. And if the implementation of mitigation strategies lags behind the ambitious IPCC scenarios, climate risks will increase even more.
Hence, while SRM is risky, so is rejecting it. Ignoring either type of risk would distort judgment. We therefore recommend a “Golden Rule” of assessing SRM: The risks of researching, developing and possibly implementing SRM must be balanced against the climate risks SRM would attenuate [C2G 2022, Sovacool et al. 2022]. Given the severe risks of climate change, disregarding SRM may have significant repercussions for future generations. Rejecting SRM is therefore not the obvious default option, but a choice which needs to be scrutinized both scientifically and ethically, in the same way as pursuing SRM needs scrutiny.
Research and transparent assessment is needed to lay the groundwork for future decision making
If one or more of the aforementioned concerns materializes, there is a significant chance that a decision on the use of SRM has to be made in the future, especially if the (perceived) pressure from climate impact calls for fast and drastic action. Should such a situation arise, ignorance would add to the risk of misguided decisions and compromise the legitimacy of the decision-making process.
Pursuing SRM in spite of lack of knowledge bears a risk of ineffective or harmful implementation strategies. Rejecting SRM a priori would deprive humanity of a potential auxiliary tool against climate change impacts. Rejecting SRM is the correct decision if no beneficial SRM implementation strategy exists, but it might be difficult to uphold this decision under pressure without sufficient evidence for the harmfulness of SRM. Hence, regardless of whether SRM is beneficial or detrimental, ignorance jeopardizes rational, balanced, justifiable decision making.
Currently, humanity is not well prepared for a possible decision on SRM. Only 0.2% of the climate research funding is presently spent on research on SRM, and grave knowledge gaps remain. Our generation has the chance and capability to perform responsible and critical SRM research in time to gain the knowledge needed for sound decision making. These research efforts should include thorough, impartial, interdisciplinary, and internationally legitimized assessment by international bodies such as the IPCC.
SRM research and ethics
Besides the environmental risks associated with SRM deployment, SRM research is also not without risks, especially on ethical, societal and political levels. Three main fears are raised against SRM.
Delayed decarbonisation. The fear that development of SRM, or even simply knowing that SRM is being considered, will lead to less ambitious mitigation efforts.
Sleepwalking into implementation. The fear that researching and discussing SRM will almost inevitably lead to development and eventually implementation, e.g. by shifting value judgment or establishing interest groups.
Undemocratic decision-making and governance. The fear that powerful actors, such as rich and influential nations, might impose decisions on SRM in the rest of the world. This would violate procedural justice and, given the global effects of SRM, raises concerns that groups excluded from the decision-making process would be subject to unfavorable outcomes.
These are important concerns, but they may not be inevitable. More importantly, these risks are not avoided by rejecting SRM research and, in our view, do not justify ignoring the potential of SRM. In particular, rejecting SRM research will not necessarily prevent future non-inclusive decision-making or unjustified reliance on technological solutions. Instead, SRM research should be conducted in a way that minimizes risks induced by the research itself.
Following a number of scholars who have worked on moral frameworks for SRM [Rayner et al. 2013, NAS report 2021], we propose the following broad principles:
Mitigation (including removals) and adaptation need to be the focus of any climate policy. SRM should at most serve as an addition to reducing greenhouse gas concentrations.
Knowledge and implementation of SRM must be administered in the public interest. This entails that the provision of SRM is organized by a globally legitimized body, and not based on private interests.
Legitimate governance processes must be adhered to, and societal values such as justice and equality must be central when considering the role SRM research can play in lessening the threat of climate change.
Any decision about deployment must be taken on the basis of broad public participation. Special emphasis should be placed on underrepresented and vulnerable communities, such as the so-called “Global South'' and Indigenous Peoples.
The research process should be transparent, reflexive, and cooperative (also on the international level), and provide ample space for off-ramps, in case certain findings point towards undesirable outcomes of SRM deployment.
SRM research must aim to create a comprehensive body of knowledge covering environmental, technical, political, societal and ethical sciences and properly linking and combining these domains.
A solid framework for the governance of SRM should be in place before implementation is seriously considered. This entails engaging in research and consultations on governance parallel to studying the environmental and technological aspects of SRM.
While there exists no complete framework yet to ensure adherence to these principles, the 2021 report of the (US) National Academies of Sciences, Engineering, and Medicine contains a collection of recommendations for scientists, states and the international community to promote fair and balanced SRM research.
The above principles are not intended to be the last word on the ethics of SRM research and deployment. Any ongoing ethical deliberation requires an open mind for criticism, debate and amendment. This is particularly important since the discussion around SRM research has thus far not been sufficiently representative nor global. However, this lack of representation only underscores the need for more inclusive thorough and systematic research into SRM and its impacts. By calling for ethical SRM research, we recognise that ethics has to be a central part of the research process itself.
Given the most recent projections of the IPCC, immediate and decisive action is necessary to reduce the threats of climate change. Rapid emission reduction and carbon dioxide removal are essential to keep and stabilize the climate system in a liveable state, and any further delay will increase climate-induced risks. Meanwhile, Solar Radiation Modification might help to meaningfully reduce climate-induced risks, but it may also introduce new ones. When assessing SRM, all these risks must be fairly balanced. However, knowledge on SRM is as yet insufficient to achieve this.
Given the severity of the climate crisis, there is a significant chance that humanity will eventually have to take a decision in favor of or against the use of SRM, and this possibility will not vanish if we now choose to ignore the issue or prohibit SRM research and assessment. If a choice on the use of SRM has to be made, ignorance increases the risk of inadequate decisions. We believe that society has a moral obligation to engage in SRM research - and to set up this process in such a way as to minimize potential risks stemming from the research itself. We therefore call for international, inclusive, transparent, reflexive and comprehensive research efforts to enable balanced assessment of SRM.
The Paris agreement aims at “holding the increase in the global average temperature to well below 2°C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5°C above pre-industrial levels”. The most ambitious scenario (called “SSP1-1.9”) in the recent IPCC summary for policymakers [p. 13-14 in IPCC AR6 WG1 SPM] assumes net negative CO2 emissions from 2055. The scenario has an expected 2100 warming of 1.4 degrees, with a 90% chance of staying within 1.0 and 1.8°C warming, depending on climate sensitivity. So even this scenario has a significant chance of exceeding 1.5°C warming in 2100, and an almost 50% chance of temporarily reaching or slightly exceeding 1.5°C. The next most ambitious scenario (“SSP1-2.6”), with an expected 2100 warming of 1.8°C (90% chance of staying between 1.4 and 2.4°C) still assumes net negative emissions after 2075. In other words, even the less ambitious end of the Paris agreement requires massive carbon dioxide removal, reaching at least ca ¼ of the current emissions by 2100.
The IPCC's best estimate for climate sensitivity is 3°C warming per doubling CO2, with 90% probability for values between 2 and 5°C [IPCC AR6 WG1 SPM].
Risks of climate change
The recent IPCC report [IPCC AR6 WG2 SPM] gives many examples of adverse effects of climate change, including damage to ecosystems, threats to water availability and food security, and extreme weather impacts. These effects will already be serious at 1.5°C warming but increase significantly in case of further warming.
Some damage may be irreversible even if global mean surface temperature is eventually restored. One example is severe ice loss from the Greenland and West Antarctic ice sheets, each of which bind enough water to cause roughly 7m of sea level rise. Both are believed to have tipping points that may be triggered at 1.5-2°C or slightly above, although the duration of the warm period also matters, as a short overshoot followed by cooling may not destabilize ice sheets due to their long response time [Pattyn et al. 2018].
Climate impacts may – gradually or abruptly – overwhelm the capacity to adapt, for example if physical limits to adaptation are exceeded or if means (e.g. money or skills) for adaptation measures are lacking [Felgenhauer 2015].
Carbon Dioxide Removal
Several methods for Carbon Dioxide Removal (CDR) have been suggested. Some are more industry-based, for example capturing CO2 from point sources (e.g. incinerators of organic waste or fuel-based power plants) or from ambient air by chemical filters, and utilizing it or burying the CO2 in suitable rock formations or the deep sea. Other methods are based on intervening in natural processes, as is the case for SRM. These include adding minerals to the ocean to make it less acidic and enhance its ability to take up atmospheric CO2 (“ocean alkalisation”). There are uncertainties about the total capacity of some methods (for example, the availability of land for afforestation or suitable rock formations for carbon capture may be limited physically or politically), as well as the rate at which CO2 can be taken up (for example, the annual capacity of ambient air capture may depend on available green energy, or the capacity for biomass-based methods on the growth rate of suitable plants) [Lawrence et al. 2018]. In addition, environmental side effects have to be considered, especially for those CDR methods aimed at intervening in natural processes like ocean alkalisation [Gonzales et al. 2016].
Solar Radiation Modification - IPCC definition and potential methods
Solar radiation modification (SRM, also called "solar radiation management", “solar geoengineering” or “climate intervention”) is defined in the glossary of the recent IPCC report [IPCC AR6 WG2] as “a range of radiation modification measures not related to greenhouse gas (GHG) mitigation that seek to limit global warming. Most methods involve reducing the amount of incoming solar radiation reaching the surface, but others also act on the longwave radiation budget [i.e. making it easier for the Earth’s radiation to escape into space] by reducing optical thickness and cloud lifetime.”
There are several potential Solar Radiation Modification methods.
Stratospheric Aerosol Injection (SAI) involves creating a thin, reflective aerosol cloud layer in the higher atmosphere (the stratosphere). SAI is generally seen as the most promising SRM technique because large explosive volcanic eruptions have already proven that the method works in principle.
Marine Cloud Brightening (MCB) involves spraying condensation nuclei into suitable cloud banks to induce them to develop more (but smaller) droplets, thereby making the cloud brighter. Cloud feedback processes are not yet sufficiently understood to be sure that the method can cause significant cooling.
Cirrus Thinning (CT) aims at reducing high thin clouds (cirrus clouds), which have a net warming effect because they reflect only little sunlight but strongly inhibit outgoing longwave radiation, just as greenhouse gasses do.
Space-based SRM, placing reflective sun shields at the point between Earth and Sun where the gravitational pull from both bodies cancels (Lagrange point 1) or in an orbit closer to Earth, might offer a high degree of steerability and relatively little side effects, but is seen as technically extremely challenging.
Surface albedo changes involve making the earth’s surface more reflective, and range from local measures (white roof tops) to brightening the ocean surface with foam or whitening (and hence reducing the melt of) Arctic sea ice.
Partial SRM seems likely to reduce climate change
[Irvine et al. 2019] analyzed several model simulations in which strong global warming was partially prevented by SRM. They found that in nearly all regions key climate variables - mean temperature, peak temperatures, peak rainfall and mean precipitation minus evaporation - were closer to their pre-industrial levels with SRM than without. Exceptions were mostly beneficial (e.g. more rain in South Africa). While obviously it would be preferable to reach a certain temperature target through reducing greenhouse gasses, Irvine’s results suggest that partial SRM can bring (most of) our climate system closer to the “safe” pre-industrial state. For some complex climate risks, including sea level rise, there is however still uncertainty about SRM's effectivenes to reduce them, and more generally, climate models such as those used by Irvine et al. still have uncertainties.
Environmental risks of SRM
Some environmental risks can be caused by inappropriate implementation. For example, performing Stratospheric Aerosol Injection too asymmetrically around the equator (e.g. only on one hemisphere) would cause a shift in tropical rainfall patterns, while an abrupt termination of SRM would cause sudden warming. Even deployment schemes currently deemed appropriate would come with some risks, including a possible delay in the healing of the ozone hole. SRM also will not restore climate variables perfectly: A climate characterized by high levels of greenhouse gasses and of SRM would not equal the pre-industrial state even if global mean surface temperature would be restored to pre-industrial levels. For example, the spatial distribution of temperature or rainfall might change. Nonetheless, research to date suggests that for a given greenhouse gas concentration, partial cooling by SRM would bring the climate closer to its (safe) pre-industrial state than a climate with the same greenhouse gas concentration but without SRM. There is still uncertainty on the maximum amount of cooling that can be reached, as well as SRM’s ability to prevent or reverse long-term effects such as sea level rise (which depends not only on global mean temperature but also, for example, ocean currents around Antarctica). All environmental risks of SRM probably increase with the amount of SRM used. Therefore even if humanity agrees to deploy SRM, it is vital to reduce greenhouse gas emissions swiftly such as to minimize the dosage of SRM required. Other, non-environmental risks of SRM are discussed in the main text under “Research and Ethics”.
0.2% of climate research funding spent on SRM research
The 0.2% estimate is based on an estimated 52 million USD spent on SRM research in the period 2008-2018 [NAS Report 2021], compared to 30 billion USD on all climate research in the same period [Overland & Sovacool 2020].
Uncertainties surrounding SRM
Uncertainties surrounding SRM relate to many aspects, including:
Technical feasibility: Is it possible to build and operate the required infrastructure and manufacture the required material (e.g. aerosol precursors)? For Stratospheric Aerosol Injection (SAI) and Marine Cloud Brightening (MCB), technical feasibility and implementation costs are probably not the limiting factor, but for space-based solutions they probably are.
Effectiveness: Does the method reliably lead to cooling, and what is the maximum cooling achievable? For SAI, volcanic eruptions suggest that significant cooling is achievable, though uncertainty remains about the amount of aerosol needed, and whether there is an upper limit to the possible cooling. For MCB, complex cloud feedbacks are not yet sufficiently understood to be sure that sustained cooling can be achieved at all.
Climate and environmental impact: This includes for example uncertainties on the impact of SRM on precipitation patterns, its effectiveness to reduce sea level rise (which depends not exclusively on cooling), but also side effects of the SRM method itself, e.g. the strength of SAI’s influence on the ozone layer.
Optimal scenarios: Many impacts, for example the influence on precipitation patterns, depend strongly on the implementation scenario (e.g., injection location for SAI). Finding an optimal – or at least globally acceptable – implementation scheme requires assessing multiple SRM scenarios for each greenhouse gas emission scenario, as well as agreeing on criteria to measure the desirability of each scenario.
Policy and Ethics: See main text “Research and Ethics”.
Risks of decision making on SRM under large uncertainty
We do not yet know whether SRM can be implemented in such a way as to be beneficial. But whether SRM is beneficial or harmful, ignorance endangers adequate decision-making, both if SRM is pursued or rejected.
Risks from uninformed decision-making
Loss of powerful tool against global warming, unnecessary exposure to severe climate risks and suffering
Correct decision, but potentially difficult to uphold under pressure (insufficient evidence for SRM’s harmfulness)
Correct decision in principle, but risks of inadequate implementation strategy
Large-scale environmental and political damage
Discussion on ethical concerns
There are several common counterarguments against the delayed-decarbonisation (or “moral hazard”) concern. First, it is not confirmed by polling studies and experiments; respondents confronted with information on SRM are in most cases found more willing to support emission reduction than respondents who did not receive such information, or least not less willing [Merk et al. 2015, Cherry et al. 2020]. However, these studies are not uncontested [Burns et al. 2016] and the preferences of ordinary citizens (respondents in the polls) may not be fully indicative of the choices of decision makers [Reynolds 2020]. Also, if research shows that SRM is infeasible or grossly undesirable, it might actually debunk unfounded hopes of a technofix.
Lock-in dynamics, while possible, are far from inevitable: In fact many budding technology developments never make it to implementation stage [Callies 2019], so research does not automatically lead to implementation. Also, the most difficult part in SRM implementation is not technical feasibility - a powerful government determined on sending up planes for SAI “no matter how” could probably achieve this within years. The major difficulty would be finding an implementation strategy that optimizes benefits and minimizes risks (which might well amount to abstaining from implementation altogether). Timely SRM research, even if it indeed increases the probability of implementation, may well reduce the risk of inappropriate implementation, especially under severe pressure from climate impacts.
The fear of undemocratic decision-making – either nationally or internationally – is arguably the most substantial threat, although the claim that SRM is “inherently” incompatible with democracy, has been refuted [Horton et al. 2018]. The fear is anyhow a questionable argument against SRM research: Rejecting SRM research does not necessarily prevent future undemocratic or unilateral decision-making on SRM. In fact, given the current unequal distribution of SRM knowledge, rejecting further research now might mostly be a decision of privileged actors who currently have the capacity to engage in (or oppose) SRM research [Taiwo et al. 2021], and would thus be an undemocratic act in itself. Transparent research, outreach and capacity building, especially in vulnerable developing countries [DEGREES], can empower citizens and underrepresented regions to take part in the debate and preempt rogue actors from monopolizing SRM knowledge [Smith et al. 2021].
The concern that humanity is not legitimized to perform wanton large-scale interventions in Earth systems cannot be refuted – or proven – as it depends on one’s value judgments. The authors believe that the situation of the climate system, which in itself is the result of large-scale, albeit unintentional, human intervention, is so dire that it justifies considering SRM.
Suggested guidelines for ethical SRM research
The NASEM report [Chapter 5, NAS Report 2021] provides an extensive list of recommendations to promote ethical SRM research. Here we summarize a selection of their suggestions.
Self-Governance by the scientific community: The scientific community should establish and regularly review a code of conduct for scientists and research funding agencies. Such a code of conduct could be inspired by e.g. ethical guidelines for stem cell research.
Transparency: Scientists and their funders should at minimum commit to open-access publishing for all SRM research and to data sharing. A public registry on national or eventually global level could catalog SRM research for better accessibility, similar to WHO’s Human Genome Editing Registry. Standing national or international bodies should perform and publish regular assessments, in particular of risks associated with SRM development and deployment.
Public Engagement: The scientific community should ensure broad public engagement, which includes public outreach and meetings, as well as active participation of stakeholders throughout the research process. This includes equitable and inclusive participation of partners from underrepresented communities, as well as international consultation.
International collaboration: Scientists and agencies involved in SRM research should strive for international collaboration and support currently underrepresented regions in building expertise and capacity to engage in discussions around SRM.
Regulating outdoor experiments: To foster transparency, scientific value and safety, and prevent encroaching commercial interests, (large-scale) field experiments should be regulated by a permit system on a national and potentially international level, for example modeled after the London Protocol for Ocean Iron Fertilisation research.
Non-Commerciality: All persons engaged in SRM research and development should pledge to not submit patents.
Claudia Wieners, Ben Hofbauer, Iris de Vries, Matthias Honegger, Daniele Visioni, Herman Russchenberg, Tyler Felgenhauer