Key points

• Most climate discussion focuses on the 'likely' range of warming, but society should also avoid and prepare for less likely but severe climate scenarios ("tail risks").

• These residual risks can lead to a lot of damage. Small probabilities of disaster can dominate the expected costs of climate change, which is why extreme climate risk deserves attention even if the central forecast looks manageable.

• Reducing greenhouse gas emissions is the most important lever to address these tail risks, but this field gets relatively more resources than some important but more neglected fields. These levers are solar radiation management, the management of natural feedback mechanisms, and adaptation to a hotter, more extreme climate.

• These two fields are strikingly neglected: only a few hundred researchers and advocates work worldwide on avoiding or preparing for extreme climate change, and even fewer work specifically on getting the politics of this aligned to drive government action.

• You can help by researching these risks, building governance for climate interventions, or working to put them on the political agenda. This suits entrepreneurial people, scientists, and those who value the long-term future.

About this issue

Most conversations about climate change focus on the most likely emissions, warming, and damage scenarios, but ignore the non-negligible chance that climate change will be worse than expected. Global warming could be worse than expected for several reasons:

1. We could emit more greenhouse gas emissions than we expect. While the world is rapidly installing renewables, other sectors like steel, cement, or agriculture might be harder or more expensive to decarbonise, especially if climate policies get withdrawn. Emerging economies might choose to develop fossil fuel infrastructure rather than renewable energy. While the IPCC recently dropped its most pessimistic scenario because of climate progress, emissions could still end up higher than expected.

2. Emissions could lead to more warming than we expect. The relationship between greenhouse gas emissions and global warming (climate sensitivity) is not entirely understood. Initial warming from greenhouse gas emissions affects cloud formation, ice sheets, and additional emissions such as methane release from permafrost. 

3. Global warming could be more damaging than we expect. It might be hard to adapt to a hotter world. An inability to adapt will lead to more deaths, a larger health burden, and more poverty. Even at existing warming levels, major systems, like the currents in the Atlantic that drive global heat circulation, are weakening faster than hoped, indicating that we will see more disruption, earlier, than we had thought. This is a major responsibility for governments to address.

Why this issue?

Small chances of extreme outcomes are a large part of the expected impacts of climate change. The economist Martin Weitzman argued that the uncertainty about climate change is fat-tailed: the probability of extreme outcomes falls away slowly rather than vanishing quickly. When that is the case, a small probability of an enormous loss can outweigh the more likely, more moderate outcomes in an expected-value calculation. A study from 2021 found that climate tipping points raise the social cost of carbon (SCC) by 25%, and there's a 10% chance that tipping points can more than double the SCC.

Damages rise faster than temperature. The marginal damage function is increasing, so each additional degree does more harm than the last. Moving from 3.9°C to 4.0°C of warming is far more damaging than moving from 1.9°C to 2.0°C. This convexity is another reason the worst-case pathways carry so much of the expected harm, and why keeping warming away from the high end matters more than the average forecast suggests.

Despite its importance, extreme climate risk is remarkably neglected. By one recent estimate there are merely a few hundred researchers and advocates working on it worldwide, and fewer than a hundred people in advocacy specifically. That is a tiny share of the more than a trillion dollars spent on climate each year, most of which goes to mitigation. Because so few people work on extreme climate risk, additional effort here tends to go further than it would in the crowded parts of climate work. The neglectedness of this work is understandable. Interventions such as solar radiation management have long been something of a taboo within the environmental movement, partly out of a fear that even discussing them would distract from cutting emissions. Two cognitive biases add to the neglect of this issue: present bias leads us to discount risks that lie decades ahead, and the availability heuristic leads us to underrate low-probability, high-impact events we have never actually witnessed. The result is that options that could be necessary to manage severe risks have not been engaged seriously by governments , creating a major mismatch between the scale of the risks (trillions) and the public spending on them (low tens of millions).

This work is especially important in the long term. Most climate scenarios stop at the end of this century, but the lives of people living after 2100 matter too. If humanity survives and flourishes, there could be vastly more people in the future than are alive today, which makes ensuring that civilisation can persist in a stable climate a high-stakes goal. A climate catastrophe that derails human progress would harm not only those alive at the time but everyone who comes after them.

How can we reduce extreme climate risk?

There are four broad ways to reduce extreme climate risk. Cutting greenhouse gas emissions is the foundation, but it is also where the great majority of climate resources already go. The more neglected levers, and therefore the more promising ones on the margin, are solar radiation management, the management of natural feedback mechanisms, and adaptation to a hotter and more extreme climate. We look at each in turn below.

(Meanwhile, we will also ultimately need to remove carbon dioxide from the atmosphere to phase out solar radiation management in the future, which we’ll discuss in a separate upcoming article).

Reducing greenhouse gas emissions

First and foremost, we can reduce the risk of extreme climate change by getting greenhouse gas emissions to zero. The deeper and faster emissions fall, the smaller the tail risk becomes. Other risk management approaches on this page are a supplement to that imperative, not a substitute.

Within mitigation, we think the highest-value work focuses on reducing the 'derailment risk' of the energy transition, so that decarbonisation continues even in a turbulent and crisis-prone world, and on neglected levers such as the protein transition, industrial decarbonisation, and preventing a carbon lock-in. We think that scaling up carbon dioxide removal is also a promising lever, because it offers a way to reach net zero even if some sectors or countries refuse to decarbonise, such as aviation or petrostates.

Mitigation nonetheless attracts far more resources than any other response to extreme climate risk. Fields that reduce the risk of extreme climate change without primarily reducing emissions are more promising on the margin, simply because so few people work on them. The rest of this page focuses on those neglected areas.

Solar radiation management

There are two fundamentally different ways to reduce global warming. One is to lower the amount of greenhouse gas in the atmosphere, which is what mitigation and carbon removal do. The other is to reduce the amount of sunlight the Earth absorbs, which is what solar radiation management does. A useful analogy is a greenhouse on a hot day: you can either take out some of the glass that traps heat, or you can pull a shade across the roof to let less sunlight in. Solar radiation management (SRM) is similar to the shade.

In addition to stopping warming from greenhouse gases, SRM also matters because the planet is dimming. That is in part because ice sheets are melting (ice is very reflective), but also because clouds are changing in ways that are still only partially understood. Further, removing sulfate from global shipping and coal plants – though very good for public health – made the world substantially dimmer too, essentially stopping a form of engineered global cooling by accident. The result is that Earth’s energy imbalance has more than doubled in recent decades. Managing the reflectivity addresses the problem of an increasingly more dim Earth.

There are several proposed methods to do SRM. The most studied is stratospheric aerosol injection, which would release reflective particles high in the atmosphere, imitating the temporary cooling that follows large volcanic eruptions. Others include marine cloud brightening, which makes low ocean clouds more reflective, and cirrus cloud thinning, which would let more heat escape to space. The US National Academies have laid out a careful research and governance agenda for these options.

In one sense, as noted above,  humanity has already been running an inadvertent version of this experiment. For decades, sulphur pollution from ships brightened clouds over the oceans and masked some of the warming from greenhouse gases. When the International Maritime Organization's 2020 rules cut the sulphur content of marine fuel by roughly 80% to reduce air pollution, that masking effect weakened, which appears to have given warming a measurable boost. This is a real-world illustration of both the cooling power of reflective aerosols and the abrupt warming that can follow when they are suddenly removed.

That last point is central to the case against relying on solar radiation management. Unlike carbon removal, it does not reduce emissions or address the underlying cause of warming; it only masks the effect. This arguably creates a moral hazard, because the promise of a cheap shade could weaken the momentum to cut emissions, and it creates the risk of a 'termination shock', where stopping suddenly after years of deployment causes rapid warming at the worst possible moment. 

On the other hand, it seems likely that clean energy growth will continue due to market forces, and that the derailment risk from extreme warming without SRM could be a greater concern. There are also policy proposals to manage the risks of termination shock. There are also unintended physical effects to weigh, such as potential changes to monsoon rainfall and risks to ecosystems. Assessing solar radiation management therefore means comparing its risks against the risks of extreme climate change.

What should be done follows from this. We need to increase our understanding of both the benefits and the harms, so that responsible deployment becomes possible and irresponsible deployment can be avoided or resisted. Just as important, we need a governance system that decides when and how any intervention happens, one that prevents premature or unilateral use but can still act decisively if the alternative is catastrophe.

Management of natural feedback mechanisms

Beyond deliberately reflecting sunlight, a related and even more neglected field concerns the Earth's own feedback systems, which can amplify or weaken global warming. We do not fully understand climate feedback mechanisms, and some feedback mechanisms pose the threat of tipping points, which are critical thresholds that, when crossed, lead to large and accelerating changes in the climate system. The field of work focused on the research, monitoring, and management of the risk that major Earth systems destabilize and cross tipping points is sometimes called climate stabilisation.

Several Earth systems are already showing signs of strain. Reflectivity is declining as ice retreats and pollution falls, so the planet absorbs more sunlight. Ice sheets and glaciers are losing mass in ways that could lock in metres of sea-level rise. The Atlantic Meridional Overturning Circulation (AMOC), the system of ocean currents that regulates much of the Northern Hemisphere's climate, could weaken or shut down sooner than once assumed. And natural systems are starting to become net emitters, as permafrost thaw and the destabilisation of tropical wetlands release greenhouse gases that lie outside direct human control. These feedbacks are the hazards that stabilisation work aims to understand and, where possible, forestall.

The response has several components. Most of the near-term value lies in research and monitoring, so that we can detect dangerous change early and understand which thresholds are closest. Some physical interventions are also being explored, such as glacier stabilisation, which would try to slow the collapse of vulnerable ice sheets, though these remain at an early and uncertain stage. Because the field is so small, a few more people working on this field could improve our understanding of risks valued in the trillions. Durable stabilisation ultimately still depends on returning greenhouse gas concentrations to safe levels through deep emissions cuts and large-scale carbon removal.

Adaptation to extreme climate change

In addition to stabilising warming and the earth's systems, we need to prepare for the chance of a hotter and more extreme climate. That means studying, and building resilience against, the ways in which extreme climate change could trigger catastrophic risks.

Even moderate climate change is likely to lead to droughts, extreme weather, mass migration, heat deaths, and an increased spread of some diseases. Extreme climate change could pose distinct risks. While we don't know all the risks posed by extreme climate change, there are several fields worth studying. Extreme climate change could cause catastrophic infrastructure loss, such as large, long-duration power outages, or it could trigger a multiple breadbasket failure, in which several major agricultural regions fail at once. Organisations such as ALLFED research resilient foods that could feed people even if harvests collapse or sunlight is reduced. Climate stress could also increase the risk of conflict and, in the worst case, contribute to a great-power war, which is one of the main ways climate change could raise existential risk. Understanding these cascading and interacting risks, some of which we cannot yet foresee, is itself an important and neglected form of preparation.

Reasons to work on extreme climate risks

• You think expected-value calculations are the right way to prioritise. Extreme climate risk earns its importance from the tails, so the case for working on it rests on taking low-probability, high-impact outcomes seriously rather than planning only for the central forecast.

• You are entrepreneurial. There are few organisations hiring in this field, so it is a strong fit for people who want to start a new organisation or set up a new research programme rather than slot into an existing role.

• You are good at science and research. Many of the key questions, from tipping points to the effects of solar radiation management, are open scientific problems where careful work can shift the field.

• You are good at putting novel ideas on the political agenda. Much of the bottleneck is governance and political attention, so people who can make unfamiliar issues legible to policymakers are especially valuable.

• You value the long-term future. The stakes of extreme climate risk extend well beyond 2100, so the work appeals particularly to people who place weight on the welfare of future generations.

Reasons not to work on extreme climate risks

• You do not like working on doom scenarios. Though some may see this work as a source of hope, it also requires focusing on what could go wrong. Much of this work involves engaging with catastrophic possibilities, which is not a good fit for everyone.

• You want to see immediate results. The benefits from working on SRM, natural feedback mechanisms, and societal resilience are often diffuse and far in the future. The work offers fewer visible, near-term wins that some other mitigation-focused roles do.

• You place less weight on the long-term future. If you think safeguarding future generations matters less than serving people alive today, other climate priorities may look more compelling.

• You think other existential risks are more neglected. There are other threats to humanity, such as the risks posed by AI and future pandemics, that some argue are more neglected still. If you find that case persuasive, they may be a better use of your effort.

Applying your skills to extreme climate risk

How your expertise fits in

We think a range of backgrounds can contribute to reducing extreme climate risk, since the barriers are scientific, political, and economic at once.

• Earth and atmospheric sciences: researching feedback mechanisms, tipping points, and the effectiveness and side effects of interventions such as solar radiation management.

• Political and advocacy skills: putting extreme climate risk on the political agenda and building the governance frameworks that responsible decision-making would require.

• Economics and other natural sciences: researching the feasibility, costs, and risks of climate stabilisation and solar radiation management, and modelling how they compare with the risks of inaction.

• Governance, law, and international relations: designing institutions that can prevent premature or unilateral deployment of climate interventions while remaining able to act if a catastrophe looms.

• Risk analysis and forecasting: characterising low-probability, high-impact pathways such as multiple breadbasket failure or cascading infrastructure loss, which are currently badly understood.

• Entrepreneurship and operations: founding and running the new organisations and research programmes the field needs, given how few currently exist.

Potential employers

The field is small, so the list of organisations is short and spans nonprofits, research institutes, and government. Options include:

• Reflective is driving public research on SRM and mapping out paths to manage major uncertainties.

• The African Climate Risk Intervention Hub is expanding research capacity across Africa

• India’s Council on Energy, Environment, and Water is leading work on these issue in the subcontinent

• Climate Hub is a global consortium working to build government capacity and move forward research funding at scale.

• The Federation of American Scientists has a growing program on climate interventions and a long track record of managing catastrophic risks.

• SparkClimate focuses on scaling neglected climate solutions

• SilverLining, which works on solar radiation management research

• The Degrees Initiative, which builds solar radiation management research capacity in the Global South.

• SRM360 informs people about solar geoengineering.

• Arete, which works on glacier stabilisation.

• The Alliance for Just Deliberation on Solar Geoengineering (DSG) elevates voices of climate-vulnerable communities in SRM decision-making.

• The Centre for Future Generations (CFG) in Brussels has a programme on SRM governance.

• ALLFED, which researches resilient foods and preparation for global catastrophic food shocks.

• Many major environmental groups, including NRDC, are beginning work on these issues

• Eco-right groups like the Rainey Center and the Foundation on American Innovation are building conservative-focused approaches on these issues.

• Government research programmes, such as the UK's Advanced Research and Invention Agency, which has funded work on climate cooling.

• Universities and research institutes, where much of the underlying earth-system and intervention science is done. Some prominent research centres are the Climate Systems  Engineering Initiative at the University of Chicago, the Centre for the Study of Existential Risk, and the Potsdam Institute for Climate Impact Research (PIK). There is also academic research at departments of Utrecht University, Exeter University, Harvard University, and Cambridge University. (This list is not exhaustive.)

• Private companies like Stardust, though sometimes controversial, are advancing research in the field and may be important drivers of innovation if governance develops for them.

You may also find important work that no organisation is currently doing. In that case, you might consider starting a new nonprofit or research programme, which the field badly needs.

Personalised career advice

Are you interested in working on extreme climate risk but unsure where to start, or weighing it against other climate priorities? The following services can give you tailored advice.

• The Effective Environmentalism Initiative provides free advice to professionals who want to maximise their positive climate impact. Read more about our personalised career advice.

• Probably Good provides free one-to-one career advice for people aiming to have a meaningful impact.

Other ways to help

Donate to impactful nonprofits

Donating to effective organisations is a way to support this work without changing your career. Independent charity evaluators have not yet identified a specific top climate nonprofit for extreme climate risk, in part because the field is so new and small, though Giving Green added solar radiation management governance to its recommended strategies in 2024.

In the meantime, promising organisations working in this space include SilverLining and ALLFED. Because there is no vetted top recommendation yet, we suggest treating these as promising rather than proven, and looking into each organisation's own funding needs before giving.

Volunteer your time

You might be able to contribute skills or time to organisations working on extreme climate risk, especially if you have expertise that small nonprofits find hard to hire, such as operations, research, or fundraising. Given how young the field is, there is no dedicated guide to volunteering in it yet, so the best approach is usually to contact organisations whose work you find promising.

Key questions we're unsure about

• The feasibility of climate stabilisation. We are unsure how effectively humanity could stabilise warming or the earth's major systems, and which interventions, if any, would work as hoped.

• The unintended effects of geoengineering. The side effects of solar radiation management, from monsoon disruption to ecosystem impacts, are still poorly understood, and they are central to any risk-risk judgement.

• How much we can prepare for extreme climate change. We are unsure to what extent adaptation and resilience work can meaningfully reduce the worst outcomes, and to what extent adaptation to likely new temperature ranges also prepares for extreme climate change.

Further resources

Introductory reading

• An issue brief by FP Analytics about tackling the existential threat of climate change.

• An introduction to SRM on SRM360.org.

• Giving Green's report on solar radiation management, which assesses where philanthropy can help.

• The 80,000 Hours problem profile on climate change, which examines the indirect and existential risks in detail.

• Solar radiation modification: NOAA State of the Science factsheet (Archived version on climate.gov) 

• A Feminist Approach to Solar Geoengineering, an article by Anni Pokela for the Green European Journal.

  
  

Podcasts

• Kelly Wanser (from SilverLining) on whether to deliberately intervene in the climate, an 80,000 Hours podcast episode on climate interventions that are already happening, the most promising ideas, and when they should be deployed.

• David Denkenberger on using paper mills and seaweed to feed everyone in a catastrophe, ft Sahil Shah, also on the 80,000 Hours podcast.

• Martin Weitzman on climate change, an EconTalk episode on the non-trivial risk of a climate catastrophe.

  
  

Technical reading

• Climate stabilization: a draft ITN update, an EA Forum post making the case for stabilisation as a neglected field. (Also, a response to this forum post in African Tech Futures Lab being critical about the term 'climate stabilization.)

• Davies and Vinders (2025). Geoengineering, the Precautionary Principle, and the Search for Climate Safety, analyzing how to think about climate interventions in a risk/risk frame.

• Reflecting Sunlight, the US National Academies' recommendations for solar geoengineering research and governance.

• Belalia (2019). Optimal Climate Policy in 3D. Describes how SRM can integrate with emissions reduction and carbon dioxide removal.

• Weitzman (2011). Fat-Tailed Uncertainty in the Economics of Catastrophic Climate Change.

We’re grateful to dr. Craig Segall and dr. Barbara Šiljeg for providing useful feedback on an earlier version of this article. Their advice does not imply endorsement and any mistakes remain ours.