Executive Summary
In a 25 February 2026 Nature Comment, Peter A. Stott and colleagues argue that climate governance still lacks an authoritative, internationally mandated assessment of avoidable climate-change risks. Their central distinction is technical but consequential: science assessments summarize evidence, whereas risk assessments ask which impact thresholds matter most, how likely they are to be crossed, and which outcomes rapid emissions cuts could still prevent. Supplementary examples in the article package—high-end sea-level rise threatening London and extreme humid heat threatening parts of South America—show why the authors want a more systematic, threshold-based, and interdisciplinary approach to global decision-making.
The Key Development
The immediate development is not a new climate model or a new attribution study, but a call for a new global analytical instrument. Stott and co-authors say there has still been no internationally coordinated and mandated global climate-risk analysis, despite decades of climate research and repeated warnings about escalating hazards (Met Office, 2026; University of Reading, 2026). That gap, they argue, leaves governments, businesses, and citizens with an incomplete view of what is actually at stake.
The Analytical Gap
The article’s core argument is that policymakers already recognize many climate hazards, but often do not understand their full severity, timing, or avoidability. The authors emphasize that IPCC assessments remain crucial for establishing what is known with high confidence, yet they are not designed to function as comprehensive global risk assessments. A climate-risk assessment, by contrast, is intended to clarify the scale and urgency of potential outcomes so decision-makers can prioritize mitigation and preparedness more effectively.
The Science-First Distinction
The supplementary material makes the proposed methodology more concrete. It describes a risk-assessment approach that begins by identifying a threshold of impact severity and then estimating the likelihood of crossing that threshold over time or as a function of variables such as global mean surface temperature or sea level. In effect, events that first appear only in the tail of a distribution can become central planning risks, and uncertainty does not eliminate the need to communicate that transition clearly.
The Cascading Risk Problem
A major reason the authors want a dedicated assessment is that climate risk is systemic rather than siloed. The supplementary figures distinguish direct effects, such as heat-driven mortality, from indirect and higher-order effects, such as intense precipitation interacting with degraded land to drive flooding and landslides, or drought propagating through food insecurity into civil unrest. The illustrated cascade spans health and livelihoods, food systems, critical infrastructure, ecosystems, the global economy, and geopolitics, underscoring that climate hazards compound rather than remain confined to one sector at a time.
The London Threshold Example
The article’s sea-level example is deliberately framed around impact thresholds rather than average change alone. In the supplementary material, a plausible worst-case scenario based on plausible ice-sheet instability mechanisms is described as threatening London’s viability by 2150 by pushing local sea-level rise beyond a suggested 5 m adaptation limit for flood defences. By 2200, that same scenario is associated with more than 10 m of sea-level rise, a 200% increase in flood-risk area, and more than 2 million additional people in London exposed to flood hazard based on 2025 population data; the authors also note that at 3 m of sea-level rise, the current Thames Barrier system would become ineffective and a new tidal barrage system would be required.
The Humid-Heat Example
The humid-heat example makes the same point for physiological risk. The supplementary material states that by 2100, millions of people in South America could be exposed on hundreds of days per year to humid-heat conditions exceeding human thermoregulation limits, operationalized here as daily maximum wet-bulb temperature above 29 °C. Under the illustrated plausible worst-case scenario, more than 3 out of every 4 people are affected, compared with less than 1 out of every 300 today; for Belém, Brazil, the probability of experiencing more than 7 such days per year stays near zero under low emissions but rises sharply from mid-century onward under high emissions.
The Important Caveat
A science-first reading of the piece requires taking its caveats as seriously as its warnings. The authors explicitly say that precise probabilities for both the sea-level-rise and humid-heat examples cannot currently be robustly quantified. Even so, the methods are not hand-waving: the sea-level illustration draws on 450,000-member Monte Carlo simulations, while the heat-stress example spans CMIP6 models with transient climate response values ranging from 1.32 to 2.72, making the figures best understood as structured illustrations of plausible risk rather than exact forecasts.
Why Existing Assessments Are Necessary But Not Sufficient
This is why the article should not be misread as dismissing existing climate science assessments. The authors explicitly credit global assessments such as those from the IPCC with playing a crucial role in evaluating evidence, but argue that likely-impact summaries are not the same as decision-ready risk analyses. In accompanying institutional explanations, the gap is described as the absence of an authoritative, up-to-date assessment that can help governments, businesses, and communities understand the full scale of the threat, prioritize resources, and plan effective responses (Met Office, 2026; University of Reading, 2026).
What A Global Assessment Would Need
The article implies that a credible global assessment would have to be threshold-based, interdisciplinary, transparent, and internationally accepted. It would need to combine climate modeling with health data, socioeconomic analysis, ecological knowledge, and systems thinking, because the expertise required to define socially meaningful thresholds is not the same as the expertise required to estimate the probability of crossing them. The institutional statements associated with the article also emphasize the need for a framework that can be updated regularly, identify vulnerable regions and populations, and support both mitigation and adaptation planning (Met Office, 2026; University of Reading, 2026).
Why This Is Not A Counsel Of Despair
The authors are careful to frame the proposed assessment as a tool for agency, not fatalism. Their point is that a rigorous accounting of severe outcomes should clarify which futures remain avoidable, not normalize the worst case as inevitable. That logic is visible even in the supplementary figures, where low-emissions pathways keep threshold-crossing risk far below high-emissions and plausible worst-case trajectories, reinforcing the article’s claim that emissions reductions still materially change the risk landscape.
Conclusion
The lasting value of this Nature Comment is that it reframes the climate-policy problem as an assessment problem as well as an emissions problem. If global leaders are being asked to make high-consequence decisions under deep uncertainty, then they need more than evidence that climate change is real and harmful; they need a structured account of which thresholds matter most, how risks cascade across systems, and what rapid mitigation can still prevent. On the authors’ reading, that is the missing layer between climate science and proportionate global action.
Frequently Asked Questions (FAQs)
- What is the main claim of the Nature comment? The main claim is that the world still lacks an authoritative, internationally mandated assessment of avoidable climate-change risks, and that this gap makes it harder for leaders to judge the scale, urgency, and preventability of major climate threats.
- How is a climate-risk assessment different from an IPCC-style assessment? The authors argue that IPCC-style assessments are indispensable for synthesizing evidence and confidence levels, whereas a risk assessment is designed to identify severe thresholds, estimate the chance of crossing them, and clarify which outcomes deserve highest priority for avoidance or mitigation.
- Why do the authors focus on avoidable risk rather than inevitable damage? Because the article is explicitly about decision-useful analysis. The point is not only to catalogue harm, but to show which harmful outcomes can still be substantially reduced through rapid emissions cuts and better planning.
- What do the London and Belém examples show? They show how threshold-based risk framing changes the policy conversation. London illustrates how sea-level rise can become an existential urban adaptation problem, while Belém illustrates how humid heat can cross into conditions that threaten human thermoregulation under higher-emissions pathways.
- What would make a global assessment scientifically credible? Based on the article package, credibility would require explicit thresholds, transparent methods, interdisciplinary evidence, regular updating, and honest treatment of uncertainty, including situations in which precise probabilities cannot yet be robustly quantified.
Sources
- Met Office. (2026, February 25). Global call to action: Addressing the critical gap in climate change risk assessment. Met Office.
- Nature. (2026). Supplementary information to: We need a global assessment of avoidable climate-change risks. Nature.
- Stott, P. A., Lo, Y. T. E., Marsham, J. H., Obura, D., Oliver, T. H., Palmer, M. D., Ranger, N., Sharpe, S., & Sutton, R. (2026, February 25). We need a global assessment of avoidable climate-change risks. Nature, 650, 826–828. doi:10.1038/d41586-026-00544-6.
- University of Reading. (2026, February 25). Update climate change risks list to avoid worst impacts. University of Reading.