Public conversation about climate change is heavily influenced by efforts to understand the role of global warming in specific weather-related disasters. The science of extreme event attribution (EEA) has become stronger and more rigorous in the last decade, according to a new report from the National Academies of Sciences, Engineering, and Medicine.
James Hurrell, Scott Presidential Chair of Environmental Science and Engineering at Colorado State University, served as the chair of the National Academies committee that produced the report, titled, “Attribution of Extreme Weather Events and Their Impacts in the Context of Climate Change.”
“My role,” he notes, “was to guide the committee’s deliberations, facilitate consensus among experts from a wide range of disciplines, and oversee preparation of the report. The report reflects the collective judgment of the committee rather than the views of any individual member.”
AMS Headlines spoke with Hurrell about the report’s findings. His responses to our questions are below.
“Although uncertainty remains, scientists are increasingly able to estimate how the same event might have unfolded in a world without human influence.”
—James Hurrell
Q: Why is attribution science important as we plan for a warming future?
A: As the climate continues to warm, people naturally want to know whether climate change played a role in the extreme events they experience. Attribution science provides a rigorous scientific framework for answering that question.
Attribution studies use individual extreme events to improve our understanding of how climate change is altering the characteristics, likelihood, and impacts of different types of extremes. That information can help governments, businesses, emergency managers, insurers, infrastructure planners, and the public make more informed decisions about managing future climate risks.
Extreme event attribution science not only helps explain the role of climate change in individual events, but it improves our understanding of the physical processes that produce extremes and provides information that helps society become more resilient to future climate risks.
Q: What are the fundamental ingredients required to determine the role of climate change in an extreme event?
A: Successful attribution depends on several key ingredients.
First, scientists need high-quality observations to characterize both the event itself and the long-term climate record.
Second, we need a sound physical understanding of the processes responsible for the event and how climate change may influence those processes.
Third, attribution studies require realistic representations of both today’s climate and a plausible “counterfactual” climate representing conditions without human-caused greenhouse gas emissions. Depending on the attribution approach, these representations may rely primarily on climate models, observations, statistical methods, or combinations of these. They are then used to estimate how climate change influenced the likelihood or intensity of the event by comparing the observed world with the counterfactual one.
Confidence is highest when observations, physical understanding, the methods used to represent the current and counterfactual climates, and statistical analyses all tell a consistent story.
Q: What are a few important ways that attribution science has advanced in the past decade?
A: The past decade has seen substantial progress both in the scientific foundations of extreme event attribution and the emergence of new approaches for understanding climate-related impacts.
Observational records are now longer, more complete, and of higher quality than they were a decade ago. New satellite observations, improved reanalyses, better homogenized station datasets, and the continued recovery of historical records have strengthened our ability to detect long-term changes and place individual events into historical context.
Climate models have also improved significantly. They represent many important physical processes more realistically, operate at higher spatial resolution, and are increasingly run as large ensembles. Large ensembles allow scientists to better separate the influence of human-caused climate change from natural climate variability and provide more robust estimates of changes in the probability or intensity of extreme events.
Methodologically, EEA studies have become more sophisticated. Multiple complementary approaches—including probabilistic event attribution and physically based storyline approaches—are now widely used, with each providing different but complementary insights depending on the scientific question being addressed.
Our ability to construct counterfactual climates has improved through advances in climate models, larger ensembles, and improved understanding of historical forcing. Although uncertainty remains, scientists are increasingly able to estimate how the same event might have unfolded in a world without human influence.
Rapid attribution has advanced considerably. For some event types—particularly extreme heat and many heavy rainfall events—scientists can now provide scientifically robust attribution assessments within days or weeks of an event, provided appropriate observational and modeling systems are already in place.
A distinguishing feature of this report is that it addresses not only extreme event attribution, but also the emerging field of extreme event impact attribution (EEIA), which examines how climate change influences impacts such as economic losses, infrastructure damage, ecosystem disruption, and human health outcomes. Understanding societal impacts requires closer collaboration among climate scientists, engineers, economists, ecologists, public health experts, and social scientists to better understand how climate hazards interact with exposure, vulnerability, and societal responses to shape real-world outcomes.
Q: What key limitations remain in attribution science?
A: Some of the greatest challenges involve extreme events that depend strongly on atmospheric dynamics rather than primarily on thermodynamics. While the influence of greenhouse gases on many thermodynamic aspects of climate—such as increasing atmospheric moisture and extreme heat—is well understood, the influence of climate change on atmospheric circulation is often less certain. Climate models have become much more capable, but they still struggle to fully represent certain aspects of atmospheric dynamics, including changes in blocking patterns, jet streams, and modes of natural climate variability such as El Niño and La Niña. These uncertainties make attribution more difficult for events whose occurrence depends strongly on those dynamical processes.
Similarly, many high-impact hazards—including severe thunderstorms, tornadoes, hail, and localized flash flooding—arise from interactions between large-scale atmospheric circulation and small-scale weather processes that remain difficult to simulate. Continued improvements in model resolution, process representation, and our understanding of these cross-scale interactions will be important for increasing confidence in attribution of these events.
Large observational gaps also remain, particularly in many low-income and under-resourced regions. Expanding observing networks, recovering historical records, and improving access to high-quality climate data are essential for ensuring that attribution science benefits all regions of the world.
Confidence varies considerably among extreme event types. The scientific evidence supporting attribution is strongest for events such as extreme heat and many heavy precipitation events, where the physical influence of greenhouse gases is relatively well understood. Confidence is generally lower for hazards that depend more strongly on atmospheric dynamics or occur at very small spatial scales. Recognizing these differences in confidence is an essential part of attribution science and one of the reasons the committee emphasized evaluating each event type on its own scientific merits.
Advancing EEIA will require better observational data on impacts, improved impact models and impact-response relationships, and closer collaboration across disciplines to better quantify how climate-related hazards translate into societal and environmental consequences.
Q: What opportunities or challenges do you see in the near future for this field?
A: One exciting opportunity is the continued convergence of advances in observations, physical understanding, climate modeling, and computational capability. Together, these advances should allow attribution studies to become both more accurate and more useful for society.
Higher-resolution Earth system models, larger model ensembles, improved representation of atmospheric dynamics, and expanding observational datasets will increase confidence for many types of events that remain difficult today. In particular, the committee emphasized the need for models that better represent interactions across spatial scales—from regional and storm-scale processes to large-scale circulation—to improve our understanding of how climate change may influence atmospheric dynamics. Advancing these capabilities is especially important for events that depend strongly on atmospheric circulation, such as severe convective storms and some forms of drought and flooding.
Another important opportunity lies in strengthening partnerships between scientists and decision-makers. Increasingly, attribution studies are being designed with input from stakeholders so that the information produced better addresses real-world decisions involving infrastructure, public health, water resources, agriculture, insurance, and disaster preparedness.
The committee also identified an important opportunity to strengthen collaboration across the attribution community itself. One of our central recommendations is to develop a common framework to advance both EEA and EEIA. As attribution science has matured, a diversity of methods has emerged, driving innovation but also making studies more difficult to compare and reproduce. We envision a framework built around shared best practices, rigorous treatment of uncertainty, multiple lines of evidence, and clear documentation. Developing such a framework will require collaboration across the international research community. The resulting framework would strengthen collaboration across disciplines while making attribution studies easier to compare, communicate, reproduce, and apply across scientific and decision-making contexts. Importantly, the goal is to strengthen the science—not to standardize it in ways that limit continued innovation.
A continuing challenge will be communicating both what attribution science can say with confidence and where important uncertainties remain. Maintaining that balance is essential to preserving scientific credibility and ensuring that attribution results are used appropriately.
Attribution science is not intended to answer every question surrounding an extreme event. It addresses one specific scientific question: how human-caused climate change influenced the probability or intensity of an event. Other important factors—including exposure, vulnerability, preparedness, and societal decisions—also shape the impacts that people ultimately experience.
Recognizing both the growing strengths and the remaining limitations of attribution science is essential for ensuring that it continues to provide credible, actionable information for society.
