YSE Senior Research Scientist Jennifer Marlon was part of an international team that combined global fire data and climate models to identify gaps in wildfire risk projections and outline a framework to better inform long-term planning.
How can scientists better assess the effects of a warming climate on wildfire behavior, fire risk, and long-term fire trends? A new review co-authored by Yale School of the Environment senior science researcher Jennifer Marlon outlines how climate change is reshaping wildfire patterns worldwide. The review, published in Science Advances, also details why understanding these shifts is essential for forecasting future risks.
“We’re seeing landscape fires increasingly threaten human health and property in ways that demand urgent attention,” Marlon said.
The paper, “Gazing into the flames: A guide to assessing the impacts of climate change on landscape fire,” brings together research from wildfire scientists across Australia, Europe, Canada, and the United States to focus on “the art and science” of projecting climate change impacts on landscape fire, including how fire’s drivers and impacts are modeled and how projections of the climate system are developed. Drawing on decades of fire research, satellite observations, climate model projections, and landscape fire modeling, the authors argue that traditional approaches fall short unless they integrate ecological, climatic, and human drivers.
“Smoke from wildfires is degrading air quality thousands of kilometers from where fires ignite. Entire communities are being destroyed, and the health, economic, and social consequences far outlast the fires themselves. Understanding and preparing for these risks requires more than any single discipline can offer,” Marlon said.
Wildfires don’t burn in isolation. Their frequency, size, and severity depend on fuel availability (such as dead grasses, shrubs, and trees), ignition sources, short-term weather, and longer-term climate patterns. Climate change influences all of these factors. For instance, warming temperatures dry fuels more quickly, extend fire seasons, and altered precipitation patterns can also increase fire risk.
We’re moving rapidly into a warmer world with no historical analog. The relationships between climate and fire that we’ve relied on may not remain stable.”
From boreal forests to major cities, recent fires show how wildfire threats are increasingly putting communities at risk. Canada’s 2023 season burned 15 million hectares and set records for smoke emissions. Australia’s 2019–2020 Black Summer consumed roughly one-fifth of the southeast’s forests, while the 2025 Los Angeles–area fires exposed how quickly extreme conditions can overwhelm urban areas.
The review finds that research on climate change and fire is heavily concentrated in North America, Australia, and Western Europe.
“The vast majority of research on climate change and fire comes from North America, Australia, and Western Europe,” Marlon said. “Yet, most of the world’s burned area is in tropical savannas and grasslands. That mismatch leaves significant blind spots in regions that are globally important for fire emissions and ecosystems.”
Additionally, fire risks are not changing uniformly. While satellite records show declining burned area in some savanna regions, temperate and boreal forests are increasingly experiencing fires, sometimes in places that historically saw less fire activity. Understanding these shifts requires models that capture more than climate trends alone.
A central message of the paper is that standard Earth System Models often do not include the full range of processes that drive landscape fire. Although they simulate temperature, precipitation, and large-scale climate dynamics, they may lack detailed representations of vegetation change, ignition sources, land management, and human behavior, all of which, according to researchers, strongly influence fire outcomes.
Biweekly, we highlight three news and research stories about the work we’re doing at Yale School of the Environment.
The review also identifies a potential source of underestimation. In some regions, climate models tend to overestimate atmospheric moisture, which can lead to underestimating fuel dryness, a key driver of fire activity. If fuels are drier than projected, future fire impacts could be more severe than some current estimates suggest.
Unlike greenhouse gas emissions pathways, there are no widely adopted “fire scenarios” to guide long-term planning. Modeling human behavior — from ignition sources to fire suppression — remains especially challenging. Perhaps most fundamentally, relationships between climate and fire observed in the past may not hold as the world continues to warm.
“We’re moving rapidly into a warmer world with no historical analog,” Marlon noted. “The relationships between climate and fire that we’ve relied on may not remain stable.”
To address these uncertainties, the authors recommend using multiple climate models individually rather than simply averaging them, investing in interdisciplinary research teams, and improving how fire models are evaluated, including testing them against Global Paleofire Database records that extend beyond the modern observational era.
“Fire sits at the intersection of atmospheric science, ecology, public health, economics, land-use planning, and Indigenous knowledge systems,” Marlon said. “But these fields rarely work together in an integrated way. The science of projecting future fire needs to be as interdisciplinary as fire’s impacts are.”
