Skip to main content
U.S. flag

An official website of the United States government

Publication Date
1 July 2023

Dry Weather, Abundant Fuel Contributing to Longer Fire Season in the Western United States

Subtitle
SUMMARY: A machine learning model is used to understand how and why large-fire emissions have increased in the past decade.
Print / PDF
Powerpoint Slide
Image
A forest fire.
Science

In the western United States, large wildfires are intensifying and becoming more frequent. Understanding the cause of these fires with large emissions is crucial to understanding how to prevent, prepare, and fight wildfires. Fuel for the fire, like trees and brush, is a key contributor to wildfires in addition to how dry an area is. While past research has highlighted factors linked to large fires, primarily dry conditions, understanding their combined influence has remained limited. Researchers at Pacific Northwest National Laboratory have identified four clusters representing the impact of fuel load and varying levels of dryness driven by different factors. Notably, large-fire emissions of fine particulates in the first three clusters significantly increased in September and October from 2010 to 2020 when compared to 2000 to 2009. This shift has extended the fire season into fall due to reduced fuel moisture, more frequent fire-conducive meteorological patterns, and drought conditions.

Impact

Using explainable artificial intelligence, which helps interpret predictions made by machine learning models, this study reveals an extended peak season for large-fire emissions that stretches into the fall in the western United States and the key drivers of these changes. A prolonged fire season poses increased health risks from smoke exposure. The findings underscore the importance of planning tools, such as prescribed fires, which reduce fire fuels and enhance ecosystem health to mitigate increasing fire risks.

Summary

Global warming is raising temperatures and creating dryer conditions across the western United States. This increasingly warmer climate influences the seasonal water cycle and is changing wildfire activity and its seasonality. By leveraging the power of explainable artificial intelligence and a statistical clustering method, Pacific Northwest National Laboratory researchers built a machine learning model to predict fire emissions over the western United States and grouped the grids with large-fire emissions by predictors of larger contributions to fire emissions. Four groups of large-fire emissions controlled by abundant fuel and extreme, moderate, and weak drying conditions, respectively, were identified. The dry conditions and increased fire emissions are caused by multiple factors, including warmer temperatures, drought, local dryness, and large-scale meteorological patterns which are favorable to fires like high-pressure and low-relative humidity. Additionally, the large-fire emission peak of the first three groups extends from summer to autumn. These findings underscore the importance of drying increasing the autumn risk of large-fire emissions across the western United States.

Point of Contact
Ruby Leung
Institution(s)
Pacific Northwest National Laboratory
Funding Program Area(s)
Publication