Climate change is altering wildfire seasons in North America, but the direction of the shift depends on the regional ecosystem, a new study shows.
The fire season in the northern boreal forests of Alaska and Canada has, on average, shifted forward; prairie regions have seen little change; and the fire season in the arid West and California has extended further into late fall and winter. The findings were published on February 24 in the journal Geophysical Research Letters.
The article continues below
Zhang and his colleagues used data on fire area in North America from 2001 to 2020 obtained from the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument on NASA’s Aqua and Terra satellites. They also collected data on meteorological variables, vegetation, lightning potential and other environmental factors at the time of the fires.
They found that the boreal forest, or taiga, in Canada, Alaska and the Great Lakes region sees earlier fires. This is due to earlier melting of snow, and thus earlier dryness of fuel. Canada experienced it worst forest fire season in 2023 and its second worst just two years later.
The hot desert southwest and the Mediterranean-like climate region of California saw an extension of the fire season, with more fires burning after the traditional high-risk window.
Prairies and grasslands experienced little change in fire season intensity and minimal change in seasonal timing. The Appalachians and Southeast forests also saw little in the way of seasonal shifts.
The researchers also modeled future scenarios. During high emissions climate change scenario, they found, the boreal forest fire season would be shifted by about a week, while California’s annual fire season could extend more than a month later than the current June to October window. The desert southwest could see a similar stretch of fire season, the researchers wrote.
This model will be useful for more detailed study, Zhang said. He and his team plan to use it to study the effects of other factors, such as vegetation changes and human activities. (According to the National Park Service, 85% of wildland fires in the U.S. are caused by human actions, such as arson or failure to extinguish a fire properly.) The model is also useful for predicting pollution and carbon emissions from those fires, Zhang said.
“This model is very good at predicting wildfires,” he said. “So now we want to predict the emissions of wildfires to the atmosphere.”
