Remember meForgot password?
    Log in with Twitter

article imageScientists map wildfire fuel moisture across western US

By Karen Graham     May 23, 2020 in Science
Researchers have developed a deep-learning model that maps fuel moisture levels in fine detail across 12 western states, opening a door for better fire predictions - even as drought conditions are expected to worsen.
Talks of drought and an early fire season have put a damper on an already difficult spring with local economies in our western states still suffering from COVID-19 shutdowns. The news is not good.
With temperatures expected to climb into the 80s and near 90s today and on into the coming week, by Thursday, temperatures are anticipated to be 30 degrees higher than the normal high for this time of year. But it is the drought monitor that bears a closer look.
On May 21, 2020, the Drought Monitor map showed that northern California and Eastward are showing abnormally dry conditions, while many dry areas of the Plains and Midwest have not had drought development due to the unseasonably cool temperatures in May.
U.S. Drought Monitor
In the West, temperatures for the region were 3-6 degrees above normal over central Nevada, Utah, Colorado, and eastern New Mexico, with most of the rest of the region near normal to 3 degrees below normal for the week. In the Pacific Northwest, the recent rains helped to slow down further degradation in Oregon and Washington, with portions of the abnormally dry areas of western Washington improved this week.
However, as California and the American West head into fire season amid the coronavirus pandemic, scientists are harnessing artificial intelligence and new satellite data to help predict blazes across the region.
Monitoring moisture content
To this end, a team of researchers at Stanford University in California has developed a deep-learning model that maps fuel moisture levels in great detail - across 12 western states, from Colorado, Montana, Texas and Wyoming to the Pacific Coast.
Woolsey Fire on November 10  2018.
Woolsey Fire on November 10, 2018.
The research team published a description of their technique in the August 2020 issue of Remote Sensing of Environment. According to the senior author of the paper, Stanford University ecohydrologist Alexandra Konings, the new dataset produced by the model could “massively improve fire studies.”
The new model uses what's called a recurrent neural network, an artificial intelligence system that can learn to recognize patterns in vast mountains of data. The scientists trained their model using three years' worth of field data - starting in 2015, when SAR data from the European Space Agency’s Sentinel-1 satellites became available.
Using this data and additional data from the National Fuel Moisture Database, the model was then put it to work estimating fuel moisture from two types of measurements collected by spaceborne sensors.
One estimation involves measurements of visible light bouncing off Earth. The other, known as Sentinel-1 synthetic aperture radar (SAR), measures the return of microwave radar signals, which can penetrate through leafy branches all the way to the ground surface.
Examples of forest dryness progressing across western states in 2019.
Examples of forest dryness progressing across western states in 2019.
Krishna Rao/Stanford University
"The western US spans a large climatological range, with a diverse mix of 6 land covers, making it an excellent location to test the robustness of the algorithm," according to the study.
"One of our big breakthroughs was to look at a newer set of satellites that are using much longer wavelengths, which allows the observations to be sensitive to water much deeper into the forest canopy and be directly representative of the fuel moisture content," said Konings, who is also a center fellow, by courtesy, at Stanford Woods Institute for the Environment.
The model's estimates feed into an interactive map that fire agencies may eventually be able to use to identify patterns and prioritize control measures. “Creating these maps was the first step in understanding how this new fuel moisture data might affect fire risk and predictions,” Konings said. “Now we’re trying to really pin down the best ways to use it for improved fire prediction.”
More about western states, Wildfires, fuel moisture content, deep learning, sentinel1
Latest News
Top News