The Science of Predicting Which Houses Will Survive Wildfires
“Is my house still standing?” The burning question on the mind of each wildfire evacuee is also something researchers want to predict—well in advance.
Scientists have been studying two parts of California to learn which factors determine whether a house will succumb to fire. The results suggest the location and density of homes are important indicators of survival.
Multicolored maps showing levels of fire hazard risk have traditionally focused on the location and density of fuels, under the assumption that if there aren’t good fuels, there won’t be a fire.
Similarly, much of the effort to prevent and control wildfires is centered on going in and changing natural land, such as thinning forests.
“If you look at recent history, you will see that more and more money is being spent on fuel manipulation, and yet we continue to have these extremely large and dangerous wildfires, and we continue to lose increasing numbers of homes,” says Syphard. “It’s time to start thinking about alternative solutions.”
To find out what those better alternatives might be, Syphard and colleagues turned to Google Earth, and began mapping all of the 700,000 buildings in two fire-prone regions in Southern California, the Santa Monica Mountain and the San Diego area. Syphard then looked at what was different about the 5,500 structures that were lost or beat up by wildfires in the last decade.
Syphard found that the primary drivers of home loss were fire history, structure density, and distance to the coast. In particular, homes were most likely to burn if they were located off by themselves, or in small clusters.
Syphard says this is because these houses are more likely to be intermingled with wildland vegetation, and firefighters have less access to them. Houses were also more susceptible if they were located on hills, or at the edge of developments.
Homes surrounded by grasses were just as or more likely to burn than homes surrounded by bigger plants.
“Most people usually associate fire risk with the largest amount of vegetation,” says Syphard. “But fires don’t necessarily differentiate between fuel types.”
Especially with extreme fire conditions, like high wind, any dry vegetation can allow the fire to spread quickly.
This example highlights where fuel mapping alone would have been misleading. Adding in the additional factors Syphard identified, she was able to improve the accuracy of the fire hazard maps.
While Syphard’s study specifically looked at California, she says many of the results are applicable to any place, including Colorado. She says land use planning is critical for limiting the effects of wildfires in the future.
"You can’t change where houses are located right now,” says Syphard. “But you can change the way new homes are placed.”
Syphard is currently working on modeling the best way for new homes to be built so that recommendations grounded in data can inform future construction projects.