Since the wildfires in January that destroyed neighborhoods in Los Angeles, displacing nearly 13,000 households and causing an estimated $30 billion in damages, California communities have been seeking ways to better protect themselves from similar disasters.

A recent study led by researchers at the University of California, Berkeley offers new data on how specific wildfire mitigation measures can cut damage from such events by up to half. The study analyzed two main strategies: home hardening and creating defensible space. Home hardening includes structural changes like using fire-resistant building materials and installing double-paned tempered glass windows. Defensible space involves clearing vegetation around homes to create a buffer zone.

Researchers used advanced simulation tools along with real-world data from five major California fires prior to 2022 to measure the effectiveness of these methods. According to their findings, combining home hardening and defensible space could double the number of buildings that survive a wildfire. They also found that simply removing vegetation within five feet of homes—an approach considered under proposed Zone Zero regulations—could reduce losses by 17%.

“I view this as really powerful evidence that the mitigation measures that are available to us, hardening and defensible space, actually have some real-world effectiveness,” said Michael Gollner, associate professor of mechanical engineering at UC Berkeley and senior author of the study.

Gollner explained that these strategies not only help prevent property loss but also slow fire spread, giving residents more time to evacuate and emergency services more time to respond. “We can’t always change the spacing between structures or the exposure from flames and embers,” he said. “But even within those limitations, we still have the power to cut the destruction in half, if not more. That is very powerful.”

The study was published August 28 in Nature Communications with support from CAL FIRE’s Forest Health program, the Gordon and Betty Moore Foundation, and the National Science Foundation.

To conduct their analysis, Gollner’s team used CAL FIRE’s Damage Inspection Database (DINS), which contains records from surveys of all structures damaged or destroyed in major California wildfires since 2013. They focused on five significant incidents: the Tubbs and Thomas fires (2017), Camp Fire (2018), Kincade Fire (2019), and Glass Fire (2020). Additional geospatial data provided information about building spacing, construction materials, and nearby vegetation.

By integrating machine learning techniques with this dataset, researchers developed a model predicting whether a structure would survive a wildfire with 82% accuracy. This allowed them to isolate how factors like distance between buildings, fire exposure levels, construction type, and defensible space interact to affect risk.

“We wanted to identify the risk factors that make a structure susceptible to loss,” said Maryam Zamanialaei, postdoctoral scholar at UC Berkeley involved in the research.

“It’s possible that a well-protected home may have a low chance of survival because of everything around it,” Gollner added. “The model allows us to tune in to see the impact of each factor and how they interplay.”

The research found that separation distance between structures was most important for reducing losses in densely built areas where fires can leap from one building to another. Flame length was also identified as a key contributor. Building features like siding material and window type were shown to play significant roles as well.

However, Gollner emphasized that community-wide adoption is necessary for these mitigation strategies to be most effective—a point underscored by ongoing debates over regulations such as Zone Zero requirements.

“Much of what you can do to prevent these fires from spreading through the whole community happens on an individual’s property and depends on what your neighbor does,” he said. “This is a really challenging social, economic and political problem that requires a lot of groups working together.”

He hopes these findings will help justify investments into wildfire prevention efforts across California communities: “We need to justify the investments we’re making in mitigation, and I was glad to see that for many of them, we do see significant payback in terms of risk reduction,” Gollner said.

Co-authors include Daniel San Martin (Universidad Técnica Federico Santa María); Maria Theodori and Dwi Purnomo (UC Berkeley); Ali Tohidi, Arnaud Trouvé and Yiren Qin (University of Maryland); Chris Lautenberger (Cloudfire).

Related work shows UC Berkeley scholars are actively helping Bay Area communities prepare for wildfire season through simulations showing how fires might spread through neighborhoods: https://news.berkeley.edu/2023/05/31/wildfire-season-is-here-uc-berkeley-scholars-are-helping-bay-area-communities-prepare;https://news.berkeley.edu/2022/08/16/how-uc-berkeley-researchers-simulate-how-wildfires-spread-through-communities