When UC Berkeley postdoctoral researchers Maya Zomer and Lucia Layritz first drove through burned portions of Mojave National Preserve in Southern California, the sight stunned them. "It looks like a pretty dystopian landscape," Layritz said.
More than a million Joshua trees had burned when the 43,000-acre Dome Fire tore through in 2020, with many plants blackening in the flames. Over time, the charcoal had chipped off, revealing wizened white stumps that appeared like skeletons. Zomer, Layritz and their companions were quiet as they passed miles of Mojave desert resembling a moonscape.
PhiLip via Wikimedia
On closer inspection, the researchers and park employees saw tiny signs of life. Walking around sun-bleached Joshua tree remnants, they spied green resprouts. Gradually, they noticed that some of the trees had survived, and bright desert wildflowers grew among the dead plants. Yet despite these hopeful signs, the group knew that the Mojave Desert and its iconic Joshua trees were existentially threatened by increasing - and increasingly severe - wildfires.
"People were very concerned," Layritz said. "It was like, okay, are we, I don't know, five big fires away from destroying the majority of the Joshua trees? And what do we do if that happens?"
Parks have limited time and resources to deploy after natural disasters such as fires. That's why researchers at Berkeley's Eric and Wendy Schmidt Center for Data Science & Environment are teaming up with the National Park Service to develop digital fire recovery tools - software that can help park managers respond quickly and effectively after a fire, protecting and restoring as many Joshua trees as possible.
Plants and people navigate new climate conditions
Joshua trees are quintessential Mojave Desert plants, though they also grow in the Great Basin and Sonoran Deserts. Technically two species, the yuccas can grow 40 feet tall and live for more than a century. But after thriving in the desert for thousands of years, they are threatened by a hotter and drier climate caused by greenhouse gas emissions. They also contend with increasingly severe wildfires, driven by the spread of nonnative grasses that carry flames across their desert habitat. These factors have put them in danger of almost completely disappearing from their namesake Joshua Tree National Park, according to a 2019 study from UC Riverside.
"The frequency, magnitude and spatial area of the burns have definitely experienced a huge ramp-up in the last eight to 10 years within the greater Mojave Desert," said Maya Weltman-Fahs, a senior program manager at the Berkeley research center. She attributes the increasing number of fire ignitions to human activity, including camping.
The Mojave Desert did not evolve with wildfires of the size and frequency that occur now. What grows back after a burn can be very different from the ecosystem that existed before, and many desert species may not be able to recover on their own. Because Joshua trees take up to 70 years to become reproductive and the region continues to warm, those resprouts that Zomer and Layritz found are growing in a much harsher environment than did the adults that burned. In many places, they will not survive these new conditions.
"The Mojave Desert may be at a critical shift from small, patchy fires to continuous large fires," Zomer said. "That's the motivation of a lot of work we are doing."
Jacob Sohr via Wikimedia
What's more, areas that scientists consider to be "climate refugia" - locations in Joshua Tree National Park that are high enough in elevation for Joshua trees to persist even as lower elevations become too hot - are also the most fire-prone because of that increasing grass coverage. "There have been fires in the climate refugia, which are a really dense Joshua tree habitat, that have been really devastating to some of the established reproductive adult Joshua tree populations," Weltman-Fahs said.
After a fire, parks have limited time to report on the recovery actions they plan to take and to request necessary federal funding. Collaborators at Joshua Tree National Park told the researchers that it is difficult to make these fast decisions in a desert landscape without historic fire information to draw from, unlike in more forested places with a history of fires, such as California's coniferous forests.
"What we talked a lot about was how wonderful it would be to have a tool that could use the most current scientific modeling expertise to be able to do post-fire scenario planning," said Zomer. She and her colleagues are now creating one.
In forests, satellite imagery quickly reveals the severity of a wildfire without people having to hike across a charred landscape to visually assess the damage. This information helps managers decide where to focus recovery efforts. But in the desert, those interpretations of satellite data don't apply.
"The desert ecosystems are very different from, say, a burned versus an unburned forest and what that would look like from satellite imagery," Weltman-Fahs said. "The desert had not been well-served by the existing comparative indices to examine the burn severity from satellite imagery. So that was the place that we started with our Joshua Tree partners - can we give you the information really quickly in the aftermath of a fire of what got burned, where, and how bad was it?"
Parks could then use that information to target their ecosystem recovery efforts in places where they are most likely to succeed, and to know where in the park to accept the change that's happening. "Resource limitation can also refer to how many boots they have on the ground to actually be going out and planting seedlings and watering them and helping support their establishment," Weltman-Fahs said. "If they're going to put forth that effort, they want to have a science-based, defensible reason to do that here rather than here."
Last year, the Park Service used the research group's burn severity assessment tool for the first time after the Eureka Fire in Joshua Tree National Park, which burned just over 200 acres.
"We were able to do a quick assessment of fire severity and help the park with some analysis of how much vegetation and which types of vegetation were within the burned area, and which were in the more severe parts of the fire," Zomer said.
Courtesy of the National Park Service
The park also used the tool after the 2025 Black Rock Fire, which started near a campground in the park and burned 72 acres.
"We used DSE's Disturbance Severity Tool twice last year to assess post-fire conditions following the Eureka and Black Rock fires," said Nick Graver, the botany and vegetation monitoring program lead at Joshua Tree National Park and a graduate student in UC Riverside's Department of Evolution, Ecology and Organismal Biology. "Combined with field assessments conducted by park biologists, it provided the foundation for evaluating fire severity and developing recovery plans for both incidents."
In addition to the burn severity assessment tool, researchers are now developing a platform to model the future outcomes of potential recovery actions. The platform, called Josh, just launched this year and will be available when the next wildfire sweeps through either park.
"Josh is a scenario planning tool where you can play out different restoration strategies, like outplanting or herbivory protection, and understand what their potential is, especially when you consider climate change," Layritz said. The platform is a user-friendly website where park managers can run different scenarios to see what the predicted outcomes of efforts made now will be far in the future, then decide what restoration actions to take and where after a fire.
The goal is to help managers make decisions about where to focus recovery efforts right after a fire, with an understanding of what the results will be in the future. "You would be able, if there's a fire, to first run the fire severity module, which gives you a map of fire severity," Zomer said. "That would be moved into the modulating platform, where then you can get these different future scenarios."
Eventually, Weltman-Fahs said, the hope is that parks won't need to ask Berkeley scientists for input when using these platforms to make quick decisions about wildfire recovery. "Our goal is that nobody needs to call DSE when a fire breaks," Weltman-Fahs said. "They can just utilize the tools that we are offering to make their decisions and rehabilitate landscapes."
