UNIVERSITY PARK, Pa. — As wildfires relentlessly sweep across Southern California and other parts of the world, Manzhu Yu, an assistant professor of geography at Penn State, offered insights into the ongoing crisis in Los Angeles (LA). Her expertise lies in atmospheric modeling, environmental analytics, big data and cloud computing — fields that intersect closely with wildfire behavior, smoke prediction and exposure analysis.
Q: Could you explain the primary factors contributing to the increased frequency and intensity of wildfires in California recently?
Yu: One of the most significant factors is climate change. Warmer temperatures, reduced precipitation or a changed rainfall season and longer fire seasons have dried out California’s landscapes, increasing the potential for ignition and rapid fire spread. Extended periods of drought further exacerbate the issue. Dry vegetation becomes ready to ignite with the slightest spark. These changes have contributed to an annual average burned area in 2020-23 that is three times higher than in the 2010s.
While some fires are fuel-dominated due to century-long fire suppression and changes in land management, others are wind-dominated, particularly in southern California, like the recent LA wildfires. Winds like Santa Ana and Diablo winds act as accelerants for wildfires. These strong, dry gusts push flames across vast distances, spreading fires at an alarming rate. Combined with already dry conditions, these winds make controlling wildfires exceptionally challenging.
California’s growing population has expanded into the wildland-urban interface — areas where human development meets natural landscapes — with 12.7% of the state's population now residing in high-growth census tracts along the wildland-urban interface. This increases the chances of fire ignitions from human activities, like faulty power lines, vehicles and even campfires. It also heightens the stakes, as more homes and lives are at risk.
Q: What are the different phases of a wildfire, and what are the challenges associated with each one?
Yu: Wildfire spreads include phases of ignition, active spread, fully developed and decay. During the ignition phase, wildfires can be sparked by natural causes, such as lightning, or human activities like power line failures or campfires. Approximately 84% of wildfires in the United States are caused by human activities. Early detection of ignition would be the most critical but challenging thing to minimize the fire’s impact.
The active spread phase is when the fire begins to grow rapidly, driven by dry vegetation, strong winds and terrain. Fires in California have been found to spread up to 14 times faster under high winds, like the Santa Ana winds, which exacerbate fire intensity and movement. High winds not only accelerate the fire’s spread but also carry embers and firebrands over long distances, igniting new spot fires far ahead of the main fire. This is particularly the case in the LA wildfires burning right now. These flying embers can travel up to a mile or more, making it difficult to predict and contain the fire’s advance. Firefighting and resource coordination during this phase is challenging as high winds can hinder the deployment of aerial resources, such as water-dropping helicopters and firefighting planes.
In the fully developed phase, the fire reaches its peak intensity, consuming significant fuel and spreading aggressively. Extreme heat from wildfires can create pyrocumulonimbus clouds, which can spark new fires miles away. These clouds can exacerbate fire spread by generating erratic winds and even sparking new fires through lightning strikes.
In the decay phase, fire intensity decreases as it runs out of fuel or is under control. However, lingering hotspots can reignite under the favoring conditions. Smoldering fires in forest peat layers can last for months and reignite during dry conditions.
Q: Once a wildfire starts, what key factors determine how quickly and in which direction it spreads?
Yu: Wind is the primary driver of wildfire spread, as wind pushes flames forward, supplies oxygen, and carries embers, or firebrands, that ignite new spot fires far ahead. Strong winds, like California’s Santa Ana winds, can dramatically accelerate fire growth and shift its direction unexpectedly.
Topography is part of this process: fires move faster uphill, as flames preheat the vegetation above them. Steep slopes can amplify fire intensity. Valleys and ridges can alter wind patterns.
Fuel availability, dryness and connectivity is another one. Fine, dry fuels ignite easily, while dense vegetation sustains longer, more intense fires.
Weather elements like low humidity and turbulent air can make fires more erratic. Fire-atmosphere interactions can even create their own wind systems, such as fire whirls.
The location of the ignition point also matters. Fires near steep terrain or during high-wind conditions can spread rapidly and pose greater challenges for containment.
Q: Given the projected increase in wildfire frequency, what long-term environmental consequences do you foresee for California’s landscapes and natural resources?
Yu: Wildfires have already increased dramatically, and this trend is expected to continue, with wildfire frequency projected to escalate through the end of the century. One critical consequence is the transformation of ecosystems, impairing shrub regeneration, reducing native plant diversity and leading to a dominance by non-native. This shift threatens the biodiversity and ecological balance of these unique environments. Wildfires are also significant contributors to greenhouse gas emissions. Carbon dioxide emissions from wildfires have surged by 60% since 2001 globally, nearly tripling in climate-sensitive boreal forests due to rising temperatures and fire-favorable weather conditions. These emissions exacerbate climate change, creating a feedback loop that increases future wildfire risks.
Q: What are some of the most effective modern technologies and strategies being used to predict and control wildfire spread?
Yu: The most common and effective methods used right now are remote sensing and satellite monitoring. Satellites like NASA’s MODIS and VIIRS provide real-time data on fire location, size and intensity. These tools help identify new ignitions and monitor fire progression, even in remote areas. Thermal imaging from drones and ground-based radar complements satellite data, offering high-resolution insights for on-the-ground operations.
California already uses an artificial intelligence (AI)-powered wildfire detection system using AI to train a forest-based camera network to recognize early signs of fire.
There have been many controlled or prescribed burns implemented to reduce excess vegetation, like thinning dense forests and clearing underbrush, to mitigate fire severity. There have been robotics companies starting to try out these controlled burns with the help of robots, like the BurnBot.
Autonomous helicopters, like the modified Black Hawk equipped with Sikorsky’s MATRIX autonomy, have demonstrated the ability to detect and suppress fires independently.
There are a couple of computational models, such as ELMFIRE with wildland-urban interface extensions, that simulate fire spread through communities. These tools incorporate data on vegetation, structure materials and fire dynamics to reconstruct past events and predict future risks.
Google, in partnership with Muon Space and the Environmental Defense Fund, will launch a constellation of satellites called FireSat that can detect as small as a classroom, roughly 5x5 meters.
Editor's note: This Q&A was originally published as an Institute of Energy and the Environment blog.