Refining fire behavior modeling

Research by USDA Forest Service Southern Research Station biometrician Bernie Parresol takes center stage in a special issue of the journal Forest Ecology and Management due out in June. Parresol is lead author of two of the five articles—and co-author of two more—in an issue that focuses on methods that incorporate fine-scale data into the tools Southeastern forest managers use to assess wildfire potential and plan mitigation treatments.

Most fire behavior analyses rely on sparse plot inventories and data from satellites, and often do not address the complexity found at the ground level where managers operate. Parresol and fellow researchers demonstrated a statistical approach that can incorporate hundreds to thousands of fuel observations into models that managers can easily use to prioritize areas to treat to reduce the wildfire hazard.

Fire is an important part of forest ecosystems in the southeastern United States, especially in the Coastal Plain. European settlers cleared most of the native longleaf pine forests of the region; industry later planted many of the same acres in loblolly pine plantations. Meanwhile, fire suppression policies broke the cycle of frequent low-intensity fires in the remaining natural forests, causing the buildup of fuels that leads to wildfires.

Over the last decades, southeastern land managers added prescribed fire to other forest treatments to reduce wildland fires, promote forest restoration, and improve wildlife habitat. Because of budget constraints and public concerns about fire and smoke, managers need to prioritize the areas where they will use prescribed fire. To do this, they use wildfire hazard assessments such as LANDFIRE and the Southern Wildfire Risk Assessment (SWRA), both of which use satellite images and other supporting data to represent fuels across a landscape. Although these tools work well enough at the state and regional levels, they don't offer enough detail to land managers trying to decide which of their hundreds or thousands of acres should be burned first.

The special issue of Forest Ecology and Management focuses on a study conducted on the 200,000-acre Savannah River Site as representative of an actively managed forest landscape in the Southeast. Researchers used studies on the site to assess wildland fuels, potential fire behavior and treatments to reduce fire hazard. In his first article, Parresol and fellow researchers develop equations to describe fuel loads for both dead and alive materials on the site based on vegetation type, stand age, recent fire history and other aspects. These equations were then used to create custom landscape models based on the actual data from the site, then compared with results from LANDFIRE and SWRA to assess the effectiveness of those tools.

"Taken together, the research reported in these articles shows that fine scale measurements repeated over time can be put into a manageable framework and reduced to create dynamic fire behavior models useful to managers," says Parresol. "They can also be used to help address scientific questions and to evaluate the effect of management conditions."