Sonoma County Complex Fires of 2017: Remote sensing data and modeling to support ecosystem and community resiliency

In the western U.S., long-term fire suppression has led to a buildup of surface and ladder fuels, increasing the severity of fires. Coupled with increased home building in the wildland urban interface and global climate change, much of the western U.S. is facing unprecedented risk of catastrophic wildland fires. Given the almost 30 million acres of forestland in California, and the impacts to human community health and safety and natural systems that stem from uncontrolled fires, it is imperative that we understand the underlying processes and conditions in the landscape that determine fire impacts. In October of 2017, Sonoma County, California experienced three significant fires that resulted in loss of life and property, as well as impacts to natural systems. Sonoma County Ag + Open Space—with support from a team of technical consultants and in partnership with NASA and other experts—researched the impacts of the fires to woody vegetation within areas that burned during wind-driven and non-wind driven events. Using high-resolution aerial imagery, we mapped canopy condition of woody vegetation and used machine learning techniques to determine the importance of landscape measures of vegetation structure, land cover type, topography, climate and weather, and nearness to streams as predictors of woody canopy condition for areas that burned during the October 2017 fires. Across the landscapes, riparian and mesic vegetation types exhibited the least canopy damage, followed by upland hardwood forest types. Shrub and upland conifer types exhibited the most canopy damage. Measures of vegetation structure derived from lidar data are the most important predictors of post-fire woody canopy condition, in addition to slope-aspect, proximate vegetation community types, and distance to streams. In general, the higher the density of shrubs and fire-adapted vegetation types, the higher the density of ladder fuels, and the greater the distance from streams, the higher the canopy damage. This study emphasizes the value of high resolution airborne lidar for mapping vegetation type and structure and building locations at a scale large enough to inform local management decisions. The study also documents pre- and post-fire baseline conditions to support the long-term evaluation of vegetation impacts and provides remote sensing and analysis tools to better plan for, manage, and mitigate future extreme wildfire through the lens of climate and extreme event resiliency, community