Fuel mapping with remote sensing data on the Kaibab Plateau, Arizona
A mountain view on the Kaibab National Forest.
Remote sensing methodologies allow for fuel mapping across heterogenous landscapes, which would otherwise be difficult to characterize from ground observations alone. Airborne lidar, in which a laser aboard an airplane acquires point cloud data measuring 3D vegetation structure, is particularly well-suited for measuring and mapping fuel loads.
Relatively few studies, however, have demonstrated how airborne lidar can be used to map surface fuels, where most fuel available for burning resides. Researchers and collaborators created statistical models predicting field-observed fuel loads from airborne lidar and satellite fire history metrics. Models predicting canopy and surface fuels were 39-59 percent accurate, considerably higher than prediction accuracies of most previous studies that have attempted to predict surface fuels from airborne lidar. Fire history metrics, derived from satellite remote sensing, helped increase prediction accuracy by accounting for surface fuel accumulation over time.
These models, which were trained with fuel measurements at field plots, were then applied across the entire Kaibab Plateau, where lidar data were available, to create canopy and surface fuel load maps for the years 2012, 2019, and 2020. In a space-for-time approach, researchers calculated surface fuel load accumulation rates for several forest types that occur on the plateau. Also, by differencing 2019 and 2020 fuel load maps, researchers were able to estimate fuel consumption totaling around 150 Gg for the 2019 Castle and Ikes Fires combined. Airborne lidar can be used to map both canopy and surface fuel loads across landscapes with moderate accuracy. By combining fuel load maps with fire history information, RMRS researchers accounted for surface fuel load accumulation through time. Multiple lidar acquisitions over the same area make it possible to map fuel loads both before and after fire, so that fuel consumption from fire can be estimated.