U.S. Joint Fire Science Program


Date of this Version


Document Type



JFSP Project Number: 10-1-02-14


US government work.


Our research used light detection and ranging (LiDAR) systems coupled with sequential harvesting of Pitch pine (Pinus rigida Mill.) to quantify canopy fuels in three dimensions across a large, heterogeneous landscape impacted by multiple wildfires, prescribed burns and insect defoliation events. We used a three-tiered approach; 1) calibration of upward sensing profiling LiDAR data with sequential harvesting of 20 x 20 meter plots to quantify the mass of foliage, branches and stems in Pitch pine canopies in 1-meter height layers, 2) scaling results to the landscape scale using previously-published relationships between upward sensing and downward sensing scanning LiDAR systems in similar Pitch pine stands, and 3) evaluation of predicted canopy fuel loading using an independent set of 20 x 20 meter field plots. Five 20 x 20 m plots were harvested, ranging in total live tree biomass from 67 to 108 Mg ha-1. Crown fuel weight (CFW; kg m-2) ranged from 0.83 to 1.16 kg m-2, and maximum canopy bulk density (CBD; kg m-3) ranged from 0.15 to 0.23 kg m-3. Allometric relationships between parabolic bole volume, calculated from height and DBH measurements, and available fuels, needle mass, and 1-hour and 10-hour fuels were highly significant, with regression coefficients ranging from 0.89 to 0.91. Regression coefficients calculated for maximum CBD and its height using biometric data were 0.81 and 0.72, respectively. Relationships between upward-sensing profiling LiDAR returns and available fuels, needle mass, and 1-hour and 10- hour fuels biomass of canopy fuels were highly significant, and regression coefficients were > 0.9 between crown fuel weight or maximum CBD and LiDAR returns. Across all equations, the poorest fits were for 1000-hr fuels and dead needle mass. Previous research has demonstrated that relationships between upward sensing profiling LiDAR and downward sensing scanning LiDAR are highly significant in Pitch pine – dominated stands, facilitating the scaling of crown fuel estimates across the landscape. Initial analyses of data from validation plots indicate that biometric and LiDAR-derived estimates of CFW and maximum CBD are not significantly different; CFW estimates were 1.15 ± 0.27 vs. 1.22 ± 0.28 kg m-2 (n = 17, Paired-sample T = 0.22, ns) and maximum CBD estimates were 0.22 ± 0.08 vs. 0.22 ± 0.08 kg m-3 (n = 17, Paired-sample T = 0.59, ns) for biometric and LiDAR-derived estimates, respectively. The results of our project will assist state and federal wildland fire managers, because highly accurate canopy fuel maps can be produced for large forested areas in the Pinelands, and for areas in and near wildland-urban interface. Our results can also be used to evaluate the effectiveness of prescribed burns and mechanical canopy fuel reduction treatments. In addition, we can now generate highly accurate estimates of crown bulk density (CBD) and other canopy fuel characteristics, which are appropriate for current fire behavior models such as the FVS-Fire and Fuels Extension, and for the next-generation of fire behavior models such as WFDS, which require high resolution canopy fuel loading information.