Thu, Aug 18, 2022: 8:30 AM-8:45 AM
518C
Background/Question/MethodsUnderstory fuels play an important role in wildfire spread and severity. It is important to quantify changes in fuels before and after wildfires across multiple spatial scales. Developments in LiDAR remote sensing present new avenues for quantifying these changes across different forest types. Here, we evaluated the use of handheld mobile LiDAR (HMLS) to measure surface fuels in Northern California redwood forests and oak woodlands. We aimed to validate the use of HMLS to estimate surface fuel mass and determine the ideal voxel size for the most accurate measure of mass. We hypothesized that voxelized HMLS data will accurately measure surface fuel volume and smaller voxel sizes will better estimate dry mass. First, we built a 0.5x0.5x1 m PVC frame, placed it over vegetation, and scanned both the frame and vegetation with the HMLS. Once scanned, vegetation in the frame was visually assessed for occupied volume, destructively sampled, and dry mass was calculated. HMLS data were processed using Lidar360 and R. Leave one out cross-validation with an expected linear model was used to assess the relationship between field and voxel occupied volume as well as the voxel size with the least error for predicting vegetation dry mass.
Results/ConclusionsHMLS voxels had a strong positive, linear relationship with field assessed occupied volume (R2 = 0.72, RMSE = 1.71). The comparisons of dry mass estimated using various voxel sizes suggested that 1 cm voxels resulted in the lowest error (R2 = 0.46, RMSE = 50.77). However, the effect size of this relationship is not as strong as that of the field vs voxel occupied volume. Preliminary results suggest that vegetation type may affect how well voxel count measures vegetation: frames that had higher amounts of woody material were further from the regression line. The application of HMLS in this study presents an opportunity to expand the range of tools that land managers and conservationists could use to quantify and understand understory vegetation in fire-dependent ecosystems. In the future, we look forward to using this approach to compare surface fuel changes pre- and post-management.
Results/ConclusionsHMLS voxels had a strong positive, linear relationship with field assessed occupied volume (R2 = 0.72, RMSE = 1.71). The comparisons of dry mass estimated using various voxel sizes suggested that 1 cm voxels resulted in the lowest error (R2 = 0.46, RMSE = 50.77). However, the effect size of this relationship is not as strong as that of the field vs voxel occupied volume. Preliminary results suggest that vegetation type may affect how well voxel count measures vegetation: frames that had higher amounts of woody material were further from the regression line. The application of HMLS in this study presents an opportunity to expand the range of tools that land managers and conservationists could use to quantify and understand understory vegetation in fire-dependent ecosystems. In the future, we look forward to using this approach to compare surface fuel changes pre- and post-management.