Assessing post-fire tree mortality and biomass change by integrating LiDAR and hyperspectral data

Accurate characterization of post-fire changes in tree mortality and carbon storage is critical in understanding and simulating future forest responses to wildfires in response to climate change. LiDAR remote sensing has been successfully used to estimate aboveground biomass (AGB) in undisturbed forests, but few studies focused on effects of burn severity on post-fire tree mortality and AGB. Specifically, it is not clear how burn severity would affect forest mortality and post-fire AGB patterns in mixed forests in Eastern U.S. In this study, we examined short-term changes in tree mortality and AGB across a burn gradient in a Pine Barrens ecosystem in the Eastern United States, using field observations, LiDAR and hyperspectral data. Results show that fusion of LiDAR and hyperspectral images can characterize key forest ecosystem attributes at fine spatial and spectral resolutions, and provide consistent and accurate estimation of post-fire forest conditions at landscape scales. Both post-fire mortality and AGB are highly correlated with burn severity, and we found a linear relationship between burn severity and post-fire AGB. Our results can be used to predict post-fire mortality and AGB changes and provide quantitative evidence for informed forest fire management in similar ecosystems in the U.S.