Spatially Quantifying Forest Loss at Landscape-scale Following a Major Storm Event

Abstract

Large scale forest disturbances are becoming more frequent across the world, and remote sensing must play a role in informing and prioritizing immediate, short-term and long-term disaster response and recovery. However, such evaluations from remote sensing are currently limited (e.g., burned area severity and change NDVI) and do not always explicitly relate to change in resources of interest. Herein we demonstrate a novel method to predict basal area loss, validated by independent field evaluations. Hurricane Michael made landfall on Mexico Beach in the Florida panhandle as a Category 5 storm on October 10th, 2018. The storm affected roughly 2 million hectares of largely forested land in the area. In this study, we use Sentinel-2 imagery and 248 forest plots collected prior to landfall in 2018 in the forests impacted by Hurricane Michael to build a general linear model of tree basal area across the landscape. The basal area model was constrained to areas where trees were present using a tree presence model as a hurdle. We informed the model with post-hurricane Sentinel-2 imagery and compared the pre- and post- hurricane basal area maps to assess the loss of basal area following the hurricane. The basal area model had an r-squared value of 0.508. Plots were revisited to ground truth the modelled results; this showed that the model performed well at categorizing forest hurricane damage. Our results validate a novel method to create a landscape scale spatial dataset showing the location and intensity of basal area loss at 10-m spatial resolution which can be used for quantifying forest disturbances worldwide.