Monitoring forest foliar moisture using sentinel-2 reflectance and radiative transfer model inversion in Australia

Foliar moisture content (FMC) of tree canopies is important for understanding many aspects of biodiversity, including habitat quality for forest-dependent fauna and fire risk due to fuel conditions. In this article, we modeled spatiotemporal variation in FMC using optical remote sensing. We inverted a radiative transfer model (RTM) on Sentinel-2 satellite reflectance data of 20-m ground resolution to retrieve FMC, and then predicted FMC at subcontinental scales using a random forest (RF) regression emulation of the RTM model. An RTM look-up table was filtered by ecological criteria from 24 southeastern Australian sites sampled along precipitation and canopy cover gradients and was passed to the spectral angle, merit function. The RF emulator allowed efficient computation and presentation of a forest FMC data cube that is novel in its large extent and high spatial resolution, and in its direct and immediate utility for research and management in animal, fire, and plant ecology. Our RTM-based, predictions of FMC had a root mean square error (RMSE) of 19.9% dry matter content and explained more than 60% of variance (r2 = 0.62). This represents an improvement over previous models using reflectance data of coarser spatial resolution, particularly in terms of explained variance (r2 = 0.17 and RMSE = 32%). The emulator model achieved similar performance to the RTM with slightly larger error (RMSE = 21.77%) and smaller explained variance (r2 = 0.54). Overall, both models performed best in canopies of moderate to high density (>∼0.75 LAI). This article demonstrates that FMC can be monitored at spatial resolutions that allow intra or interlandscape patterns to be resolved surpassing previous capabilities.