Moisture content estimation of forest litter based on remote sensing data

As a fine fuel, forest litter plays an important role in fire danger rating systems, so forest litter moisture data are necessary and meaningful for fire risk management and prevention. An optical remote sensing technique can provide continuous and regional data for litter moisture estimates, but such an approach is restricted in separating the background reflectance of the forest floor from pixel reflectance because the litter moisture information is included only in background reflectance while pixel reflectance in the forest area consists of both canopy reflectance and background reflectance. Therefore, we present a geometrical-optical model to estimate forest litter moisture by separating contributions of background reflectance from the remote sensing image and use a statistical model to estimate the forest litter moisture content based on the calculated background reflectance. The results show that the model had an R2, root mean square error (RMSE), and average precision of 0.595, 0.372, and 69.654%, respectively. This approach provides a new way of estimating forest litter moisture content from an optical remote sensing image, and it can potentially be applied in large-scale forest litter moisture content mapping.