Enhancing global rainfall interception loss estimation through vegetation structure modeling

Accurate estimates of rainfall interception loss are crucial for modeling the water balance of forested areas. However, considerable regional variability exists in the interception process, and much uncertainty remains. This study enhances the estimation of rainfall interception loss at the global scale by integrating remote sensing products into the parameterization of Gash's analytical model. We refer to this enhanced configuration as the Global Interception Model (GIM). High-resolution satellite imagery was used to derive vegetation indices and spectral reflectance, which were then employed in linear regression models to estimate canopy cover fraction (c) and vegetation water storage capacity (Sv). Their importance in ecological processes, and land and water resource management, makes the modeling of these parameters of particular interest. The other two parameters required to run the Gash model, namely the mean rainfall and evaporation rates under saturated canopy conditions, were obtained via the integration of MWSEP and ERA5-Land meteorological products. Modeling performance was evaluated using in situ measurements and gridded datasets. GIM estimates exhibited a superior performance statistic when compared to PMLv2 and GLEAMv3.7a. Our results demonstrate the high potential of this approach for improving the accuracy of rainfall interception loss estimates from local to global scales.