Estimating tree aboveground biomass using multispectral satellite-based data in Mediterranean agroforestry system using random forest algorithm

Forest aboveground biomass (AGB) is a key biophysical variable to assess and monitor the spatio-temporal changes of forest ecosystems. AGB should be accurately timely estimated through remote sensing provide valuable information better support sustainable management strategies. QuickBird WorldView-2 satellites data Random (RF) regression model were used estimate tree in Mediterranean agroforestry systems. Spectral bands, vegetation indices Grey-Level Co-occurrence Matrix (GLCM) texture features 140 plots with without mask as independent variables, while total per plot was dependent variable. A good performance obtained for complex system, an R2 82.0% RMSE 10.5 t/ha (22.6%). The top 11 most important variables have 80.3% relative importance, 59.6% GLCM textural features, 12.3% 8.4% spectral bands. results highlight importance texture, use RF collect accurate spatial on crown cover attributes, by excluding contribution understory soil characteristic,