Estimation of Forest Parameters Combining High Resolution Radar and Optical Spaceborne Sensors

Mapping forest resources is a major issue to carry out environmental policies and assessing terrestrial carbon stocks and emissions. The objective of this research was to evaluate the ability of freely available spaceborne radar, optical and textural metrics to estimate forest parameters on different forest types. The Combination of these data using a machine learning approach was able to accurately estimate the basal area and DBH with relative RMSE below 25% on mono-specific plantations as for different mixed broadleaved forests. This method could help to spatialize forest inventories over large areas with a reduced cost and regular updates.