Estimation of forest parameters combining multisensor high resolution remote sensing data
Résumé
Forest monitoring is a major issue to carry out energetic and environmental policies. Actual context in spaceborne remote sensing data is very promising. Our study aims to test the ability of SAR, optical and textural data to estimate forest parameters (biomass, height, diameter and density), and to evaluate the improvement of combining these remote sensing data. We worked on monospecific pine forest stands. The first results highlighted the synergy between SAR and spatial texture informations. Sentinel-1 C-band SAR data is very promising for the estimation of forest parameters in monospecifics stands. Biomass was estimated with 29.4% relative error (20.7 tons/ha) and height with 14.6% (2.1m) combining four SAR and optical sensors.