Investigation of the Possibilities of H-α Decomposition for Dual Polarization in Radar Monitoring of Afforestation

Abstract

Assessment of the processes of afforestation and restoration of forests after fires is relevant for a significant territory of Russia, including the problem of carbon neutrality. The paper considers the possibilities of radar monitoring of the afforestation process based on the Cloud-Pottier decomposition of L-band data time series with dual polarization. Preliminary segmentation is based on the minimum values of the radar backscatter over the entire observation period. This makes it possible to distinguish treeless areas and sparsely wooded areas into a separate class, both existing before the start of the study and formed later. Next, Cloud-Pottier polarimetric decomposition is performed to obtain the parameters H (entropy) and α (scattering angle) and form time series from them. Studies have shown the principal possibility of afforestation dynamics monitoring on the H-α plane, where the points of the test areas form characteristic time tracks. A mature dense forest, whose characteristics are considered permanent, was used as a reference for estimating the changes rate on the H-α plane.