Multifrequency and multipolarization radar backscattering from mangrove forests

In this paper, we examine the dependence between multifrequency/multipolari zation synthetic aperture radar (SAR) data and mangrove forest parameters. AIRSAR data at P-, L-, and C-band were acquired over French Guiana in June 1993. Structural parameters, namely, tree height, tree diameter at breast h eight (DBH), tree density, basal area, and total aboveground biomass were e stimated in 12 stands representative of different mangrove forest stages. A t the study site, total biomass ranges from about 5 to 437 tons of dry matt er per ha (t DM ha(-1)). Strong relationships were found between most fores t parameters and radar data, with P-HV showing the greatest sensitivity to total biomass, Estimation of the forest parameters was performed through mu ltiple stepwise regression techniques. While sigma(o) at P-HV saturated at about 160 t DM ha(-1), the use of several frequencies and polarizations all owed total biomass to be accurately estimated up to about 240 t DM ha(-1), Furthermore, the use of polarization ratios at different frequencies has pr ovided useful information about the penetration capability of the radar wav e as well as the dominant mechanisms occurring between the radar wave and t he mangrove canopy.