Effects of forest structure on radar response to forest biomass

Recent experimental data collected by airborne imaging radars has shown, that for a given forest type, radar backscattering coefficient /spl sigmaspl deg/ has a power-law dependence upon aboveground biomass. These data also demonstrate that this dependence is frequency and polarization dependent. A first-order, vector radiative transfer model, MIMICS, is used to investigate polarimetric backscatter from forests. The model results show that the observed frequency and polarization effects are determined by forest structural properties. The model is used to examine the scattering behaviors of several structurally dissimilar vegetation archetypes: (1) herbaceous, (2) woody shrubs and (3) woody trees. Three structural growth forms of woody trees are investigated: (1) columnar, (2) excurrent and (3) decurrent. The model shows each archetype to yield distinctive scattering behaviors as functions of frequency and polarization. These results are compared to AIRSAR observations of each archetype at P-, L- and C-bands. Structural classification of vegetation is shown to be possible and is found to be an important component of forest biomass retrieval algorithms.<<ETX>>