Speaker
Description
The European Space Agency’s BIOMASS mission provides the first global observations of forests at P-band, offering unprecedented sensitivity to vegetation structure and above-ground biomass. While primarily designed for biomass estimation, the sensitivity of P-band Synthetic Aperture Radar (SAR) observations to the vertical structure of vegetation also opens new opportunities for detecting forest disturbances. Such disturbances are defined here as structural deviations from undisturbed forest, including reduced vegetation density caused by natural processes such as fires and insect outbreaks or anthropogenic activities such as selective logging and fragmentation, as well as changes in stand height and residual vegetation left on the ground following forest clearing.
This study demonstrates the potential of BIOMASS observations for identifying such forest disturbances. The proposed approach combines polarimetric and tomographic information to characterize the presence and vertical distribution of volume scattering associated with vegetation. Polarimetric information is exploited through the polarimetric volume indicator, $v_{POL}=I_{VH}/span$ which provides sensitivity to the presence of volume scattering within the observed scene. This information is complemented by the normalized vertical power distribution derived from SAR tomography, $v_{TOMO}=P(z_{30})/\sum_iP(z_i)$ which quantifies the relative amount of tomographic power located at 30-m height. The combination of these complementary descriptors enables the classification of a scene into stable forest, bare soil, water, and forest disturbance classes.
Single-stack observations are first exploited to demonstrate the sensitivity of the proposed descriptors to different disturbance conditions, while the final forest disturbance product is designed to exploit multi-temporal BIOMASS observations. Preliminary validation based on visual interpretation of high-resolution optical imagery over sites distributed across the humid tropics, including Brazil, Colombia, French Guiana, and the Democratic Republic of the Congo, demonstrates the potential of P-band observations to identify a range of forest structural disturbances. Furthermore, the exploitation of BIOMASS IOC data provides an initial demonstration of the multi-temporal capabilities of the approach for mapping incremental disturbance.
These results highlight the capability of BIOMASS to extend beyond its primary biomass-mapping objective and provide a novel P-band perspective on forest disturbance monitoring, with sensitivity to structural changes that may remain difficult to characterize using conventional optical and shorter-wavelength SAR observations.