Boreal Forest Biomass Estimation using Radar Derived Vertical and Morphological Forest Structure Indicators

Radar and lidar remote sensing techniques have demonstrated the ability to estimate the above ground biomass within a reasonable (10%-35%) accuracy level over a wide range of forest types. The conventional local approaches are based on relationships of biomass to polarimetric radar backscatter and to vertical height profiles derived from either lidar or radar interferometry. In the framework of the designated space--borne missions DESDynI and BIOMASS to measure the global biomass (carbon stocks and dynamics), this study examines the potential enhancement of biomass estimation using the model-based vertical and morphological structure information derived from polarimetric SAR interferometry (PolInSAR) together with forest species composition information from optical imagery. The lidar samples the first-contact heights and is therefore most sensitive to the upper layer of the canopy, though still providing information about the lower vertical structure dependending on the density and gap structure. The total backscatter from high and medium frequency radar is mostly dominated by the canopy with a ratio of 0.4 to 0.95[*]. To some extent, using structural functions or regression parameters for the individual tree species makes it possible to relate the canopy extension to the total tree biomass. However, for instance the biggest biomass contribution comes from the stems, where the ratio of stem biomass to total biomass is between 0.71 and 0.83[*]. Polarimetric interferometry height and the ability to estimate the polarimetric signature at arbitrary vertical position provide the necessary sensibility to structure components of the forests. Based on the derived parameters, the biomass estimation is examined in order to evaluate the importance of different indicators and models. The experiments are conducted on the Krycklan catchment boreal forest site in Northern Sweden, where an extensive data set has been acquired and provided by ESA, DLR and FOI. This set consists of ground field measurements, lidar, and optical data, as well as polarimetric interferometric SAR data with several baselines acquired by DLR's E-SAR sensor at L- and P-band frequencies. This research was supported by an appointment to the NASA Postdoctoral Program at the Jet Propulsion Laboratory, administered by Oak Ridge Associated Universities through a contract with NASA. [*] Based on the data for the Krycklan catchment.