Retrieval of forest aboveground biomass through the synergy of X-band (TeraSAR-X/TandDEM-X) and L-band (PALSAR-2) InSAR data

Forest parameters, especially the forest aboveground biomass over large area and in higher accuracy, become more and more important for the research on global climate change and carbon cycle [1]. Synthetic Aperture Radar (SAR) was considered as an effective tool for assessing forest aboveground biomass because it could penetrate farther into forest canopies, especially at longer wavelength such as L-and P-bands. In 1990's, many researchers investigated the relationship between the forest aboveground biomass and radar backscattering coefficients under various conditions through the campaigns of airborne and space-borne SARs [2–4]. The basic conclusions from these researches were that (1) cross-polarization is more sensitive than co-polarization; (2) The saturation point increased as the wavelength becomes longer [5]. Besides of radar backscattering coefficients, another variable used for the estimation of forest aboveground biomass is coherence since 2000's [6]. Coherence was initially used to describe the quality of interferometric SAR image pairs. Some researchers found an interesting phenomenon that forest stands with higher biomass level always showed lower coherence, i.e coherence is negatively correlated with forest aboveground biomass[7].