3-D Structure of a Forest at L-band: Experiments Towards a Spaceborne Implementation

Abstract

Due to its ecological importance, the monitoring of the vertical structure of forests is continuously raising the interest of the synthetic aperture radar (SAR) scientific community. In particular, SAR tomography allows reconstructing the 3-D distribution of the radar power backscattered by a forest by combining more than two SAR acquisitions with baseline diversity. Beyond specific estimation algorithms, the link between the estimated tomograms (depending in general on frequency, polarization and acquisition geometry) and physical forest structure is essential for establishing potential applications. The main objective of this seminar is to present and discuss the results of some airborne tomographic experiments, considering forest stands with different biomass levels and different species composition. The analysis has been carried out by processing an L-band multibaseline fully polarimetric data set acquired over the temperate forest of Traunstein (Germany) by means of the DLR’s E-SAR platform, and by exploiting an auxiliary Lidar acquisition and a set of ground measurements. The effects of the weather conditions on the estimated profiles have been analyzed as well by processing data acquired after a dry and a rainy period. It is worth noting that the variability of the profiles with the environmental conditions can be seen as a critical aspect for the implementation of 3-D forest structure monitoring from space. For this reason, considering a single-pass interferometric implementation as proposed for the Tandem-L mission, the impact of the inhomogeneity of the interferometric coherences on structure estimation has been quantified.