A Comparison of P- and L-Band PolInSAR 3-D Forest Structure Estimates: A Study Case in the Traunstein Forest

Kurzfassung

Forest vertical structure is a key variable for assessing biodiversity, structural degradation and/or regeneration, and for the development of accurate biomass estimators. Moreover, the knowledge of forest structure changes is important to determine the type and the intensity of forest disturbance as well as the seasonal and annual forest cycle variation. In this framework, the next generation of low-frequency spaceborne SAR configurations will play a crucial role in forest structure monitoring as they will provide the unique ability to image large areas at high spatial and temporal resolution. Among the different SAR techniques that have been developed in the last decades, SAR Tomography (TomoSAR) allows the estimation of the 3-D distribution of the backscattered power from multiple SAR images, thus providing a proxy to forest vertical structure. Although TomoSAR is an established technique with several algorithmic implementations a still underdeveloped element is the physical interpretation of tomograms. The link between the obtained tomograms (that depend in general on frequency, polarisation and acquisition geometry) and physical forest structure parameters is essential for establishing potential applications and is not well understood yet. The enhanced observation scenario provided by SAR sensors on one side and the need for 3-D structure products for ecosystem modeling on the other, motivate the exploration of the information content of multi-baseline polarimetric interferometric measurements for the retrieval of physical forest structure parameters by means of TomoSAR techniques. As different carrier frequencies are in general sensitive to different structure elements, the comparison of the related tomograms represents an additional way to assess scattering model assumptions and to advance in their physical interpretation. With this in mind, the objective of this work is to compare TomoSAR vertical profiles of the backscattered power obtained at L- and P-band over the forest site of Traunstein. The forest of Traunstein is a temperate mixed mountain forest and is located in south-eastern Germany. Dominant tree species are European spruce and Silver fir accompanied with small fractions of European beech. It is a managed forest composed of even-aged stands which cover forest heights from 10m to 40m. Mean biomass level is on the order of 210t/ha, while some old forest stands can reach biomass levels up to 600t/ha. Compared to other managed forests in this ecological zone the biomass values of Traunstein forest are significantly higher. This very high ecological variety, together with the availability of detailed ground measurements, make the Traunstein forest an interesting study case. Multibaseline fully polarimetric SAR data were acquired by the DLR’S E-SAR platform during the TempoSAR 2009 campaign. The tomographic focusing will be carried out by means of (model-based) high resolution techniques in order to increase the vertical resolution as much as possible, as well as to overcome the imaging limitations intrinsic to a low number of baselines. By extensively using ground and airborne LiDAR measurements, the key processes and structure elements will be identified in each band, and their influence on the scattering will be discussed. To this purpose, different polarimetric channels will be combined. The capability of separating between ground and volume scatterers by using multibaseline PolInSAR models will be evaluated for the two frequencies. In addition, the structure information content of a full-pol configuration will be compared with a dual-pol one. It is expected that these experiments will allow to asses (1) the physical information content of L-band and P-band, and, therefore, (2) their synergies and complementarities towards the derivation of structure-related products. At the same, implications on SAR system design and acquisition strategies will be outlined for the derivation of forest products at a global scale.