Spaceborne Single Pass Pol-InSAR For Monitoring Forest Vertical Structure: Recent TanDEM-X Experiments and Future Tandem-L Perspectives

Abstract

Forest vertical structure is widely recognized to be an indicator of the above ground biomass, whose knowledge is very important to understand the global carbon cycle. Moreover, vertical structure allows the characterization of the state of a forest ecosystem in terms of biodiversity, degradation and regeneration. Within this framework, the employment of space borne synthetic aperture radar (SAR) systems results to be particularly suitable as they can acquire data continuously with large coverage, beside the possibility to reach areas in which extensive and continuous ground measurement are difficult, if not impossible. In particular, the coherent combination of single- and multi-baseline SAR data at different polarizations by means of polarimetric interferometric (Pol-InSAR) and tomographic (Pol-Tomo) techniques provides sensitivity to the vertical distribution of scattering processes and allows their characterization. However, in repeat pass spaceborne implementations, the inherent presence of temporal decorrelation degrades such sensitivity and limits the performance in forest monitoring. This is the main reason why polarimetric space borne missions could not contribute either with respect to large scale demonstrations of Pol-InSAR and Pol-Tomo techniques or for the development of new techniques and products. The acquisition and analysis of Pol-InSAR data without the disturbing effect of temporal decorrelation on a global scale is now possible by means of the TanDEM-X (TDX) mission, which provides for the first time single pass (single- and dual-) polarimetric interferometric data from space. After two years of missions, time series with variable baseline over the same forest site are available, allowing to (1) explore their information content, (2) assess penetration capabilities, (3) assess scattering model assumptions, and (4) estimate vertical structure and monitor its dynamics. Recent TDX experiments for forest applications have demonstrated rather surprising penetration capabilities of X-band radar signals into volumes. Different scattering behaviours have been observed for different seasons, for instance between summer and winter in the case of boreal forests and between wet season and dry season in the case of tropical forest systems. Moreover, Pol-InSAR forest height inversion has been demonstrated to be possible using either dual pol acquisitions or a single pol acquisition in combination with an external terrain DEM. Combining forest height and allometric equations, the possibility has also been shown of generating high resolution biomass classification maps of boreal forests improving previously available thematic maps. Very recently, first encouraging experiments have been carried out in the combination of multiple single-pass interferogram by using a dataset on a tropical forest, and potentials for a richer 3-D analysis of forest scattering have been individuated. In this work, new experimental results will be presented with the objective of further exploring the potentials of TDX for monitoring the forest structure in different ecosystems. Contributions of multiple baselines and polarimetry will be further investigated. Special emphasis will be set on the evaluation of time series data, with reference also to the robustness of the inversion algorithms against variable seasonal and environmental conditions. Moreover, future perspectives on the systematic monitoring of forest dynamics will also be addressed. In this framework, a prominent role is played by the German mission proposal Tandem-L. The Tandem-L mission concept consists of two L-band cooperating satellites flying in close formation and using advanced techniques and technologies. Beside single-pass interferometric imaging, a more frequent revisit time and novel high-resolution large-swath imaging modes, forest observation will benefit from L-band radar signals, which are a richer source of information than X-band signals as they can penetrate and interact with different vegetation layer until (almost) the ground. For this reason, Tandem-L is expected to provide estimates of forest height, ground topography, vertical structure and biomass as basic products.