Analysis of seasonal effects on forest parameter estimation of Indian deciduous forest using TerraSAR-X PolInSAR acquisitions
Résumé
TerraSAR-X/FanDEM-X form the first space-borne interferometer providing full-PoIInSAR acquisitions with a near-zero temporal gap. This paper explores the potential of X-band SAR data for forest height estimation over tropical Indian forests. Unique TanDEM-X data sets with varied spatial baselines are acquired over a tropical, deciduous Indian forest Multiple dual polarimetric and fully polarimetric acquisitions are analysed and the best performance is achieved using fully polarimetric data acquired over both high- and low-biomass forests (r(2) =0.86 and r(2)=0.80 respectively). It is shown for the first time that, for deciduous Indian tropical forests, seasonal changes such as leaf-fall affect the PoIInSAR based height inversion performance. For species exhibiting leaf-fall, acquisition during leaf-on season provides better height inversion accuracy (r(2)=0.75) than for leaf-fall season (r(2)=0.65). Results show consistent PoIInSAR height inversion performance for both high- and low-biomass forests in both natural and managed forest tracts. Spatial baseline is a critical factor for selection of suitable acquisitions for PoIInSAR based height estimation. The effect of spatial baseline on height inversion performance is also discussed. (C) 2017 Elsevier Inc. All rights reserved.