Estimating forest structure change by means of wavelet statistics using TanDEM-X datasets

Abstract:
Understanding forest structure is crucial because it is shaped by forest dynamics and biophysical processes, and provides valuable insights into a forest's state and dynamics when monitored globally. The scale of observation is critical in understanding forest structure, as disturbances tend to homogenize patterns across different scales, while fine-scale processes introduce heterogeneity. Radar-based techniques like SAR interferometry and SAR tomography offer ways to characterize forest structure, primarily by analysing variations in radar reflectivity profiles. However, conventional methods, such as phase histograms, sacrifice spatial resolution. TanDEM-X, with its high spatial resolution allows the characterisation of structural heterogeneity by examining the centroid of the reflectivity profile. To fully capture structural variations, a multi-scale analysis using wavelet frames is applied. This research combines experimental SAR data from Tan- DEM-X and forest simulations to comprehensively assess and understand forest structure.