Answers are grounded in this paper and cite their sources.
Demonstrates forest height and ground elevation mapping in tropical forests, suggesting improved accuracy in dense forests.
This paper presents a method to estimate ground elevation and forest height using spaceborne lidar data guided TanDEM-X Synthetic Aperture Radar (SAR) Interferometry (InSAR) with complete modeling analysis of canopy penetration capability. The InSAR phase height histogram is an efficient means for characterizing vertical forest structure by assigning each pixel of InSAR phase height to a specific height bin. Taking into account the penetration constraints of high-frequency spaceborne InSAR in dense forests, an extended modeling analysis of the statistical behavior of InSAR phase height histograms reveals one key insight: the standard deviation of the InSAR phase height histogram as a key radar-driven feature of forest vertical structure for both detecting penetration-limited regions in tall tropical forests and assisting ground finding. Leveraging this insight and integrating the spaceborne lidar samples for calibration, we develop an innovative processing workflow for the retrieval of ground elevation and forest height, which is designed to accommodate varying lidar sampling densities and penetration conditions using three strategies. Specifically, lidar-measured terrain elevation and the phase height histogram’s standard deviation is utilized to identify penetration-limited regions and then guide ground finding process. In areas with penetration limitations, forest height is more accurately estimated by directly extracting a lidar-like relative height metric from the histogram, while in full-penetration areas, forest height is retrieved using the Random Volume over Ground (RVoG) model. The theoretical modeling analysis and processing framework were validated across three representative sites (i.e., Tapajós, Howland, and Lopé forests) under different climate and penetration conditions. Across these sites, ground elevation was best estimated with accuracies of 3.16 m, 2.32 m, and 7.37 m, and the forest height was best inverted with RMSEs of 4.52 m, 3.56 m, and 7.17 m at one-hectare pixel size. This study presents a lightweight method for retrieving forest biophysical parameters from single-baseline, single-polarization X-band InSAR data. The developed model and approach also show promise for enhancing biophysical parameter retrieval in tropical regions using X-band InSAR data and for improving the BIOMASS mission’s coarsely gridded biophysical products through integrating fine-scale information, provided penetration biases are corrected with BIOMASS Digital Terrain Model (DTM) products.
No takes yet. Share an insight, caveat, or question.
Yu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: