Baseline-Dependent Retrieval of Underlying Terrain and Forest Height From Hongtu-1 Bistatic X-Band InSAR

This study presents a baseline-dependent assessment of underlying terrain and forest height retrieval using bistatic interferometric synthetic aperture radar (InSAR) data from the Hongtu-1 mission, the first spaceborne X-band cartwheel constellation providing six different baselines. An improved X-band few-look InSAR phase height histogram method is applied, in which the underlying terrain is inferred statistically by accounting for the vegetation-induced phase center bias, while forest canopy height is retrieved through inversion of a simplified physical scattering model. To overcome the limited availability of calibration samples in previous X-band few-look studies, sparsely yet extensively distributed spaceborne LiDAR observations from Global Ecosystem Dynamics Investigation and Ice, Cloud, and land Elevation Satellite 2 are incorporated to calibrate the key histogram model parameter. This enables a systematic examination of the previously adopted constant proportionality assumption between the phase center bias and the histogram width over a wide range of tropical forest canopy heights. In addition, the multibaseline configuration of Hongtu-1 allows a quantitative evaluation of baseline-dependent inversion performance, identifing an optimal range of baseline for reliable parameter retrieval. The proposed approach is demonstrated over the Tapajós National Forest in the Brazilian Amazon. Validation against spaceborne and airborne LiDAR measurements shows that the underlying terrain can be retrieved with an RMSE of 5.46 and 3.6 m, respectively, while forest canopy height is estimated with an RMSE of 6.82 m for tropical forests up to 45 m in height (average height of 30 m; relative error of about 20%). These configuration- and site-specific results confirm the data quality of Hongtu-1 bistatic X-band InSAR and highlight the importance of baseline selection and LiDAR-constrained calibration for robust retrieval of forest structural parameters.