Geocoding Refinement via Planar Block Adjustment for Spaceborne SAR Imagery in Low-Relief Terrain: A Case Study of LuTan-1

The geolocation accuracy of spaceborne synthetic aperture radar (SAR) imagery is crucial for a wide range of applications, yet it is often degraded by errors in onboard parameters, making accurate geocoding refinement essential. Conventional refinement methods typically involve coregistering SAR images with digital elevation model (DEM)-simulated SAR images (DSSIs) to extract planar control points (PCPs), which are then utilized to construct polynomial models for correcting geolocation errors. However, a key drawback of this approach lies in its strong reliance on slope variations in the reference DEM to generate textural features in the DSSIs. Consequently, in low-relief terrains such as plains and deserts, the coregistration process often produces an inadequate number of unevenly distributed PCPs, leading to reduced refinement accuracy. When a standard SAR scene provides limited coverage, planar block adjustment (PBA) is commonly applied to integrate PCPs and propagate their constraints through tie points, alleviating the limitations of single-scene refinement. Nevertheless, most PBA techniques assume prerefined single scenes as a prerequisite and aim to compensate for residual errors from the initial refinement, rendering them ineffective in low-relief terrains. To overcome this limitation, we propose a robust PBA method that directly achieves holistic geocoding refinement, bypassing the need for error compensation. Experiments using LuTan-1 (LT-1) bistatic InSAR data from three representative test sites—including plains and deserts—demonstrate the effectiveness of the proposed approach. Our method consistently attained subpixel geolocation accuracy, overcoming the limitations of conventional PBA methods. In terms of elevation accuracy, the resulting LT-1 DEM achieved a root mean square error better than 2 m across all test sites, outperforming the Copernicus DEM in the same regions.