Terrain-Guided Block Adjustment for InSAR DSM Elevation Calibration

The block adjustment of interferometric synthetic aperture radar (InSAR) digital surface model (DSM) is a crucial step in elevation calibration. Local slope and aspect can directly affect the occurrence of geometric distortions in InSAR DSM, such as foreshortening, shadow, and layover, severely decreasing elevation accuracy. Planimetric geolocation errors can lead to larger elevation errors, particularly in areas with significant terrain variation. Therefore, this study proposes a block adjustment method that considers local terrain information. Firstly, terrain information such as slope and aspect are introduced into the construction of the elevation error model. Next, the new model is used to construct the error equation in block adjustment and the least squares method is used to solve the parameters and correct the elevation. Finally, twelve pairs of TanDEM-X co-registered single-look slant range complex (CoSSC) images covering an area of approximately 1° × 1° in Guangdong, China, are used for experimental validation. Filtered Ice, Cloud, and land Elevation Satellite (ICESat)-2 points are used as ground control points (GCPs) and checkpoints, whereas Copernicus DSM30 is used for qualitative assessment. The results indicate that (1) the root mean square error (RMSE) of the adjusted DSMs elevation error decreases from 3.09 m to 2.60 m after the introduction of the terrain information, resulting in an improvement in accuracy of approximately 15%. (2) The new method that considers terrain information outperforms the traditional polynomial model across different slope ranges. These results demonstrate the superiority of the proposed method.