Combination of ascending/descending ERS-1 InSAR data for calibration and validation

Reliable information extraction using repeat-pass ERS-1 InSAR becomes problematic in foreshortened and layover areas due to the severely degraded ground resolution on facing slopes. Variable scattering area sizes in sloped areas also lead to a requirement that calibrated SAR backscatter values use a height model for compensation. Combination of ascending and descending datasets supplies at least one non-laidover dataset for almost all scene locations. Transformation of synthetic aperture radar (SAR) data into a map projection geometry (terrain geocoding) makes the ascending/descending overlay possible. The terrain geocoding process transforms slant (or ground) range SAR data into a map geometry suitable for intercomparisons. In addition, a “synthetic” interferogram calculated from a DEM is used to “flatten” a raw interferogram, enabling coherence estimation without contaminating slope effects. Provided that forest stands remain coherent, the technique also appears to be able to delineate forest stand boundaries. It may also possibly give an indication of tree height Selection of the highest available local range resolution from ascending and descending datasets significantly improves the effective mean ground resolution of the geocoded images. Application of this optimal resolution approach (ORA) is especially important in areas with severe terrain variations. Knowledge of the local slope enables application of a radiometric correction to the backscatter, significantly improving thematic interpretability in sloped regions