View article

1. Improvement of effective baselines by combining interferograms Topographic corrections for our interferograms were calculated using a digital elevation model (DEM) constructed from an ERS tandem pair of SAR images with a perpendicular baseline of-129 m, and calibrated with the SRTM30 global topographic dataset to remove long-wavelength tilts and warps. One drawback of using such a DEM is the possibility of incorporating atmospheric noise in the topographic signal. As the baseline of the coseismic pair of images for the Bam earthquake was over four times larger than that of the tandem pair used to generate the DEM (Table 1), this meant that the effects of any atmospheric noise present in the DEM would be carried over into the ‘corrected’coseismic interferogram, and amplified by the ratio of baselines. To improve the baseline of the coseisimic interferogram, and therefore mitigate this effect, we use the result of Massonnet and Feigl [1998] whereby the difference image of two interferometric pairs is shown to have an effective baseline equal to the difference of the two baselines of the two pairs. A preseismic interferogram (Table 1), corrected for topography using our tandem pair DEM, and with a perpendicular baseline of~ 450 m, was subtracted from the~ 570 m-baseline coseismic interferogram (Table 1), also corrected with our DEM. The resulting differenced interferogram had an effective baseline of~ 120 m, smaller than the baseline of the tandem pair, and contained an appreciably reduced component of noise due to atmospheric errors in the DEM when compared with the undifferenced coseismic interferogram. This method differs …