Ghost images in azimuth
Posted: Thu Aug 13, 2026 5:23 pm America/New_York
I was recently looking at some NISAR data over Mexico City. Specifically, I was looking at the descending data from frame 79 and track 113. In the provisional data release, I was noticing some significant anomalies in the amplitude images. There appear to be ghost images of bright targets appearing periodically in azimuth. That is, for a given bright target, I can see a similar image appear offset by a set number of points in amplitude, with 4+ images visible, each of lower amplitude. I can see these after downloading and processing RSLC level data, and in the preview images available on ASF vertex. The images with this problem include the images from 20251118, 20251212, 20260105, 20260306, 20260411, 20260423, and 20260704.
The 20251118 image is particularly interesting because it is available as a provisional data release and as a beta data release. I have attached cropped, downsampled amplitude of both the provisional and beta releases. This azimuth ghosting issue is clearly present in the provisional data release but does not seem to be present in the beta data release. I also noticed that when run through GAMMA's par_NISAR_RSLC command, the output par files are different in both the single stated prf and the doppler polynomials. All of the bad scenes have higher prf values in the GAMMA produced par file than the scenes without this problem. The provisional file I used is NISAR_L1_PR_RSLC_005_113_D_079_4005_DHDH_A_20251118T011330_20251118T011405_P05023_N_F_J_001.h5 and the beta file is NISAR_L1_PR_RSLC_005_113_D_079_4005_DHDH_A_20251118T011330_20251118T011405_X05009_N_F_J_001.h5 .
Would somebody be able to check if there has been an issue with focusing the data for these scenes?
The 20251118 image is particularly interesting because it is available as a provisional data release and as a beta data release. I have attached cropped, downsampled amplitude of both the provisional and beta releases. This azimuth ghosting issue is clearly present in the provisional data release but does not seem to be present in the beta data release. I also noticed that when run through GAMMA's par_NISAR_RSLC command, the output par files are different in both the single stated prf and the doppler polynomials. All of the bad scenes have higher prf values in the GAMMA produced par file than the scenes without this problem. The provisional file I used is NISAR_L1_PR_RSLC_005_113_D_079_4005_DHDH_A_20251118T011330_20251118T011405_P05023_N_F_J_001.h5 and the beta file is NISAR_L1_PR_RSLC_005_113_D_079_4005_DHDH_A_20251118T011330_20251118T011405_X05009_N_F_J_001.h5 .
Would somebody be able to check if there has been an issue with focusing the data for these scenes?