HDF5 Statistics Attributes Set to Zero for Multiple RUNW Datasets
Summary
Multiple 2D raster datasets in RUNW HDF5 products have statistics attributes (min_value, max_value, mean_value, sample_stddev) incorrectly set to 0.0, despite the actual data containing non-zero values.
Affected Products
- RUNW: ✗ Affected - 4 datasets have incorrect statistics (see details below)
- RIFG: ✓ No issues found
- GUNW: ✓ No issues found
- ROFF: ✓ No issues found
- GOFF: ✓ No issues found
Affected Datasets in RUNW
| Dataset |
Attr Min |
Attr Max |
Attr Mean |
Attr Std |
Actual Min |
Actual Max |
Actual Mean |
Actual Std |
[..]/interferogram/HH/coherenceMagnitude |
0.0 |
0.0 |
0.0 |
0.0 |
2.45e-05 |
9.37e-01 |
1.59e-01 |
1.23e-01 |
[..]/pixelOffsets/HH/alongTrackOffset |
0.0 |
0.0 |
0.0 |
0.0 |
-6.46 |
1.33e+02 |
2.79e+01 |
3.00e+01 |
[..]/pixelOffsets/HH/slantRangeOffset |
0.0 |
0.0 |
0.0 |
0.0 |
-4.41e+01 |
7.12 |
-8.71 |
1.03e+01 |
[..]/pixelOffsets/HH/correlationSurfacePeak |
0.0 |
0.0 |
0.0 |
0.0 |
-7.17e-02 |
9.08e-01 |
1.66e-01 |
1.91e-01 |
"Attr" metrics are what is written into the HDF5 file's dataset attributes.
"Actual" metrics are computed post-facto by opening the HDF5 file and computing e.g. numpy.nanmin(..) on the full dataset raster itself.
Impact
- Downstream users relying on statistics attributes for quick min/max/mean queries will get incorrect (zero) values
- Impacts data quality assessment and validation workflows
- May affect visualization tools that use these attributes for color scaling
Tested Files
RUNW (✗ Issue found in both sample files):
NISAR_L1_PR_RUNW_007_106_A_140_008_7700_SH_20251211T140849_20251211T140926_20251223T140850_20251223T140926_X05016_N_P_J_001.h5
NISAR_L1_PR_RUNW_020_106_A_140_021_4000_SH_20260516T140853_20260516T140929_20260528T140852_20260528T140928_X05016_N_F_J_001.h5
RIFG (✓ No issues):
NISAR_L1_PR_RIFG_013_118_A_171_014_2000_SH_20260222T102514_20260222T102520_20260306T102515_20260306T102521_X05016_P_P_J_001.h5
GUNW (✓ No issues):
NISAR_L2_PR_GUNW_001_062_A_019_006_2000_SH_20250927T114508_20250927T114542_20251126T114516_20251126T114545_X05016_P_F_J_001.h5
ROFF (✓ No issues):
NISAR_L1_PR_ROFF_007_106_A_140_008_7700_SH_20251211T140849_20251211T140926_20251223T140850_20251223T140926_X05016_N_P_J_001.h5
GOFF (✓ No issues):
NISAR_L2_PR_GOFF_005_106_A_140_006_7700_SH_20251117T140848_20251117T140925_20251129T140849_20251129T140925_X05016_P_P_J_001.h5
cc: @hfattahi @xhuang-jpl
HDF5 Statistics Attributes Set to Zero for Multiple RUNW Datasets
Summary
Multiple 2D raster datasets in RUNW HDF5 products have statistics attributes (
min_value,max_value,mean_value,sample_stddev) incorrectly set to 0.0, despite the actual data containing non-zero values.Affected Products
Affected Datasets in RUNW
[..]/interferogram/HH/coherenceMagnitude[..]/pixelOffsets/HH/alongTrackOffset[..]/pixelOffsets/HH/slantRangeOffset[..]/pixelOffsets/HH/correlationSurfacePeak"Attr" metrics are what is written into the HDF5 file's dataset attributes.
"Actual" metrics are computed post-facto by opening the HDF5 file and computing e.g.
numpy.nanmin(..)on the full dataset raster itself.Impact
Tested Files
RUNW (✗ Issue found in both sample files):
NISAR_L1_PR_RUNW_007_106_A_140_008_7700_SH_20251211T140849_20251211T140926_20251223T140850_20251223T140926_X05016_N_P_J_001.h5NISAR_L1_PR_RUNW_020_106_A_140_021_4000_SH_20260516T140853_20260516T140929_20260528T140852_20260528T140928_X05016_N_F_J_001.h5RIFG (✓ No issues):
NISAR_L1_PR_RIFG_013_118_A_171_014_2000_SH_20260222T102514_20260222T102520_20260306T102515_20260306T102521_X05016_P_P_J_001.h5GUNW (✓ No issues):
NISAR_L2_PR_GUNW_001_062_A_019_006_2000_SH_20250927T114508_20250927T114542_20251126T114516_20251126T114545_X05016_P_F_J_001.h5ROFF (✓ No issues):
NISAR_L1_PR_ROFF_007_106_A_140_008_7700_SH_20251211T140849_20251211T140926_20251223T140850_20251223T140926_X05016_N_P_J_001.h5GOFF (✓ No issues):
NISAR_L2_PR_GOFF_005_106_A_140_006_7700_SH_20251117T140848_20251117T140925_20251129T140849_20251129T140925_X05016_P_P_J_001.h5cc: @hfattahi @xhuang-jpl