The SSMI(S) Hydrological Products are a long-term record of monthly, globally gridded passive-microwave fields of hydrological variables, including precipitation rate and frequency, cloud liquid water, water vapor, snow cover, and sea ice, beginning in July 1987, at 2.5 and 1.0 degree resolution. They are derived from the Defense Meteorological Satellite Program (DMSP) Special Sensor Microwave/Imager (SSM/I) and Special Sensor Microwave Imager/Sounder (SSMIS) series and produced operationally at NOAA's National Centers for Environmental Information (NCEI).
The record is used well beyond its own archive. It provides passive-microwave input to the Global Precipitation Climatology Project (GPCP) Climate Data Record, and it has contributed every month to the NOAA Climate Prediction Center (CPC) Climate Diagnostics Bulletin. Sustaining a consistent, well-characterized multi-decadal record across an evolving satellite constellation, so that downstream products can rely on it month after month, is the central problem this project addresses.
This repository holds the version 2 processing system, an open-source Python 3 reimplementation of the legacy Fortran77, IDL, and GrADS processing from the daily 1/3-degree antenna-temperature grids onward, with an automated test suite and continuous integration. Testing it against the legacy code on identical inputs found three latent defects in the legacy code, an unbounded index into the antenna-temperature correction table, an unbounded index into the scattering-index table, and uninitialized variables in the emission branch of the precipitation retrieval. Version 2 removes them, and with them removed the legacy code and version 2 agree within 0.001 in product units on the months tested. Version 2 also corrects the dual-satellite merge of the GPCP input product and updates the snow-cover anomaly baseline to the World Meteorological Organization 1991-2020 climate normals. The historical record has not yet been reprocessed with version 2.
Hilawe Semunegus (NOAA NCEI) is the Principal Investigator for the SSMI(S) Hydrological Products, responsible for the production, scientific stewardship, validation, and modernization of the record. Ralph R. Ferraro (University of Maryland ESSIC) is a co-author and the original author of the retrieval algorithms, which build on the long-established NOAA passive-microwave hydrology lineage.
Northern and Southern Hemisphere snow cover and anomaly, January 2026:
The DMSP satellite constellation contributing to the record:
Global rainfall, January 2026:
- Climate algorithms (
monthly/climalg_ssmis.py): the antenna-to-brightness temperature conversion and the precipitation, cloud, water vapor, snow, and sea ice retrievals, fully vectorized with NumPy. - Monthly assembly (
monthly/combine.py,monthly/pentad_ssmis.py): multi-year record management and 5-day pentad averaging. - GPCP merge (
monthly/gpcp_processing.py): snow and ice masking and the dual-satellite precipitation merge. - NetCDF-CF export and validation (
monthly/netcdf/): CF-1.5 output and a validation harness against the legacy reference. - Imagery (
monthly/grads/run_mon_image.py): matplotlib and cartopy figures that replaced the legacy GrADS scripts. - Configuration (
monthly/satellite_config.py,monthly/product_version.py): a single source of truth for the satellite constellation and the product version, so a satellite transition is a configuration change. - Tests (
tests/): a pytest suite covering the radiometric conversion, the binary and NetCDF I/O contracts, grid geometry, retrieval invariants, and a deterministic bitwise golden-month regression. Continuous integration runs it on every push.
pip install -r requirements-test.txt # plus numpy, scipy, netCDF4
python -m pytest testsTests that require a third-party library or the operational input grids skip cleanly when those are unavailable, so the suite runs on any host with the scientific Python stack. The golden-month regression runs only where the 1/3-degree input grids are mounted.
- A multi-decadal, multi-sensor record kept as two chains, morning and late morning, each a sequence of DMSP satellites, so users have two samplings of the diurnal cycle.
- Operational input to the GPCP Climate Data Record and to the monthly CPC Climate Diagnostics Bulletin.
- Defects found by testing against the legacy code are removed, and the effect of removing them is measured over a full year of matched inputs rather than assumed.
- The binary format contract (float32, south-first rows, longitude-fastest) and the NetCDF metadata are pinned by the test suite, not just documentation.
- The version 2 processing corrections and their evaluation are described in "Evaluation and Processing Corrections for the SSMI(S) Hydrological Products" by Hilawe Semunegus and Ralph R. Ferraro, in revision at the Journal of Applied Meteorology and Climatology.
- The archive of record is held at NOAA's National Centers for Environmental Information, doi:10.7289/V5QF8R3K.
- A corrected 2024 sample year produced by version 2 is archived on Zenodo, doi:10.5281/zenodo.23424220.
- The processing source is also archived on Zenodo under the concept DOI 10.5281/zenodo.20127244, which resolves to the latest released version.
Public domain under Creative Commons Zero 1.0 (see LICENSE).