These data originate from NOAA/NCEP.

The NOAA Climate Prediction Center/NCEP/NWS is making the data available originally in binary format, in a weekly rotating archive. The NASA GES DISC is acquiring the binary files as they become available, converts them into CF (Climate and Forecast) -convention compliant netCDF-4 format, and stores the product in a permanent archive. The original record started from February, 2000, but in June, 2025 it was extended back to January, 1998.

The leading edge of data availability is delayed by about 24 hours from real-time to abide by international data exchange agreements between NOAA and EUMETSAT (the METEOSAT data providers).

The data contain globally-merged (60°S-60°N) 4-km pixel-resolution IR brightness temperature data (equivalent blackbody temps), merged from the European, Japanese, and U.S. geostationary satellites over the period of record (GOES-8/9/10/11/12/13/14/15/16/17/18/19, METEOSAT-5/7/8/9/10/11, and GMS-5/MTSat-1R/2/Himawari-8/9).

The global geo-IR are dynamically calibrated to GOES East, using a 35 day trailing inter-calibration using time/space-matched IR Tb’s at the mid-point between sub-satellite positions. In the event of duplicate data in a grid box, the value with the smaller zenith angle is taken. The data have been corrected for "zenith angle dependence", in which IR temperatures for locations far from satellite nadir are erroneously cold due to a combination of geometric effects and radiometric path extinction effects (Joyce et al. 2001). Finally, the data are re-navigated for parallax, which shifts the geo-location of the GEO-IR footprints to approximately account for the cloud tops that the IR “sees” being displaced away from their actual geographic location when viewed along a slanted path. These corrections allow for the merging of the IR data from the various GEO-satellites with greatly reduced discontinuities at GEO-satellite data boundaries. In the event of duplicate data in a grid box, the value with the smaller zenith angle is taken.

The NASA GES DISC is curating these data in a self-documenting, CF-compliant, netCDF-4 format, which allows a broad range of applications to access the data directly, without the need to cope with the original binary data format. In addition to the direct download of netCDF-4 data, the GES DISC provides data download in binary, ASCII, and netCDF-3 formats using the OPeNDAP interface which also provides remote data access.

Similarities with the original

As in the original binaries, every netCDF-4 file covers one hour, and contains two half-hourly grids, at 4-km grid cell resolution.

Differences from the original

1. The data in the netCDF-4 files are already converted to physical values of Brightness Temperatures in Kelvin. Because the original data values are round with no decimal precision, the data type in the netCDF-4 files has been changed to 2-byte signed integer, a transition that took place in mid-August, 2025. This reduces the file size and speeds up data download and remote access. There is no need to further scale these data. The netCDF-4 format is machine-independent and users need not worry about the endian-ness of their machines.

2. To meet the requirements of collection spatial metadata, the grid is re-ordered from the original and now goes from -180 (West) to 180 (East). It is also starting from -60 (South).

The data and time units are reflected in the corresponding "units" attributes, and grid dimensions are described by longitude ("lon"), latitude ("lat") and "time" vectors. Thus, any CF-compliant tool should automatically understand the setup in the data files and the starting time for each half-hourly grid. Even without such tools, simple "ncdump" or "h5dump" command line tools will easily disclose the netCDF-4 files configuration.

Acknowledgements

The creation of the original data at NOAA/NCEP is supported by funding from the NOAA Office of Global Programs for the Global Precipitation Climatology Project (GPCP) and by NASA via the Tropical Rainfall Measuring Mission (TRMM).

The permanent archive is supported by NASA's HQ Earth Science Data Systems (ESDS) Program.