In order to better keep track of the constantly changing download URLs of the various satellite data referenced in this blog, this page has been created to host the most recent known working server locations.
NOAA GOES SERIES
For full-disk images in pseudo-true color, these are the two most useful products:
Product: ABI-L1b-RadF
Format: Individual bands; 10848 x 10848 resolution or higher.
Product: ABI-L2-MCMIPF
Format: Combined bands; 5424 x 5424 resolution.
GOES-16 (EAST) Download — 2017-2025:
https://noaa-goes16.s3.amazonaws.com/index.html#ABI-L1b-RadF/
https://noaa-goes16.s3.amazonaws.com/index.html#ABI-L2-MCMIPF/GOES-17 (WEST) Download — 2018-2023:
https://noaa-goes17.s3.amazonaws.com/index.html#ABI-L1b-RadF/
https://noaa-goes17.s3.amazonaws.com/index.html#ABI-L2-MCMIPF/GOES-18 (WEST) Download — 2022-Present:
https://noaa-goes18.s3.amazonaws.com/index.html#ABI-L1b-RadF/
https://noaa-goes18.s3.amazonaws.com/index.html#ABI-L2-MCMIPF/GOES-19 (EAST) Download — 2024-Present:
https://noaa-goes19.s3.amazonaws.com/index.html#ABI-L1b-RadF/
https://noaa-goes19.s3.amazonaws.com/index.html#ABI-L2-MCMIPF/NOTES:
Pseudo-true color requires visible bands C01 (blue), C02 (red), and C03 (near-infrared). For the individual full-resolution radiance bands, here are example files for September 22, 2025 at 16:20 UTC (autumn equinox, ~local solar noon):
OR_ABI-L1b-RadF-M6C01_G19_s20252651620209_e20252651629517_c20252651629572.nc (Band C01, 10848 x 10848)
OR_ABI-L1b-RadF-M6C02_G19_s20252651620209_e20252651629517_c20252651629561.nc (Band C02, 21696 x 21696)
OR_ABI-L1b-RadF-M6C03_G19_s20252651620209_e20252651629517_c20252651629557.nc (Band C03, 10848 x 10848)Combined bands contain all 16 available bands in a single file. Here is the example file for the same date:
OR_ABI-L2-MCMIPF-M6_G19_s20252651620209_e20252651629523_c20252651629598.nc (5424 x 5424)The two products have different levels of processing and must be handled differently in Python. For details, refer to the respective Python script pages (links to come).
This is what the current GOES pseudo true-color script produces, using the above ABI-L2-MCMIPF product:
 |
| GOES-19 at approximately solar noon on the 2025 autumnal equinox. Processed by the author. No retouching was done to the image, though it probably needs it. |
HIMAWARI-8/9
Catalogue: Bureau of Meteorology Satellite Observations (ra22)
Format: Individual bands; 11000 x 11000 resolution or higher.
Himawari-8 Download — 2015-2026:
https://thredds.nci.org.au/thredds/catalog/ra22/satellite-products/arc/obs/himawari-ahi/fldk/latest/catalog.htmlHimawari-9 Download — 2026-Present:
https://thredds.nci.org.au/thredds/catalog/ra22/satellite-products/nrt/obs/himawari-ahi/fldk/latest/catalog.htmlNOTES:
Like GOES radiance products, Himawari full-disk bands come as individual files. For near true-color, visible bands B01 (blue), B02 (green), B03 (red), and B04 (near-infrared) are required. Despite having a greenish band, the near-infrared band is still required to boost the chlorophyll and soil signal in it.
Here are example Himawari-19 bands for September 22, 2025 at 02:30 UTC (autumn equinox, ~local solar noon):
20250922023000-P1S-ABOM_OBS_B01-PRJ_GEOS141_1000-HIMAWARI9-AHI.nc (Band B01, 11000 x 11000)
20250922023000-P1S-ABOM_OBS_B02-PRJ_GEOS141_1000-HIMAWARI9-AHI.nc (Band B02, 11000 x 11000)
20250922023000-P1S-ABOM_OBS_B03-PRJ_GEOS141_500-HIMAWARI9-AHI.nc (Band B03, 22000 x 22000)
20250922023000-P1S-ABOM_OBS_B04-PRJ_GEOS141_1000-HIMAWARI9-AHI.nc (Band B04, 11000 x 11000)Here is a preview as produced by the current Python script:
 |
| September 22, 2025 equinox from Himawari-9. Credit: JMA/ABOM. Processed by the author with no further retouching. |
METEOSAT-12
Product: FCI Level 1c Normal Resolution Image Data - MTG - 0 degree
Format: Combined bands; 11136 x 11136 resolution or higher.
Meteosat-12 FCI Download — 2024-Present:
https://user.eumetsat.int/catalogue/EO:EUM:DAT:0662/overviewNOTES:
EUMETSAT requires an account and license to download and use its products. Both are available in free versions.
Meteosat-12 is similar to Himawari in that that it has an imperfect green band for near true-color images (requiring the same tweaks). The blue band is actually closer to ideal than the blue band Himawari uses. Unlike the other platforms though, the data is broken up into 40 chunks in a single ZIP archive that have to be stitched together. In Python, the Satpy library can handle this.
Here is an example Meteosat-12 *filename* (no direct link is possible) for September 22, 2025 at 12:00 UTC (autumn equinox, local solar noon):
W_XX-EUMETSAT-Darmstadt,IMG+SAT,MTI1+FCI-1C-RRAD-FDHSI-FD--x-x---x_C_EUMT_20250922120344_IDPFI_OPE_20250922120007_20250922120935_N__O_0073_0000.zip (All bands, >= 11136 x 11136)For an actual test granule without an account, there is a free tutorial with a FCI file download link in the description. Save the FDHSI_example.zip archive to your computer. I will be posting a Python script to show how to generate a near true-color full disk example from it.
Here is a preview of what the test granule will look like:
 |
| Earth as imaged by Meteosat-12 on January 31, 2024. Credit: EUMETSAT. Processed by the author. |
All three satellite examples are probably much too dark. That's something that can easily be adjusted to preference though.
CREDIT
Australian Bureau of Meteorology
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Comments
Post a Comment