Is the GEO KOMPSAT 2A reader still working in 2.0 beta?

[cmhm7]

Hello,

I have dowloaded beta 2.0.0b1 for windows and it's great for MTG and MSG images. I am trying to open GeoKompsat 2A images. I added to /config/limit_available_reader.yaml the line -ami_l1b to enable it in the reader menu However when I select an AMI.nc and use the ami_l1b reader I get an error saying on the GUI "error: no known product could be loaded from file" and on the console I have:

024-10-02 17:01:36 DEBUG yaml_reader:create_filehandlers:L622 Assigning to ami_l1b: ['.\gk2a_ami_le1b_ir087_ela020lc_202406190710.nc'] 2024-10-02 17:01:36 WARNING common:create_scenes:L88 cannot rename 'dim_image_x' because it is not a variable or dimension in this dataset 2024-10-02 17:01:36 ERROR open_file_wizard:validateCurrentPage:L226 No known products can be loaded from the selected files.

The .nc is correct, it opens fine in SNAP.

Is there something I am doing wrong? Should the GEO KOMPSAT 2A reader still work in 2.0.0b1 ?

Regards

[ameraner]

Hi @cmhm7 , thanks for reporting this. Sounds like the steps you have done to activate the reader are correct, and that the reader is initialised correctly. So I suspect an issue with the filepattern... could you kindly post an example of a filename you're trying to read? Satpy supports filepatterns as in e.g. gk2a_ami_le1b_ir087_fd020ge_201901260310.nc, if yours is deviating from this, that might be the reason.

[cmhm7]

You're welcome!

Indeed it seems to be an issue with the file, gk2a_ami_le1b_ir087_ela020lc_202406190710.nc is an Extended Local Area file which uses a slightly different format from the Full Disk files (the projection is different for instance).

When I use a full disk image I don't have an error, I can choose the channels and it works fine (except I can't choose a projection from GEOKOMPSAT's point of view)

[cmhm7]

Local Area (LA) files open fine too, but Korea (KO) files don't

[ameraner]

Thanks for checking! So the issue is with the satpy reader not supporting this file type. I would suggest to open an issue (or even better a Pull Request) to the satpy github in order to support your specific format in the reader.

As for the area, if you want the original geostationary projection (otherwise you can just use plate_carree): the cleanest way would be to add an appropriate area definition to the satpy areas.yaml file, and then add it to the SIFT defaults or your local config (see https://sift.readthedocs.io/en/latest/configuration/area_definitions.html). In the meanwhile you can also do it locally, using local satpy config plugin (see https://sift.readthedocs.io/en/latest/configuration/external_satpy.html).