Forum

Yes, to make an L2 file you should provide the ancillary data. We make an L1B file from the L1A and GEO and then pass the L1B to l2gen.
I think l2gen has the code to make L2 directly from L1A and GEO but I will defer to one of our SeaDAS experts on running that...

Tommy

Are you setting the correct environment?
export OCSSWROOT=path_to_your_ocssw_directory
export OCVARROOT=$OCSSWROOT/var

We stage the operational GEO files so you don't have to make them.
Replace "L1A" with "GEO" in the filename.

Tommy

That query is missing the short_name: 'https://cmr.earthdata.nasa.gov/search/granules.csv?page_size=2000&sort_key=short_name&sort_key=start_date&short_name=&provider=OB_DAAC.... 'https://cmr.earthdata.nasa.gov/search/granules.csv?page_size=2000&sort_key=short_name&sort_key=st...

We have the same gaps at L0

Thanks
Tommy

I'm seeing missing data on our source. I'll compare our counts with theirs and see if any can be recovered on our side. I doubt MODIS will be doing any recovery on their side. I'll update when I know more.

Tommy

I can trace the problem better if you tell me how you are accessing the data and provide an example of a missing granule.

Thanks
Tommy

l2gen only needs the first file:
ifile=Oa01_radiance.nc
geofile=geo_coordinates.nc

It processes all from there.

Tommy

The Raman Correction was computed already, that output is l2gen processing the file. These are chunky L1B files that take a while to process.

Tommy

Try adding share to the OCDATAROOT

os.environ['OCDATAROOT'] = '/.../l2gen/ocssw/share'

and there should be an olci directory under share

Tommy

The forum software may have filtered the code attachment to prevent distribution of malicious code.
Do not run the code if you cannot verify its actions.

Tommy