Page 1 of 1

Issue with certain bands in PACE V3_0 AOP files

Posted: Mon Feb 24, 2025 2:56 pm America/New_York
by dbotis
Hi,

I am using the gpt mosaic and bandmaths operators in a graph with some PACE AOP L2 V3_0 files and am having an issue with some bands toward the red. When I try to run the graph on all the bands, it hangs on Rrs_689 and above. If I take out Rrs_689 from the xml graph, it then hangs on Rrs_691 and so forth. All bands above 689 seem to fail (Error in the mosaic node: Undefined symbol 'Rrs_689'). Everything up to 689 runs fine.

If I open the input L2 file in SeaDAS, I see all the bands and things are fine. It's just when I go to mosaic that the issue pops up. I can get away w/o these bands, but am curious what might be happening. I had no issues with the same approach for V2 files. I had to change some of the wavelengths for V3. It seems Rrs bands between 588 and 613 are not included in the V3 AOP files.

Thanks,

Dan

Re: Issue with certain bands in PACE V3_0 AOP files

Posted: Thu Feb 27, 2025 1:54 pm America/New_York
by OB.DAAC - SeanBailey
Dan,
I've pinged the SeaDAS team about this and they'll (hopefully) be getting back to you soon on GPT mosaic issue

I can confirm that the v3 AOP data do have fewer bands, specifically in the 590-610nm region. Our instrument calibration specialists are chasing down a complicated issue with the calibration in that region and strongly recommended we do not report data for those bands until the issue is resolved (see https://oceancolor.gsfc.nasa.gov/data/reprocessing/V3.0/pace-oci/)

Sean

Re: Issue with certain bands in PACE V3_0 AOP files

Posted: Mon Mar 03, 2025 3:31 pm America/New_York
by OB SeaDAS - xuanyang02
Dan,

Thanks for reporting the gpt issue!

Could you please do SeaDAS-Toolbox -> Software & System Info and post the results here? We would like to replicate the issue first?

Bing

Re: Issue with certain bands in PACE V3_0 AOP files

Posted: Mon Mar 10, 2025 3:48 pm America/New_York
by dbotis
Hi Bing,

Thanks for your reply.

Looking more closely at this, I see that the version of gpt that I am using is from esa-snap. So, perhaps the issue is not with SeaDAS. However, the same method I'm using worked fine for V2 PACE files, so I am curious about what the issue might be.

I will try the SeaDAS version, but I may need to have our IT lead install it first. Are the SeaDAS and ESA-SNAP versions of gpt quite different?

Thanks,

Dan

Re: Issue with certain bands in PACE V3_0 AOP files

Posted: Mon Mar 10, 2025 4:46 pm America/New_York
by OB SeaDAS - xuanyang02
Hello Dan,

Unfortunately, SeaDAS 9 has issues with GPT, and SeaDAS 8.3.0's GPT behaves the same as SNAP 10's GPT.

For some reason, SeaDAS 8.3.0 could only get the OCI (V3) bands up to Rrs_688, and I suspect that's why GPT (SeaDAS 8.3.0 and SNAP) fails with any bands above Rrs_689.

We are trying to fix the GPT issue with SeaDAS 9 for the next release.

Bing