G2G AGRI geotiff images show artifacts

[kathys]

Working with the November 4 G2G tarball, I am creating all default AGRI products from an FY4A dataset. The resulting infrared images show some artifacts including a non-white backing in Channel 7, and cutoff and smearing of data along the right side of the image in the infrared bands 8-14.

data: bumi/data/test_data/agri/2022_01_20
logfile: bumi:/data/users/kathys/geo/agri4a

See the montage below:

[djhoese]

I think this was just fixed by @simonrp84 in pytroll/satpy#2276

[djhoese]

Could also be this one: pytroll/satpy#2165

[simonrp84]

I have not seen the smearing in any of my processing through satpy, but the non white background should be fixed in my recent pr yes.

[djhoese]

I wonder if the smearing is an issue in the image viewer. I've seen that with some when dealing with tiled geotiffs.

[kathys]

I can confirm that you are right about the viewer. I tried displaying the image using two other viewers, and neither of them showed the smearing artifacts in the IR bands. The problem viewer is ImageMagick.

[kathys]

G2G V1.1 11/25 tarball still creates AGRI-FY4A Channel 7 non-transparent background. I do not consider this a show-stopper for the release of the tarball.

[djhoese]

@kathys if you have time, could you confirm transparent backgrounds on all ABI channel geotiffs?

[simonrp84]

Any chance you can send one of the problematic files over, or at least give the filename? Agri files are sometimes inconsistent so maybe you've got some with a different fill value or something.

[kathys]

@djhoese No non-transparent geotiffs from any of the abi instruments/bands that I have seen.

[kathys]

@simonrp84 I put the data that I am using here:

https:

[simonrp84]

@kathys Thanks, but it looks like the /fy4a/ directory is missing.

[kathys]

Sorry @simonrp84 it was not supposed to get scrubbed. I placed it here: https://bin.ssec.wisc.edu/pub/CSPP/data/