Document ABI L2 combined processing troubles

[djhoese]

In non-CSPP processing of ABI L2 products you get resolutions that are not equal factors of each other. This is a problem when processing all of the products at the same time as the "native" resampler will try to (by default) upsample/replicate the lower resolution product to the resolution of the higher resolution. The data not being a nice whole number means that the native resamplers fails and processing can't continue. The workaround for that is to process each product in a separate G2G call. This is likely desired by users anyway as you probably don't want the 5km HT product upsampled to 1km for no reason.

[djhoese]

This may still be a problem, although Satpy has had many updates to workaround some issues related this (native resampler rechunking arrays that needed rechunking, geolocation fixes, etc). We don't really have time to investigate all use cases especially those involving non-CSPP workflows (since we are funded to support CSPP products), but we'll move this to the G2G 1.2 milestone in hopes that any users running into this issue let us know on the forum or on this repository (feel free to comment on this issue) and provide more details about these edge cases.

Honestly, I don't know the exact use/data case I was referring to when I originally created this issue. I'll provide more details if I find them.

[kathys]

I will document this is Geo2Grid Version 1.2.