"resample_nearest" GEO swath to eqc has tiny periodic "glitch" in results

[LSUESL]

"resample_nearest" GEO swath to eqc has tiny periodic "glitch" in results

[djhoese]

I'm very sorry for taking so long to answer this. I'm on paternity leave and had this issue open in a browser tab that got lost in the sea of tabs I've opened while checking emails. If I'm understanding your code and issue correctly, I think the main thing stems from your input longitude and latitudes. Have you checked them for consistency? I ask because it is a (well?) known issue that ABI L1b NetCDF files (C01-C16 for example) have 32-bit float scale factor and offset attributes for the X and Y coordinate variables. 32 bits is not enough precision for the distance covered by these coordinate variables. In Satpy (another pytroll project) I went to a lot of work to convert these scale and offset attributes to 64-bit floats and even round some of the decimal values when computing the geostationary X/Y metered coordinate arrays to get the theoretical/accurate values.

My guess (hope?) is that your lon/lats in your files were not generated with this level of care and are seeing the effects of the 32-bit scale factor attribute.