...
For GLL data, the best algorithm to use (see Transition to TempestRemap for Atmosphere grids ) is TempestRemap’s “intbilin”, as this map is accurate, monotone, and can be used for both downscaling and upscaling. The only drawback is that it is not exactly conservative. If exact conservation is needed (so that the mass computed on the interpolation grid is the same as on the native grid), then TempestRemap’s less accurate “mono” should be used.
For PG2 data, to remap from FV (unstructured) to FV (latlon), we dont don't yet have an integrated bilinear option. ESMF’s “aave” option ( which can be obtained most efficiently using NCO’s ncremap) is conservative and good for downscaling, but will produce blocky plots when upscaling. For nicer looking plots when upscaling, ESMF’s “bilinear” is currently the only option.
...
pyNGL can construct this dual grid from a triangulation if the user specifies sfXArray and sfYArray. WE dont conver We don't cover this option here, but instead cover the second option where the user reads a dual grid from the SCRIP file and specifies that via sfXCellBounds and sfYCellBounds. For PG2 data, this is the most natural approach, with the PG2 dual grid made up of uniform quadrilaterals given in the PG2 SCRIP file.
...
It may be possible to simplify this example, but so far this is the best solution I’ve come up with to do this particular thing in mpl. Some caveats: this does not seem to work well with different map projections, and I’m not sure why yet. Attempts to first project vertices to the map coordinate system and then draw the polygons has also been largely unsuccessful, so this is a work in progress. Regardless, both of these approaches have the downside of representing the GLL point data as cell-centered data, which is maybe not completely appropriate for data on the GLL grid (however, it probably is the best representation of data on the alternative pseudo-finite-volume physics grids, e.g, ne4pg2).
...