[BUG/ISSUE] HCO_MISSVAL is incorrect for GCHP/GEOS which use MAPL

[lizziel]

HEMCO assigns parameter HCO_MISSVAL (-1e31 float) to denote missing values. If HEMCO handles file read, all missing value grid cells are assigned value HCO_MISSVAL and those grid cells are later skipped during regrid interpolation. During HEMCO run, missing data are then specially handled during scaling and masking.

This works for models in which HEMCO handles the file read and regridding, but not for other models if they assign a different value to missing data. This is the case for GCHP and GEOS which use MAPL's missing value parameter MAPL_UNDEF, equal to 1e15 float. A temporary kludge was put in back in 2017 for masking in GCHP, but was not extended to scale factors, and mostly went unused because MAPL did not use the _FillValue attribute to identify missing values. Instead, MAPL expected missing values to have data values 1e15 in the input files as regular data or assumed _FillValue was 1e15.

The fix is to set HCO_MISSVAL to the MAPL_UNDEF value in HEMCO. A separate fix will be applied to MAPL to identify missing values from netCDF _FillValue attribute without any assumption of what that value is.

This issue is fixed HEMCO in #171.

Heads up @jimmielin that we should keep an eye out for whether there are any missing value handling issues in CESM or WRF-GC.

[jimmielin]

Thanks @lizziel. Copying @fengx7 for WRF-GC as well. We follow the handling in the GEOS-Chem Classic code as we use the same libraries so hopefully we won't have to do anything extra to be consistent with other model(s).

[lizziel]

As long as we are using HEMCO to read and regrid we should be good.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. If there are no updates within 7 days it will be closed. You can add the "never stale" tag to prevent the Stale bot from closing this issue.

[msulprizio]

The fix for this is now merged (see #171). I will close out this issue.