M21C exports and LONG_NAMES for land water and energy balances

[gmao-rreichle]

As in #621 but for M21C.

A couple of open questions:

1. I had always thought that the "lnd" uses conservative regridding, but this doesn't seem to be the case:
   

   GEOSgcm_App/HISTORY_R21C.rc.tmpl

   Line 1215 in 66fb5df

   lnd_tavg_1hr_glo_L@HIST_IMx@HIST_JM_slv.regrid_method: 'BILINEAR_MONOTONIC' ,

   
   There is no "regrid_method" in the M-2 HISTORY file, but the MAPL version used in M-2 probably works differently.

2. Do we know what "deflate" does? Is it a lossless or lossy compression?
   

   GEOSgcm_App/HISTORY_R21C.rc.tmpl

   Line 1214 in 66fb5df

   lnd_tavg_1hr_glo_L@HIST_IMx@HIST_JM_slv.deflate: 2 ,

[sdrabenh]

As in #621 but for M21C.

A couple of open questions:

1. I had always thought that the "lnd" uses conservative regridding, but this doesn't seem to be the case:
   

   GEOSgcm_App/HISTORY_R21C.rc.tmpl

   Line 1215 in 66fb5df

   lnd_tavg_1hr_glo_L@HIST_IMx@HIST_JM_slv.regrid_method: 'BILINEAR_MONOTONIC' ,

BILINEAR_MONOTONIC is the default method for most collections. However, this could be changed to any of these regridding options if desired - especially for budget terms.

There is no "regrid_method" in the M-2 HISTORY file, but the MAPL version used in M-2 probably works differently.
2. Do we know what "deflate" does? Is it a lossless or lossy compression?

GEOSgcm_App/HISTORY_R21C.rc.tmpl

Line 1214 in 66fb5df

lnd_tavg_1hr_glo_L@HIST_IMx@HIST_JM_slv.deflate: 2 ,

The deflate attribute specifies lossless compression. On the other hand, setting nbits performs "bit shaving" and would be lossy compression.

[sdrabenh]

@gmao-rreichle merging based on preliminary results. Will continue testing and make further changes if necessary.