Bugfix #2070 main_v5.0 var list numeric order

[georgemccabe]

See #2072 for review instructions. Fixes issue in a less invasive approach to reduce the chance of unintended side effects.

NOTE: Failure in checks is a difference in surrogate severe use case output that is unrelated to these changes.

[JohnHalleyGotway]

I approve these changes. I also agree that it's wise to make the 5.0 bugfix as limited in scope as possible.

Please see this #2072 comment for the testing setup.

I ran this main_v5.0 bugfix branch and confirmed that the order of output written now matches the order in which they're listed in the METplus conf file.

mkdir -p OUTPUT/main

# Run main bugfix branch
export METPLUS_TEST_OUTPUT_BASE=`pwd`/OUTPUT/main
METplus-main/ush/run_metplus.py  -c system.conf METplus-main/parm/use_cases/model_applications/medium_range/GridStat_fcstGFS_obsGFS_Sfc_MultiField.conf

List the output variable order:

grep SL1L2 OUTPUT/main/met_out/GFS/sfc/201706130000/grid_stat/grid_stat_GFS_TMP_vs_ANLYS_TMP_Z2_240000L_20170613_000000V.stat | grep CONUS | awk '{print $10, "/", $12}' | sed -r 's/$/, /g' | tr -d '\n'

Matches the expected order:

TMP / Z2, RH / Z2, SPFH / Z2, HPBL / L0, PRES / Z0, PRMSL / L0, TMP / Z0, UGRD / Z10, VGRD / Z10, TSOIL / Z10-0, SOILW / Z10-0, WEASD / Z0, CAPE / Z0, CWAT / L0, PWAT / L0, TMP / L0, HGT / L0, TOZNE / L0