Fix some NFIX variables, update cime/cism for upgraded hobart new glade, new diagnostic fields, update cmip6 output

Add some new diagnostic fields. Some needed for CMIP6. Update the CMIP6 user-mods output.
Fix a couple issues. Get full list of history tapes working correctly. Check for valid range
of CO2. New IC file interpolated from the previous one for f19_g17_gl4 for 2000 Clm50BgcCrop

Update cime and cism so can work on upgraded hobart. Also fix an issue that the Nag6.2 compiler found
in CTSM. Also get working with glade changes that happened on cheyenne, DIN_LOC_ROOT_CLMFORC was
changed to a new directory.

Nitrogen Fixation flux arrays were being set to missing value over non-vegetated landunits. This sets them to zero everywhere
and averages in the zero's at the gridcell level for history output. It also reads in restarts with missing values and converts
them to zero's.

There are also some tools updates, getting the tools working on the new upgraded hobart.cgd.ucar.edu.

And mksurfdata_map is updated to add some *_MAX files on the landuse.timeseries files that will allow us to conserve memory
for transient cases.