ERA5 DATM data - reference height discrepancy

As far as I can tell, the ERA5 data-atmosphere set is using the 2 meter surface data for temperature and humidity, and the 10 meter height for wind.

https://github.com/E3SM-Project/E3SM/blob/master/components/data_comps/datm/cime_config/namelist_definition_datm.xml#L128-L131

and:

https://github.com/E3SM-Project/E3SM/blob/master/components/data_comps/datm/cime_config/namelist_definition_datm.xml#L716-L719

ELM has a hard-coded assumption in CanopyFluxesMod.F90 that atmospheric boundary conditions are at 30 meters. This reference height is intentionally a good deal higher than the surface so that it is decoupled from the surface, thus allowing us to apply MOST and the log wind profile to scale fluxes (unless we have something new and fancy and I'm way out of touch) from the reference point to the surface. If the data is measured at 2 and 10 meters, it is well inside this blending height and is using a value that is coupled to the land-surface instead of one away from it.

The GSWP3 (v1), seemed ok and uses it's model's lowest atmopheric layer, which is correct as far as I can tell. But I didn't do a thorough check on this.

Mentioned to @bishtgautam and @dmricciuto already, cc: @thorntonpe and @jenniferholm @ckoven

Read the ERA5 entries in components/data_comps/datm/cime_config/namelist_definition_datm.xml at the referenced lines, then inspect the hard-coded atmospheric boundary-condition height in CanopyFluxesMod.F90. Compare these assumptions with the GSWP3 setup and relevant comments or discussion. Done means the ERA5 reference heights and ELM's expected boundary-condition height are reconciled and documented or corrected.

Written by the indexing model from the issue text.