What was new with CLM4 in CESM1.0.1 (in CESM1.0.1) since the April 1st, 2010 CCSM4.0 release?

From, CLM4.0.00 in the CCSM4.0 release to CLM4 in CESM1.0.1 there were several developments made to CLM. A glacier multiple elevation class option was added that allows the use of CLM4 with a glacier land ice model the Community Ice Sheet Model (CISM). A bug-fix for the snow hydrology was added. Several new namelist items were added a few new history fields. Also the capability of reading aerosol and nitrogen deposition from stream files at one resolution and regridded on the fly rather than with datasets at the model resolution was added in. This was important for higher resolutions so that large datasets do not have to be created before running the model, nor are datasets for every resolution required.

What was new with CLM4 in CESM1.0.1 Science since CCSM4.0?

In general, snow layers should not be thinner than

dzmin = wice/rhoice + wliq/rholiq
If dz < dzmin, then the value of "void" computed in subroutine SnowCompaction is negative, which is unphysical. This doesn't cause problems with the compaction itself, but results in unrealistic values of vol_ice, vol_liq, and eff_porosity in subroutine SnowWater. We can have vol_ice = 1 and vol_liq = 0 even when liquid is present, which cuts off the runoff (qout) from the lowest snow layer. Liquid water then accumulates in the snow column without draining, which leads to further problems and eventually a code crash.

The solution to this problem was to adjust layer thickness dz for any water+ice content changes in excess of previous layer thickness, e.g.,

dz(c,j) = max(dz(c,j),h2osoi_liq(c,j)/denh2o + h2osoi_ice(c,j)/denice)
at appropriate steps in the snow hydrology subroutines.

Snow hydrology bug fix.
Add multiple elevation class option for glaciers so can interact with the land ice sheet model.

What was new with CLM4 in CESM1.0.1 Software since CCSM4.0?

New configuration options


glc_nec can be 1,3,5, or 10 and MUST match the number on the input surface dataset the elevation classes themselves are read from the surface dataset

New namelist items:


carbon_only = If true, and CLMCN carbon-nitrogen model is on, Nitrogen is unlimited rather than prognosed and vegetation will be over-productive (replaces the supplemental Nitrogen #ifdef)

create_glacier_mec_landunit (= T when these landunits are created; F by default)

glc_smb (= T if passing surface mass balance to GLC; else pass PDD info; T by default)

glc_dyntopo (= T if CLM topography changes dynamically; currently F) (NOT fully implemented yet)

ice_runoff = If true, river runoff will be split up into liquid and ice streams, otherwise ice runoff will be zero and all runoff directed to liquid stream

ndepmapalgo = Mapping method from Nitrogen deposition input file to the model resolution (can be bilinear,nn,nnoni,nnonj,spval,copy, bilinear by default)

scaled_harvest = If true, harvesting will be scaled according to coefficients determined by Johann Feddema, 2009

New history fields:

aais_area Antarctic ice area (km^2)
aais_mask Antarctic mask (unitless)
gris_area Greenland ice area (km^2)
gris_mask Greenland mask (unitless)
QICE ice growth/melt (mm/s)
QICEYR ice growth/melt (mm/s)
QTOPSOIL water input to surface (mm/s)
VOLR RTM storage: LIQ (m3)
VOLR_ICE RTM storage: ICE (m3)