Bill Sacks
June 6, 2013

This file documents the procedure that was used to create the initial
conditions with date stamp 130606:

clmi.I1850CRUCLM45BGC.0241-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45BGC.0241-01-01.1.9x2.5_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45BGC.0241-01-01.360x720cru_hcru_simyr1850_c130606.nc
clmi.I1850CRUCLM45BGC.0241-01-01.ne30np4_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45BGCDV.0241-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45SP.0521-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.I2000CLM45CRUBGC.2000-01-01.0.9x1.25_gx1v6_simyr2000_c130606.nc
clmi.ICRUCLM45BGCCROPmp24.0241-01-01.1.9x2.5_g1v6_simyr2000_c130606.nc
clmi.ICRUCLM45SP.2000-01-01.0.9x1.25_g1v6_simyr2000_c130606.nc

--- GENERAL OVERVIEW ---

These are all initial conditions for CLM4.5. There were older initial
conditions files (generated in May) corresponding to each of
these. However, an upcoming CLM tag (clm4_5_11) is making two changes
that make these older initial conditions unusable:

(1) Changing the numbering of landunits and columns

(2) Adding some 0-weight urban columns

Thus, each of the earlier CLM4.5 initial conditions files needed to be
adjusted via interpinic and other modifications to make them
compatible with the upcoming version of CLM.

Note that, of the above 9 files, only 4 are "original" initial
conditions files; the other 5 were interpinic'ed from these
originals. The originals corresponded to:

clmi.I1850CRUCLM45BGC.0241-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45SP.0521-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.I2000CLM45CRUBGC.2000-01-01.0.9x1.25_gx1v6_simyr2000_c130606.nc
clmi.ICRUCLM45SP.2000-01-01.0.9x1.25_g1v6_simyr2000_c130606.nc


--- STEP 1: DEALING WITH NEW NUMBERING OF LANDUNITS AND COLUMNS ---

interpinic assumes that the landunit numbering is the same on the
input & output files. To deal with this, I did the following:

(A) Created a branch (landunit_percent_CLM_a) in which the landunits
and columns were renumbered as in my main branch, but which did NOT
have the new 0-weight urban columns. Thus, the number of pfts, columns
and landunits were the same in this branch as in the trunk.

(B) For each of the 4 "original" initial conditions files listed
above, ran a case from the landunit_percent_CLM_a branch to create a
similar restart file.

(C) Created new versions of the 4 "original" initial conditions
files. These started as copies of the originals. I then overwrote the
following variables with their versions from the restart files
generated in step B: land1d_ityplun, cols1d_ityplun, cols1d_ityp,
pfts1d_itypveg, pfts1d_ityplun (note: I don't think I needed to
overwrite pfts1d_itypveg in this case, but doing so wouldn't hurt
anything).


--- STEP 2: DEALING WITH NEWLY-ADDED 0-WEIGHT URBAN COLUMNS ---

I then needed to use interpinic to create initial conditions files
that had the newly-added 0-weight urban columns. To do this, I
performed the same 4 runs as in step 1B, but now from my main branch
(landunit_percent_CLM).

I wanted the resulting initial conditions to be as close as possible
to the originals, partly to aid testing. Towards this end, I created a
temporary version of interpinic that was able to interpolate nearly
all fields (in contrast to the trunk version, which skips a number of
fields either explicitly or implicitly):
https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/landunit_percent_CLM_tags/landunit_percent_CLM_14_clm4_5_07

However, note that there were still diffs in 4 fields: EFLX_LWRAD_OUT,
albi, albd, and TWS (see bug 1702 and bug 1731 for details).

I then did the 4 interpinics, creating the files:

clmi.I1850CRUCLM45BGC.0241-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45SP.0521-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.I2000CLM45CRUBGC.2000-01-01.0.9x1.25_gx1v6_simyr2000_c130606.nc
clmi.ICRUCLM45SP.2000-01-01.0.9x1.25_g1v6_simyr2000_c130606.nc


--- STEP 3: CREATING THE OTHER FILES ---

Finally, I needed to recreate the 5 other initial conditions files,
which were NOT "originals". For this step, I reverted interpinic back
to a version that had similar behavior to the trunk:
https://svn-ccsm-models.cgd.ucar.edu/clm2/branch_tags/landunit_percent_CLM_tags/landunit_percent_CLM_15_clm4_5_07. Note
that I could not use the trunk version itself for this, because
interpinic needed to be changed based on the new landunit numbering. 

I then ran the standard interpinic procedure (run a short test case to
produce a template file, then interpinic the input file to the output
file) to produce each of the following files; note that the input file
for each of these was
clmi.I1850CRUCLM45BGC.0241-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc:

clmi.I1850CRUCLM45BGC.0241-01-01.1.9x2.5_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45BGC.0241-01-01.360x720cru_hcru_simyr1850_c130606.nc
clmi.I1850CRUCLM45BGC.0241-01-01.ne30np4_g1v6_simyr1850_c130606.nc
clmi.I1850CRUCLM45BGCDV.0241-01-01.0.9x1.25_g1v6_simyr1850_c130606.nc
clmi.ICRUCLM45BGCCROPmp24.0241-01-01.1.9x2.5_g1v6_simyr2000_c130606.nc