OK. Thanks for the information!It's not clear when or if the GSWP3 atm forcing data will be updated for more recent years. You could check with the dataset creator: Hyungjun Kim, firstname.lastname@example.org
An alternative forcing dataset is CRUJRA, which is through 2017:
Hi, Keith! I configured a Case using I2000Clm50BgcCruGs compset. For this case, I changed the datam stream file by creating user_datm.streams.txt.CLMCRUNCEPv7.Precip/ Solar/ TPQW in my Case directory. In the user_ stream files, I pointed to /glade/p/cgd/tss/people/dll/TRENDY2020_Forcing/three_stream_files.
Thanks for pointing that out. I am running a branch simulation with RUN_REFDATE=0090-01-01, and the RUN_STARTDATE was ignored. So I guess I should use DATM_CLMNCEP_YR_ALIGN=90 in this case. And it works! Now the first year atmospheric forcing is the 2000 forcing data.Your "RUN_STARTDATE" is 0001. If you want the model start year to line up with the first year of atmospheric forcing, then you would need to set DATM_CLMNCEP_YR_ALIGN=1, not 2000.
Let us know if that doesn't work.
Thanks a lot. It is much easier for me to understand now.Well not necessarily always.
The datm guide here explains two common cases (this is for CPLHIST but CLMNCEP operates the same way):
DATM_CPLHIST_YR_ALIGN -999 integer run_component_datm env_run.xml
- Valid Values 
- Description Simulation year corresponding to DATM_CPLHIST_YR_START (only used
when DATM_MODE is CPLHIST). A common usage is to set this to
RUN_STARTDATE. With this setting, the forcing in the first year of
the run will be the forcing of year DATM_CPLHIST_YR_START. Another
use case is to align the calendar of transient forcing with the
model calendar. For example, setting
DATM_CPLHIST_YR_ALIGN=DATM_CPLHIST_YR_START will lead to the
forcing calendar being the same as the model calendar. The forcing
for a given model year would be the forcing of the same year. This
would be appropriate in transient runs where the model calendar is
setup to span the same year range as the forcing data.
Another case would be where one doesn't have forcing data for the full duration of a desired transient run.
For example, our land-only transient simulations run 1850-2014, but we only have forcing data for 1901-2014.
So, we've decided we want to use the first 20 years of our forcing data (1901-1920) for simulation years 1850-1900. Then we want to use our 1901-2014 forcing data for simulation years 1901-2014.
For the first part of the simulation we set:
So here we are telling the datm that we want model year 1901 to (eventually) line up with year 1901 of the forcing data.
The datm will actually figure out that it needs to start with 1910 forcing data for the first year of simulation (1850), eg.,
Model year 1850 1859 1860 1861 1881 1901
Atm year 1910 1919 1920 1901 1901 1901
We run the simulation through the end of 1900 with this setup, then we switch to using atm years that correspond to model years for simulation years 1901-2014. RUN_STARTDATE stays the same because we are doing a restart (CONTINUE_RUN - True).
Hello Keith. I have a small follow up question about using this forcing data.