What should I do if I want to use the pft distribution and lai representative of year 2011 conditions other than 1850 or 2000 ?

[Xueli Huo]

I am running clm5 offline with a forcing data set representative of year 2011.
What should I do if I want to use the pft distribution and lai representative of year 2011 other than the 1850 or 2000 conditions ?

I am wondering whether the following settings would realize it.
1) Set the fsurdat in the user_nl_clm file. No matter simyr1850 or simyr2000 would be used (right ?)
2) Turn on the dynamic pft/crop/harvest options in the user_nl_clm file.
3) Assign a file to the flanduse_timeseries.

I am not sure which file should I assign to flanduse_timeseries.
If I set
flanduse_timeseries = '/glade/p/cesmdata/cseg/inputdata/lnd/clm2/surfdata_map/landuse.timeseries_0.9x1.25_hist_78pfts_CMIP6_simyr1850-2015_c170824.nc',
whether clm will automatically use the the variable values in 2011 other than the first year 1850 ?

Or should I manually extract the variable values in 2011 from the the simyr1850-2000 file and output them into a single file and then assign it to the flanduse_timeseries ?

Sincerely,

 

[oleson]

If the model start year is 2011 your steps 1-3 will work. Just set flanduse_timeseries and CLM will automatically use the data for year 2011.

 

[AndrewY]

If the model start year is 2011 your steps 1-3 will work. Just set flanduse_timeseries and CLM will automatically use the data for year 2011.

Hi Keith,

I have some questions about the start date of model running based on your response to Xueli's post:

Is it initial condition file that decides on which year the model start to run? So the model start year would be 2000 if "simyr2000" is included in the name of the input initial condition file (finidat).

For Xueli's case, in addition to changes made to those namelist variables, would it be necessary to also change RUN_STARTDATE to 2011 in env_run.xml?

Thank you!

 

[oleson]

The initial conditions files doesn't determine the start date of the model. It is RUN_STARTDATE.

 

[AndrewY]

The initial conditions files doesn't determine the start date of the model. It is RUN_STARTDATE.

Thank you Keith,

I did find some other threads about similar questions but I still don't fully understand how the model works concerning the start date and input data. My questions are listed as follows:

1. If I set RUN_STARTDATE=0001-01-01 and the initial conditions file used is '/work6/data/cesm/inputdata/lnd/clm2/initdata_map/clmi.I2000Clm50BgcCrop.2011-01-01.1.9x2.5_gx1v7_gl4_simyr2000_c180715.nc', then which year does '0001' refer to? Is it year 2000?

2. If I'm using the compset F2000climo, will the model use the forcing of the year 2000 for every single year simulated?

Your help would be much appreciated!

 

[oleson]

1. It refers to year 2000 for the surface dataset. If the case is out of the box (you haven't changed anything) in cesm2.1.3 then the atmospheric forcing would be 1991-2010 (set the case up and look at "DATM_CLMNCEP_YR_START" and "DATM_CLMNCEP_YR_END"). So year 0001 would correspond to 1991 atmospheric forcing.

2. I don't know much about coupled cases, but the name implies that the "forcing" data (input datasets for cam and clm) would either use year 2000 or a year 2000 climatology ("climo").

 

[Maggie Xia]

1. It refers to year 2000 for the surface dataset. If the case is out of the box (you haven't changed anything) in cesm2.1.3 then the atmospheric forcing would be 1991-2010 (set the case up and look at "DATM_CLMNCEP_YR_START" and "DATM_CLMNCEP_YR_END"). So year 0001 would correspond to 1991 atmospheric forcing.

2. I don't know much about coupled cases, but the name implies that the "forcing" data (input datasets for cam and clm) would either use year 2000 or a year 2000 climatology ("climo").

Hi Keith,

I also have several querstions based on this post:

1. When I run an I2000 composet, the run_startdate is 2000-01-01 and the default GWSP3 forcing data is from 1991-01 to 2010-12. What shall I do to force 2000-01-01 with the corresponding data at the same time?

2. I am confused about how surface data and initial data work? Let's say if I am creating a regional run and want to modify the PFT or LAI with satellite data, what shall I do? Do I have to modify the surface data and spin-up for several hundreds of years, and then create a new run pointing to the new initial data?

Thank you for the help!!

 

[oleson]

1. xmlchange DATM_CLMNCEP_YR_START=2000
xmlchange DATM_CLMNCEP_YR_END=2000
xmlchange DATM_CLMNCEP_YR_ALIGN=2000

2. I'm not sure I understand the question. But, modifying LAI implies that you are using satellite phenology (SP) mode, in which case you'd probably only need to spinup up for about 50 years at most, then use the restart file at the end of that for your production run. BGC runs require hundreds of years for spinup. See the following for an explanation of SP spinup:

1.5.6. Spinning up the Satellite Phenology Model — ctsm CTSM master documentation