Model Start date of F_2000_CAM5

[chen63@illinois_edu]

I know that F_2000_CAM5 is running with year 2000 emissions. How about the actual model start date of this compset? I ran compset F_2000_CAM5 with all default settings, and found the name of the CAM output files is like $CASE.cam2.h0.0001-01.nc. Does this mean the model start from year 0001 but not 2000? One more question is that what does this year 0001 mean? Is is 0001 A.D.? If I want the model starts from December 1999, should I change the xml_run variable RUN_STARTDATE? Are there any other changes I have to make?  Thanks!

 

[hannay]

When you run the F_2000_CAM5 compset you cycle over datasets that represents present day (typically the year 2000).
The start date 0001-01 means it is the first year of tour un that represents the year 2000.
If you run 10 years for instance, the date will increment from 0001-01 to 0010-12 (all these years representing the year 2000 but with some climate variability).
Then you can compute the climatology over these 10 years.
Hope this helps

 

[chen63@illinois_edu]

Hi, Hannay,
THanks for your answer. One more question is that how can I start the run from the previous winter, i.e., December 1999, so I can still have the data starting from the beginning of year 2000 after I throw out the first month? How do I change the RUN_STARTDATE?
Thanks!

 

[hannay]

You need to set the RUN_STARTDATE to 1999-12. However, if you do this you need to provide initial conditions for December 1. We provide initial conditions for January 1. So you would need to create initial conditions for December 1999 in another run. It is especially important for the land that takes a long time to spin up.

 

[chen63@illinois_edu]

Since I am new to CESM, could you please provide me some hints on how to create initial conditions for December 1999 in another run?
Which compset do I have to run to create the initial condition for December 1999?
Is there any other ways to let the F_2000_CAM5 start from a winter - December? Can I change the RUN_STARTDATE to 0001-12? Does this mean the model starts with the climatology of December 2000? Will the model automatically pick the emission of December 2000?
Thanks a lot!

 

[hannay]

You can create monthly initial conditions by changing the namelist parameter "inithist"
See: documentation about inithist in:
http://www.cesm.ucar.edu/cgi-bin/eaton/namelist/nldef2html-cam5_1_1
However if you are focus on the year 2000 specifically, I am not sure the compset: F_2000_CAM5 is what you need. This compset represents "present day" and not specifically the year 2000. For instance, it uses climatological SSTs computed over the years: 1982-2001.If you want to focus on the year 2000, you may want to run a F_AMIP_CAM5 compset.

Also keep in mind, that running 1 single year is not meaningful as you don't capture the variablility of the model. So you may want to run multiple instances of the year 2000 perturbing initial conditions using the namelist parameter  "pertlim"

 

[chen63@illinois_edu]

Thanks for your answers. I still have some questions here.So F_2000_CAM5 is running with year 2000 emissions but climatological SSTs from 1982 to 2001. Does this mean for year 1 in this run, the model is using climatological SSTs from 1982, for year 2 using 1983 and so on?In the cesm userguide, the description of compset F_AMIP_CAM5 is "Description: INVALID: AMIP runs not supported at this resolution"What does this mean?

 

[hannay]

By climatological SSTs over the period 1982-2001, I mean an average over the period 1982-2001.
The SSTs have an annual cycle: it is a 12-month datasetThe month 01 is an average of January 1982-2001
The month 02 is an average of February 1982-2001
... About the question:"Description: INVALID: AMIP runs not supported at this resolution" What does it means.

It means you don't get all the input datasets out of the box at this resolution. You have to provide your own datasets.

 

[chen63@illinois_edu]

Hi, Hannay,How can I know which SST data file is used for the compset F_2000_CAM5 in CESM1.0.4 as I wasn't able to find the variables SSTICE_DATA_FILENAME, SSTICE_YEAR_ALIGN, SSTICE_YEAR_START and SSTICE_YEAR_END in env_run.xml.Secondly, just want to confirm that, F_2000_CAM5 is a free-running setting as the case is constrained by realistic SST and sea ice, and all the atmospheric processes are computed within the model. Am I correct?Thanks! 

 

[hannay]

The variables that control SST file have been renamed in cesm1.1 (and later version).  ## In cesm1.0 (your case)The variables are in env_conf.xmland the variable names are:DOCN_SSTDATA_FILENAME
DOCN_SSTDATA_YEAR_START
DOCN_SSTDATA_YEAR_END ### In cesm1.1 (and later), the variables are located in:
env_run.xml
and the variable names are:
SSTICE_DATA_FILENAME
SSTICE_YEAR_START
STICE_YEAR_END

 

[hannay]

About your second question:I am not sure what you mean by "free-running setting"

In F_2000_CAM5, SST and sea-ice are prescribed using the Hurrell dataset. Hurrell, James W., James J. Hack, Dennis Shea, Julie M. Caron, James Rosinski, 2008: A New Sea Surface Temperature and Sea Ice Boundary Dataset for the Community Atmosphere Model. J. Climate, 21, 5145–5153.Other datasets are prescribed too (CO2, ozone, aerosol emission, etc..).
The model computes the atmospheric state using the governing equations for hydrostatic atmosphere (se documentation for details).

 

[chen63@illinois_edu]

Thanks! But I wasn't able to find the SST file for compset FC5 (F_2000_CAM5) in CESM1.0.4 as variables in env_conf.xml are set as follows:As these variables are set only for amip runs, where can I find the SST file for this compset which is not an amip run?Thanks! 

 

[chen63@illinois_edu]

Now I have figured it out! Thanks very much for your kind help!

 

[hannay]

YOu can see the name of the sst files in the namelist files (ice_in and docn.streams.txt).For instance in ice_in:
look for the variables: stream_fldfilenameIf you want to use another SST, you need to change this in env_conf.xmlHope this helps