Cesm release only supports a fixed set of out-of-the-box compsets. So if I don't use any new data_mode and %phys fordifferent component, but just use a different forcing data set, do I need to define a new comset? Case 1: One out-of-the-box I compset is "I2000Clm50SpGs" which is driven by GSWP3v1 forcing. If I just want to drive the modelwith CruV7 forcing while keeping other components and physics uchanged (Cruv7 is one of the supported forcing, but the combination isnot supported out of the box), should I create a new compset? Or I just use "I2000clm50SpGs" to create the case,then changethe streams in "datm_in"? If I need to create a new compset, which file(s) I should edit? 2. case 2: similar to case 1, but use a forcing data which is not supported by cesm at all, such as MERRA2. I assume I should create a new compset following section 2.3 on https://esmci.github.io/cime/users_guide/compsets.html ? Any information is appreciated.
define new compset for DATM
- Thread starter xgao304
- Start date
Hi Xgao, in my experience there is no need to define a new compset when you only want to change the forcing dataset in CLM. You may try the following:1. Setup a case and run ./preview_namelist
2. from the CaseDocs directory, copy these files to some other directory:datm.streams.txt.CLMCRUNCEP.Precip
datm.streams.txt.CLMCRUNCEP.Solar
datm.streams.txt.CLMCRUNCEP.TPQWNOTE: They may be named differently in your case. Importantly they contain the part "streams.txt" , which means that they contain the filenames for the forcing "streams" for temperature, insolation and the like. I will call these text files "stream files" from now on.3. Copy the stream files again, now into the $CASEROOT and rename themuser_datm.streams.txt.CLMCRUNCEP.Precip
user_datm.streams.txt.CLMCRUNCEP.Solar
user_datm.streams.txt.CLMCRUNCEP.TPQW4. Edit these files to your liking (i.e. change them so that they point to the other forcing dataset. If you have example stream files ready from another case with that forcing would greatly speed up the process. If not, you can try doing it manually).5. ./preview_namlists again and inspect the newly generated stream files in CaseDocs6. Test with a single month. Check the CLM log that the correct file is read at the right time (if not, fiddle with these settings in env_run.xml : DATM_CLMNCEP_YR_START, DATM_CLMNCEP_YR_ALIGN)Good luck.Leo
2. from the CaseDocs directory, copy these files to some other directory:datm.streams.txt.CLMCRUNCEP.Precip
datm.streams.txt.CLMCRUNCEP.Solar
datm.streams.txt.CLMCRUNCEP.TPQWNOTE: They may be named differently in your case. Importantly they contain the part "streams.txt" , which means that they contain the filenames for the forcing "streams" for temperature, insolation and the like. I will call these text files "stream files" from now on.3. Copy the stream files again, now into the $CASEROOT and rename themuser_datm.streams.txt.CLMCRUNCEP.Precip
user_datm.streams.txt.CLMCRUNCEP.Solar
user_datm.streams.txt.CLMCRUNCEP.TPQW4. Edit these files to your liking (i.e. change them so that they point to the other forcing dataset. If you have example stream files ready from another case with that forcing would greatly speed up the process. If not, you can try doing it manually).5. ./preview_namlists again and inspect the newly generated stream files in CaseDocs6. Test with a single month. Check the CLM log that the correct file is read at the right time (if not, fiddle with these settings in env_run.xml : DATM_CLMNCEP_YR_START, DATM_CLMNCEP_YR_ALIGN)Good luck.Leo