Transient (diagnostic) future RCP CO2 concentration for off-line CLM simulations

[koichi_sakaguchi@pnnl_gov]

Dear all,

I would like to run off-line CLM4 simulations using model outputs from CCSM4 MOAR RCP simulations.
Currently supported component sets for RCP scenarios (e.g., I_RCP8.5_CN) seems to use transient land cover and aerosol deposition under RCP scenarios, but the atmospheric CO2 concentration is kept at 2000 level.

I do not find such an example in the user guide yet and also do not find future CO2 concentration file (such as fco2_datm_1765-2007_c100614.nc for the historical period) under common CESM data directory on bluefire.

So my approach is to download CO2 concentration from RCP data website and replace the values for CO2, time, and date variables in fco2_datm_1765-2007_c100614.nc with those from RCP data, then use it as a stream into datm.

If anyone knows future CO2 concentration file already made for each RCP scenarios somewhere on bluefire, or any other easier options, please let me know.


Sincerely,

Koichi

 

[slevis]

Dear Koichi,

I'm surprised that the I_RCP compsets do not use the transient CO2...
Your solution to this problem seems reasonable. Here's another option that you could try, but I have not tried it, so no guarantees:
Begin to set up one of the B_RCP compsets. I expect that this would use transient CO2, so this way you'd gain access to the corresponding file.

Sam

 

[koichi_sakaguchi@pnnl_gov]

Dear Sam,

Thank you so much for your suggestion!

I tried configuring B-RCP8.5 case, and the model seems to get the transient CO2 (and other greenhouse gas) concentration from

$DIN_LOC_ROOT/atm/cam/ggas/

where I can see these files:

ghg_rcp26_1765-2500_c100405.nc
ghg_rcp45_1765-2500_c100405.nc
ghg_rcp60_1765-2500_c100901.nc
ghg_rcp85_1765-2500_c100203.nc


these files do not have the spatial coordinates which may be needed by data atm model. So I am thinking to use the values and time coordinates in these files to replace the CO2 and time variables in the historical CO2 forcing file.

I will update how it goes here later.

Again, thank you for your help.

Koichi