how to spin us for a tranisent land cover simulation with f2000

[xiaoxiaokuishu]

Hi all,
I made a simulation using F2000 from 1990-2014 with transient land cover, for flanduse_timeseries data , I modified the land cover data for the Amazon region for the period 1990/2014. For this case, how to get the initdata from the new landusetime data? What i now use is the finidat = '/scratch/snx3000/rxu/CCLM2_inputdata/cesm_inputdata/lnd/clm2/initdata_map/clmi.F2000.2000-01-01.ne120pg3_mt13_simyr2000_c200728.nc from NCAR, but i am wonder do
i need to get the inidata/restart data by myself as the landuse_timeseires data is different from what i use now.

Besides, for my case that aims to evalute the real land cover change, i should select f2000 or fhist ?

Best

 

[oleson]

When you say you made a simulation using F2000 with your modified landcover, does that mean you actually ran it? If so, that simulation will produce restart files that can be used as an finidat for another simulation. They should be in your archive directory (*.clm2.r.*.nc). If you haven't run a simulation yet, then the initial file you mentioned will be interpolated to your new surface data upon initialization. If your landcover is transient, e.g., 1991 is different from 1990, then you'd use FHIST so that your flanduse_timeseries data is used.

 

[xiaoxiaokuishu]

When you say you made a simulation using F2000 with your modified landcover, does that mean you actually ran it? If so, that simulation will produce restart files that can be used as an finidat for another simulation. They should be in your archive directory (*.clm2.r.*.nc). If you haven't run a simulation yet, then the initial file you mentioned will be interpolated to your new surface data upon initialization. If your landcover is transient, e.g., 1991 is different from 1990, then you'd use FHIST so that your flanduse_timeseries data is used.

Yes, I use the f2000 compset, and set do_transient_xx=true, and give the modified landusetime series data as I have mentioned, it works well (I have got the xx.clm2.r.2014-01-01-00000.nc), so i do think it is a transient land cover simulation even its compset is f2000, right?

Actually i use the initidal data from your side, finidat = '/scratch/snx3000/rxu/CCLM2_inputdata/cesm_inputdata/lnd/clm2/initdata_map/clmi.F2000.2000-01-01.ne120pg3_mt13_simyr2000_c200728.nc. Do i need to make such inidat by myself, for example, first run a 50-years simulation (e.g. from 1943 to 1992) to get the restart file, then set this restart file as
finidat ="restart file", then make the sam simulation with f2000 at 1992-2014?

 

[slevis]

@xiaoxiaokuishu this is great that you ran successfully. The answers to your questions seem to be about experimental design. You have to decide what experimental design will work best for your research questions. We do not offer support in that category of questions.