LULCC experiments

[San Diego WX]

Hi there. I would like to run CESM to quantify the impact of deforestation. The plan is to change the land cover type (PFT) in CLM, over a specific region, then compare the run results against a control case. So I wondering how can I generate a new land cover map/file for CLM? The case is FHIST and res is 0.9x1.25.
Thank you!

 

[oleson]

Easiest way is to modify the existing surface dataset (identified by fsurdat in your lnd_in) using whatever tools you use to manipulate netcdf files.
Keep in mind the following rules:

PCT_CROP
PCT_NATVEG
PCT_LAKE
PCT_WETLAND
PCT_GLACIER
PCT_URBAN

These are the CLM landunits and the sum of these has to equal 100% at each gridcell. Note that unlike other landunits, there are up to three urban landunits per gridcell (PCT_URBAN).

Also on the surface dataset is:

PCT_CFT

If you are using a 16pft dataset (SP mode), the sum of the two crop types (15 - C3 Unmanaged Rainfed Crop and 16 - C3 Unmanaged Irrigated Crop) must equal 100%, even for gridcells where PCT_CROP is zero (in this case crop type 15 is set to 100% and crop type 16 is set to 0%). These two crop types represent the non-irrigated/irrigated fractions on the crop landunit given by PCT_CROP.
if you are using a 78pft dataset (crop model active), the sum of all of the crop types must equal 100%.

Also on the surface dataset is:

PCT_NAT_PFT

The sum of these must equal 100%. They represent the fractions of pfts on the natural vegetated landunit given by PCT_NATVEG.

 

[San Diego WX]

Thank you, Keith. I hadn't thought of using a netcdf manipulator to make my land cover changes. That is a great idea.
If I may ask a follow-up questions. I noticed that my fsurdat is pointing to the file: surfdata_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr1850_c190214.nc. Does that means that I am using land cover data based on a CMIP6 scenario and not actually real data based on observations and remote-sensing data? Thanks

 

[oleson]

Sorry, I now see that you are using the FHIST compset, which specifies transient land cover (changes over time). So actually you would have to make changes to the landuse file (identified by flanduse_timeseries in your lnd_in). Or pick a non-transient F compset and modify the fsurdat file.
All of the landcover type surface data is based on Land Use Harmonized version 2 (LUH2) data as described here:

2.27. Transient Land Use and Land Cover Change — ctsm CTSM master documentation

The LUH2 historical data is based on observations and remotely-sensed data (to the extent that such observations and remotely-sensed data were available during the 1850-2015 period).

 

[San Diego WX]

Hi Keith. Thanks for the update and for sharing the information on the LULCC data.

 

[San Diego WX]

Hi Keith. I have been able to generate a new surfdata file for my runs. I am only running the model for 1 month so I do not need to worry about transient land cover for now. I have noticed that surfdata is defined in three different namelists: Buildconf/clm.input_data_list, CaseDocs/lnd_in, and Buildconf/clmconf/lnd_in. Do I need to update all three namelists? Would you mind explaining to me what each of the namelists do? Thank you.

 

[San Diego WX]

Hi Keith. Please disregard the previous message. Thank you!

 

[xiaoxiaokuishu]

Sorry, I now see that you are using the FHIST compset, which specifies transient land cover (changes over time). So actually you would have to make changes to the landuse file (identified by flanduse_timeseries in your lnd_in). Or pick a non-transient F compset and modify the fsurdat file.
All of the landcover type surface data is based on Land Use Harmonized version 2 (LUH2) data as described here:

2.27. Transient Land Use and Land Cover Change — ctsm CTSM master documentation

The LUH2 historical data is based on observations and remotely-sensed data (to the extent that such observations and remotely-sensed data were available during the 1850-2015 period).

Hi, Oleson,
I have a confusion of the relations between FHIST and use of flanduse.timeseries data, i think I can still use the FHIST with the flanduse.timeseries=blank?
Actually this combination FHIST and flanduse.timeseries="" also works well , but if i set like this, it can not be called a transient land cover simulation?
Beside, a non-transient F compset is specifically referring to F2000?

 

[slevis]

You asked similar or related questions in other threads and my responses are here:

Standardized setup method of transient land cover in IHIST

Hi, all, I perform a simulation using IHIST, but accidently set the namelist as follow: &dynamic_subgrid do_transient_crops = .false. do_transient_pfts = .false. flanduse_timeseries =...

bb.cgd.ucar.edu

https://bb.cgd.ucar.edu/cesm/threads/surfrd_get_data-error-sum-of-wt_lunit-not-1-0.6367/post-53301