Use of the CLM land use tool to create land time series data in standard CLM format

[CGL]

Hi,everyone！

I have recently generated land use data for the years 2015–2100. My data format is similar to LUH2, but the land unit in my data represents the area (in square kilometers) of different land types within each grid cell. My land use data includes seven plant functional types (PFTs), namely forest, shrub, grass, urban, snow, spare, and crops. The spatial resolution is 0.083 degree.

My CESM version is 2.1.3, CLM version is CLM5.

Based on my current understanding and research, if I want to convert this data into the standard format usable by CLM, it seems there are two approaches.

One is to use the CLM land use data tool to generate the corresponding PCT/PFT files, and then employ the mksurfdata.pl script within mksurfdata_map to obtain the standard CLM land use timeseries files and surface file. I am not familiar with CLM land use data tool. If I want to use it to convert my data, where can I obtain the tool and get help with its usage?

2.27. Transient Land Use and Land Cover Change — ctsm release-clm5.0 documentation

https://www2.cesm.ucar.edu/models/cesm1.2/clm/models/lnd/clm/doc/UsersGuide/x11868.html

The other approach is that I need to generate SCRIP grid files before I can proceed to produce the standard format used in CLM5. However, my data is a global land use dataset, while SCRIP grid files seem to be intended for regional datasets. In this case, I am not sure how to proceed.

1.3.4. Creating input for surface dataset generation — ctsm release-clm5.0 documentation

 

[oleson]

The CLM land use data tool technical note is here:

GitHub - lawrencepj1/CLM5-Land-Data-Technote: This is the working documentation for the CLM5 Land Data

This is the working documentation for the CLM5 Land Data - lawrencepj1/CLM5-Land-Data-Technote

github.com

If you need further help with that, you'll have to contact @lawrencepj1 .

However, the tool probably takes certain inputs that you may not currently have or that are incompatible with what you have. For example, you list snow as one of your pfts, but that is a prognostic quantity in the model and interacts with the prescribed vegetation. I'm also not sure what "spare" is.

Your other approach mentions SCRIP files. These can be global or regional. They can be used in the process of generating surface datasets at some unsupported resolution, e.g., your 0.083. But you will still need to translate your landcover types to "raw" input files for the surface dataset creation routines.