Soil properties in single-point mode

[thiagods@umich_edu]

One step required to prepare single-point simulations in CLM is to prescribe soil texture at the site location to the model. One single percentage of each soil separate is prescribed (sand, clay and silt).






In terms of how soil texture is distributed across all soil layers, how does CLM handle that in a single-point run? Does the model interpolate the prescribed texture across the soil layers? Or does the model assume a constant soil texture for the whole soil column?






Thanks,


Thiago.

 

[oleson]

If you are creating single-point surface datasets and prescribing sand and clay percent, then the model assumes that soil texture for all layers.However, you can modify the surface dataset after the fact to prescribe different soil textures for each layer.The global datasets have depth-varying soil texture.

 

[thiagods@umich_edu]

Thanks for the reply Keith.






I just took a peek at the "surfdata_1x1pt_CA-Oas_hist_16pfts_Irrig_CMIP6_simyr2000_c190311.nc" file that I had created for the "CA-Oas" Fluxnet site. I confirm that the same soil texture I prescribed during the creation of the surface datasets was applied for 10 soil layers (nlevsoi = 10).






However, in the output files (of a CLM5 run) I noticed that levgrnd = 25. So I am guessing that CLM5 interpolates the soil texture parameters some time at the beginning of the run to match the higher number of soil layers in the model?






Cheers,


Thiago.

 

[oleson]

The soil texture for the rest of the layers is taken from the 10th layer (nlevsoi=10) in the code, yes.

 

[thiagods@umich_edu]

I appreciate your replies to this topic Keith!