Modifying Surface Dataset with Urban Features for Mountain Cases

[AquaTech_007]

Hi,

I am running SPT simulations using CLM5 and need guidance on better representing a mountainous case study site in the surface dataset. The site is primarily grassland, located at a high elevation 2500 m in a mountainous region with underlying impervious rock. it is recommended that in areas with impervious bedrock, a shallower bedrock depth should be used. However, I'm encountering an issue where snowmelt and the energy fluxes are simulated too late for few days at the start of snowmelt, typically around May or June. Is it valid to use some percentage of Urban in a grid cell after modifications like walls, roof etc in the surface dataset.​

 

[oleson]

Unless you actually have some urban fraction in your SPT domain, I wouldn't recommend doing that.
If you want to implement a shallower bedrock depth, then one way to do that is to modify the "zbedrock" variable on the surface dataset.

 

[AquaTech_007]

I have a question regarding the elevation of the study site, which I believe was taken from the raw data file: `topodata_0.9x1.25_USGS_070110_stream_c151201.nc`.

The site is located in the Italian mountains at an elevation of approximately 2500 m but according to the index in the above file, the elevation is shown as 905 m. This value seems inaccurate, likely due to the coarse spatial resolution of the dataset. The surface dataset generated by the model have slope of 5.6 degrees and a standard deviation of elevation (STD_ELEV) = 0, which also appears to be incorrect. I downloaded the corresponding pixel from the (USGS HYDRO1K 1-km) dataset and calculated the standard deviation over a 1 km average, which came out to be around 120 m.

Because of this difference, I'm worried the model might not be correctly considering the site's elevation. Is there a way to manually set or override the site elevation in the `user_nl_clm` file?

There is another site around 15 miles away from this one, and the surface dataset generated values for slope, STD_ELEV, and topographic elevation are much closer to the correct values. Thanks in advance.

 

[slevis]

I don't think that there is a way to override elevation via the namelist, though you should check the user's guide to make sure. If there isn't, then you may consider searching where elevation appears in the code, and hardwiring the override directly in the code.

 

[AquaTech_007]

I have found it in 'src/main/TopoMod.F90'. It can be modified there, thank you Slevis.