Make your own surface data

CTSM needs lots of surface variables to run.

CTSM technical notes has the official tutorial for creating surface datasets.

CTSM technical notes 1.3.5 Creating Surface data

However, it is difficult to use the tool for users that are not familiar to CESM/CTSM. Because in pyclmuapp, we mainly focus on urban climate and the percentage of urban is set to 100%, which means the output are only urban related. Hence, we can easily over-write the urban surface data in the default surfdata.nc provided by pyclmuapp, this avoids the complex steps of using the mksurfdata_map tools.

How to get our own urban surface data of the interest point from CESM provided file

Ref: - Technical Description of an Urban Parameterization for the Community Land Model (CLMU)

Table 3 Input data required for the urban model

Workflow

get the urban surface input from `mksrf_urban_0.05x0.05_simyr2000.c170724.nc <https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata/lnd/clm2/rawdata/>`__
get the soil parameter for the pervious road in urban from `mksrf_soitex.10level.c010119.nc <https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata/lnd/clm2/rawdata/>`__
You can directly use the provided surfdata.nc by pyclmuapp, which is located at “pyclmuapp/usp/surfdata.nc”. (No need to input)

[1]:

%time
from pyclmuapp import get_urban_params
urban = get_urban_params(urban_ds='data/mksrf_urban_0.05x0.05_simyr2000.c170724.nc', # can be a xarray dataset or a path to a netcdf file
                 soil_ds='data/mksrf_soitex.10level.c010119.nc', # can be a xarray dataset or a path to a netcdf file
                 # here we use the default vancouverCAN dataset provided by CLM5.0 as template
                 lat = 51.5116,
                 lon = -0.1167,
                 #template='surfdata.nc',
                 # optional, if not provided, the default surfdata dataset will be used
                 # the default surfdata dataset is the vancouverCAN dataset provided by CLM5.0 as follows
                 PTC_URBAN=[0,0,100.0], # percentage of urban land use in each density class, sum should be 100
                 # this means that the urban land use is 100% in the MD class
                 outputname='data/surfdata_london.nc',)
urban # the created surfdata.nc file

CPU times: user 1e+03 ns, sys: 0 ns, total: 1e+03 ns
Wall time: 3.1 µs
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
No suitable point found in the search range, expanding search range.
Found suitable point at lat: 51.4116, lon: -0.1167
Found suitable point at lat: 51.4116, lon: -0.1167

[2]:

urban['URBAN_REGION_ID'].values, urban['LONGXY'].values, urban['LATIXY'].values

[2]:

(array([[19]], dtype=int32), array([[359.875]]), array([[51.525]]))

(optional) more to know about the urban parameters

fist, let’s check the mksrf_urban_0.05x0.05_simyr2000.c170724.nc

[3]:

from matplotlib import pyplot as plt
import xarray as xr

[4]:

# open the downloaded mksrf_urban_0.05x0.05_simyr2000.c120621.nc
global_surf = xr.open_dataset('data/mksrf_urban_0.05x0.05_simyr2000.c120621.nc')
fig = plt.figure(figsize=(12, 3))
for i in range(3):
    ax = fig.add_subplot(1, 3, i+1)
    global_surf['PCT_URBAN'].sel(density_class=i).where(global_surf['PCT_URBAN'].sel(density_class=i) > 0).plot(ax=ax, add_colorbar=False)
    print(f"Region {i}:", global_surf['PCT_URBAN'].sel(density_class=i).where(global_surf['PCT_URBAN'].sel(density_class=i) > 0).count().values, "grid cells")
    ax.set_title(f'Region {i}')
plt.tight_layout()
plt.show()

Region 0: 348 grid cells
Region 1: 74945 grid cells
Region 2: 278922 grid cells

../../_images/notebooks_own_prepare_urbansurf_10evlurb_9_1.png

then check mksrf_soitex.10level.c010119.nc

[5]:

soil = xr.open_dataset('data/mksrf_soitex.10level.c010119.nc')
soil = soil.assign_coords(lat=soil.LAT, lon=soil.LON)   # add lat and lon as coordinates
fig = plt.figure(figsize=(4, 3))
soil['MAPUNITS'].plot()
plt.tight_layout()
plt.show()

../../_images/notebooks_own_prepare_urbansurf_10evlurb_11_0.png

Make your own surface data

Workflow

(optional) more to know about the urban parameters

(Optional) here is the step for generate a template.