source:sandbox/clm/sva_DWD_forcing.ncl

The purpose of downscale is to substitute model2 with the community land model, CLM, and use the downscaled data as forcing. Replacing model1 with an online COSMO should then be easily conceivable. For the time being, every effort aims at CLM (Community Land Model)’s requirements. They are presented in an NCL (NCAR Command Language) file of another project TR32-Z4-Tools: sva_DWD_forcing.ncl

Variables

Tab.: The table in the header of the original sva_DWD_forcing.ncl

; *************************forcing data*****************************************

; COSMO-DE forcing data for land surface model (CLM3.5)                        *

; CLM3.5 Variable        COSMO Variable           Local Variable               *

; ---------------        --------------           --------------               *

; LONGXY                 g10_lon_1                longxy                       *

; LATIXY                 g10_lat_0                latixy                       *

; TBOT                   T_GDS10_HTGL_13          tbot            (T_2m)       *

; WIND                   U_GDS10_HTGL_13,         wind            (U,V_10m)    *

;                        V_GDS10_HTGL_13                                       *

; QBOT                   TD_GDS10_HTGL_13,        qbot            (TD_2m)      *

;                        PS_GDS10_HTGL_13                                      *

; PRECTmms               TOT_PREC_GDS10_SFC_13    prectmms                     *

; FSDS                   ASOB_S_GDS10_SFC_13,     fsds                         *

;                        ALB_RAD_GDS10_SFC_13                                  *

; FLDS                   ATHB_S_GDS10_SFC_13,     flds                         *

;                        T_GDS10_SFC_13                            (T_G)       *

; PSRF                   PS_GDS10_SFC_13          psrf                         *

This project has to provide the variables of the center column of the table, or some reasonable stand-ins. These will become necessary even though downscale uses COSMO (Consortium for Small-scale Modeling) data as a compromise between better downscaling and more suitable variables. In the project, the best possible downscaling is conducted by the full 3-step Schomburg scheme. If the 2nd step lacks the appropriate determinist a-priori rule, called Schomburg rule, for the variable, it does nothing. Therefore, even in case downscale has just the variable which CLM (Community Land Model) requires, it may pay to use one which is only similar, but better downscaled.

Example 1: Temperature

The 1st example for this is temperature. CLM (Community Land Model) requires 2 m-temperature, which downscale could gladly provide. Unfortunately, there are no Schomburg rules for 2 m-temperature. Fortunately, there are rules for the temperature of the lowest layer, an obviously similar field. So, downscale provides the latter to CLM (Community Land Model) with metadata such that the surface model will not be the wiser.

Example 2: Wind

A counterexample for this compromise is 10 m-wind. Again, downscale can provide it, and there are not any Schomburg rules for 10 m-wind. Unfortunately, there are not even Schomburg rules for any variable like wind, say the wind field of the lowest model layer. So, downscale uses the exactly accurate variable which was only spline-interpolated and not properly downscaled.

Overview

Given the 2 examples, this section sorts the remaining variables into wind-types,

• the wind itself U_GDS10_HTGL_13 and V_GDS10_HTGL_13,
• the dew point at 2 m TD_GDS10_HTGL_13,
• the precipitation TOT_PREC_GDS10_SFC_13,
• ASOB_S_GDS10_SFC_13,
• ALB_RAD_GDS10_SFC_13,
• ATHB_S_GDS10_SFC_13, and
• the ground temperature T_GDS10_SFC_13,

temperature-types,

• the 2 m-temperature itself T_GDS10_HTGL_13 corresponding to the temperature of
  the lowest layer, and
• the pressure of the lowest layer PS_GDS10_HTGL_13 corresponding to the surface
  pressure

and exceptions:

• The variables g10_lon_1 and g10_lat_0 are outside the scope of this project.
  Programs like the Climate Data Operators should work better.
• The surface pressure PS_GDS10_SFC_13 precisely fits a rule.

Processing

The script sva_DWD_forcing.ncl aims to produce a NetCDF file with the variables of the left column of the table. Since their definition is subtly different than COSMO’s, the original sva_DWD_forcing.ncl
performed some processing, some of which required the use of multiple variables. This is why some rows have a blank cell in the column “CLM3.5 Variable” - they correspond to the variable in the preceding row and are processed by the script.

The processing is not invariant to downscaling with the Schomburg scheme. In this project, therefore, model1 does the necessary processing and only then downscaling comes into play. This required major revisions to sva_DWD_forcing.ncl.

External data for the Rur catchment

/daten01/z4/database/TestCases/Rur4as/griddata_0752x0720.nc

This data forces the planned CLM (Community Land Model) runs as external data alongside what active data this project produces. It is thus the ideal source for the correct coordinates sva_DWD_forcing.ncl crops from the output of model2 and the ideal resolution, which is still just out of reach of this project.

The directory contains another 4 files with similar information.

Coarse forcing

/daten01/z4/database/clm3.5/Rur_NRW/atm_forcing/

Any number of experiments have been conducted with CLM (Community Land Model) in the past. This directory contains some of their CLM (Community Land Model) inputs, certainly produced by the unchanged sva_DWD_forcing.ncl from TR32-Z4-Tools. Ideally, this project’s CLM (Community Land Model) input files should differ from those in nothing but resolution.

1. stored here as: document:“Wrapper for COSMO”

2. stored here as: document:“Wrapper for COSMO”

3. stored here as: document:“Wrapper for COSMO”

4. stored here as: document:“Wrapper for COSMO”

5. stored here as: document:“Wrapper for COSMO”

6. stored here as: document:“Wrapper for COSMO”

7. stored here as: document:“Wrapper for COSMO”

8. stored here as: document:“Wrapper for COSMO”

9. stored here as: document:“Wrapper for COSMO”

10. stored here as: document:“Wrapper for COSMO”

11. stored here as: document:“Wrapper for COSMO”

12. stored here as: document:“Wrapper for COSMO”

13. stored here as: document:“Wrapper for COSMO”

14. stored here as: document:“Wrapper for COSMO”

15. stored here as: document:“Wrapper for COSMO”