OSVAS

Step 1. Preparation of forcing data

Notebook execution

The Python workflow launcher executes the notebook WRITE_Station_forcing.ipynb to generate SURFEX forcing data. The notebook is executed using one of these methods:

1. Automatic conversion (default): The launcher uses jupyter nbconvert to convert the notebook to a Python script on-the-fly and executes it. This approach works in environments without a Jupyter server.
2. Direct Python execution: A pre-converted WRITE_Station_forcing.py script is available and can be executed directly if notebooks are unavailable.
3. Interactive Jupyter: The notebook can be manually executed with jupyter notebook or jupyter lab for development and debugging.

Forcing data generation

The notebook generates SURFEX forcing files in ASCII or NetCDF format based on the Forcing_data block in the station YAML:

Forcing_data:
  height_T: 2                       # Height of the temperature measurement (meters)
  height_V: 10                      # Height of the windspeed measurement (meters)
  run_start: '2021-05-01 00:00:00'  # Timestamp for the forcing start
  run_end: '2021-07-01 23:30:00'    # Timestamp for the forcing end
  forcing_format: 'ascii'           # Choose between 'netcdf' or 'ascii'
  dataset1:
    doi: https://meta.icos-cp.eu/objects/fPAqntOb1uiTQ2KI1NS1CHlB
    timedelta: 30  # in minutes
    variables:
      Forc_CO2: -, 0.00062
      Forc_PS: PA, *1000
      Forc_RAIN: P, /(timedelta*60)
      Forc_SNOW:
      Forc_WIND: WS
      Forc_DIR: WD
      Forc_DIR_SW: SW_IN
      Forc_LW: LW_IN
      Forc_QA:
      Forc_SCA_SW:
      Forc_TA: TA, +273.15
      Forc_RH: RH

Output formats

ASCII format:

• Forcings are generated for the entire simulation period in a single ASCII file
• Simple columnar format suitable for small domains or single-point models like SURFEX OFFLINE
• File: forcings/{STATION_NAME}/Forc_*.txt

NetCDF format:

• Daily NetCDF files are created by the notebook
• A merge function combines daily files into a single FORCING.nc file
• Files: forcings/{STATION_NAME}/FORCING.nc and forcings/{STATION_NAME}/Forc_*.txt

Variable transformations

The variables section supports:

• Pass-through: Forc_WIND: WS — Use the ICOS variable directly
• Arithmetic: Forc_TA: TA, +273.15 — Convert to Kelvin; Forc_PS: PA, *1000 — Convert to pascals
• Rate conversion: Forc_RAIN: P, /(timedelta*60) — Convert accumulation to rate (automatically substitutes timedelta value)
• Wildcards: TS_*: TS_*, SWC_*: SWC_* — Expand to all matching ICOS variables
• Constants: Forc_CO2: -, 0.00062 — Use a fixed value

Sampling rate

If multiple datasets with different sampling rates are provided, all data are upsampled to the finest (smallest) timedelta. For example, if dataset1 has 30-minute data and dataset2 has hourly data, both will be interpolated to 30-minute intervals.

Lockfile mechanism

A lockfile (FORCING.lock) prevents overwriting of existing forcing files. To regenerate forcing data, manually delete the corresponding .txt or .nc files.