Model Information

Running the model

Plotting maps using cuwalid forecasting is done by running this command in the terminal:

# replace "input.json" with the path to your input json
python -m cuwalid.forecasting.main_impact_forecast input.json

Or you can import the function into your code like this:

from cuwalid.forecasting.main_impact_forecast import plot_maps_json

# replace "input.json" with the path to your input json
plot_maps_json("input.json")

Input Parameters

The best way to build the input files for forecast impact is to use the Input Helper.
Here you can find an easy-to-use interface to create the file and others with clear descriptions.

Here are the parameters used for the Forecast Impact model:

  • plot_scales: List of geographical scales for plotting (e.g., ["County", "Zoom"]).

  • country_name: Country to generate forecast for (e.g., "Kenya").

  • place_names: List of specific regions or counties of interest (e.g., ["Samburu", "Laikipia", "Meru"]).

  • seasons: Forecast season (e.g., "OND" or "MAM").

  • water_status: Type of water-related event being forecasted (e.g., "Flood").

  • year: The year of the forecast (e.g., 2022).

  • language: Language for the output (e.g., "English").

  • output_dir: Directory for saving outputs (e.g., "output").

  • netcdf_path: Path to the NetCDF forecast dataset. Use 'YYYY' in the filename to represent the selected year dynamically.

  • threshold_path: Path to the threshold file. Use 'SSS' in the filename to represent the selected season dynamically. Do not include _flow_quantiles.nc or _quantiles.nc in the name — these will be added automatically by the script.

  • mask_path: Path to the mask file used to limit forecast region.

  • river_path: Path to river network data for hydrological features.

  • shape_path: Optional path to custom shapefile to override built-in HAD region shapefiles.

Optional (auto-generate netcdf/threshold paths instead of specifying them directly)

  • prev_output_path: Directory from previous model run that contains the output NetCDF files (e.g., "forecasting_dataset/HAD/output/").

  • pp_path: Post-processing threshold files path (e.g., "forecasting_dataset/HAD/postpp/").

  • model_name: The model name used during forecasting (to help script construct file names correctly).

⚠️ Note: If both netcdf_path and threshold_path are provided, they take priority over the automatically generated paths using prev_output_path, pp_path, and model_name.