.. _technical_des_wildfire_wise: Technical Description ===================== The WISE Fire Spread Model --------------------------- The Wildfires WISE application is based on the Canadian Wildfire Intelligence and Simulation Engine (WISE). WISE integrates three key components to predict fire behaviour: 1. **Weather** - Hourly atmospheric conditions (temperature, humidity, wind speed and direction, precipitation) 2. **Topography** - Digital elevation model capturing terrain slopes and aspects 3. **Vegetation/Fuel** - Spatial distribution of fuel types and their combustibility characteristics The model is based on the Canadian Fire Weather Index (FWI) and Fire Behaviour Prediction system, which use weather, topography and vegetation information to calculate fire behaviour :cite:`VanWagner_1987` :cite:`Hirsch_1996`. It uses the Prometheus fire spread model algorithm, originally developed in Canada for boreal and temperate forest applications. For more information, see the `Prometheus documentation `_ and the `WISE GitHub project `_. Model Inputs ~~~~~~~~~~~~ **Climate Data** (hourly resolution) - Temperature at 2 meters above ground - Relative humidity at 2 meters - Wind speed and direction at 10 meters - Hourly precipitation **Auxiliary data** - Digital Elevation Model (DEM) - Used to calculate slope and aspect effects on fire spread - Fuel Classification Map - Defines fuel type at each grid cell based on Canadian FBP system fuel classes (such as C-3 pine fuel class) Fire Simulation Parameters ~~~~~~~~~~~~~~~~~~~~~~~~~~~ - **Spatial Resolution**: Finland 16 m × 16 m; Portugal 100 m × 100 m - **Temporal Resolution**: Daily fire spread simulations - **Simulation Period**: Fire season months (May–August) - **Ignition Sites**: Specific historical locations (e.g., Kalajoki, Koli in Finland; PIISA sites in Portugal) - **WISE Version**: v1.0.6-beta (released 2023-03-01) Model Outputs ~~~~~~~~~~~~~~ The WISE model is used to produce the following fire behaviour outputs at each grid cell: - **Burn Grid** - Binary indicator of whether each grid cell burned during the simulation - **Maximum Fire Intensity** - Maximum simulated fire intensity at each grid cell (kW/m²) - **Maximum Flame Length** - Maximum simulated flame length at each grid cell (m) In the Climate DT WISE application, these three outputs are provided as NetCDF files. Integration in the Workflow ---------------------------- The WISE application is integrated into the Climate DT workflow as follows: 1. **Data Preparation**: Climate model output is used to derive the required fire-weather inputs (for example, relative humidity from temperature and dew point temperature, and wind speed/direction from u and v wind components) for each test site location 2. **Auxiliary Data Setup**: Digital elevation model and fuel classification maps are prepared at site resolution (Finland: 16 m × 16 m; Portugal: 100 m × 100 m) 3. **Fire Simulation**: The WISE fire spread model runs daily simulations during the fire season, using the climate data and auxiliary information as input 4. **Output Generation**: Fire behaviour grids are generated and transformed into NetCDF format .. Application runs are based on the run_wise.py script. In it, the run is started by extracting the requested data from the climate models, and extracting hourly temperature, relative humidity, wind speed and precipitation values for the fire simulation area. These are written as a table-formatted text file, readable by WISE. After this, the job.fgmj file used by WISE to define starting parameters, simulation start and end times and output files is edited to match the run data and area. Finally the script runs the WISE model, using the climate and auxiliary data as inputs. Model output is written by default as ASCII or tif files, which are then read back and written as NetCDF files. Outputs are written to the output file directory bound to the container at run start. As the runs are daily and independent of each other, post-processing steps are done afterwards.