2. Configuring the Community Fire Behavior Module¶
2.1. Configuring a domain with the WRF Pre-processing System (WPS)¶
Because the CFBM was originally developed as part of the WRF model, creating a domain must be done using the WRF Pre-processing System (WPS). These instructions can be found in the WRF Users Guide. To run the CFBM with the UFS or with WRF data, users need to provide a geo_em.d01.nc file containting the interpolated static data fields.
Future releases will include a method for creating domains without needing to compile WPS.
2.2. Namelist Configuration¶
The options specific to the CFBM are controlled by a namelist file namelist.fire. This namelist file consists of three sections: &time, &atm, and &fire. The available options in each section are described below.
Example namelists can be found in the various test subdirectories under the tests/ directory.
2.2.1. &time¶
start_year: integer (Required)Start year of the simulation.
start_month: integer (Required)Start month of the simulation.
start_day: integer (Required)Start day of the simulation.
start_hour: integer (Required)Start hour of the simulation.
start_minute: integer (Required)Start minute of the simulation.
start_second: integer (Required)Start second of the simulation.
end_year: integer (Required)End year of the simulation.
end_month: integer (Required)End month of the simulation.
end_day: integer (Required)End day of the simulation.
end_hour: integer (Required)End hour of the simulation.
end_minute: integer (Required)End minute of the simulation.
end_second: integer (Required)End second of the simulation.
dt: real (Default:2.0)Atmospheric time step in seconds.
interval_output: integer (Required)[Units: s] Specifies the time interval (in seconds) for writing to the history output files
num_tiles: integer (Default:1)Number of tiles for MPI domain decomposition. Not yet implemented.
2.2.2. &atm¶
kde: integer (Default:2)Number of vertical levels for the atmospheric simulation
interval_atm: integer (Default:0)[Units: s] Time interval (in seconds) for incoming atmospheric data. When running a coupled model, this value represents the atmospheric timestep. In offline mode, it determines the frequency of reading atmospheric data from the input file.
2.2.3. &fire¶
fire_print_msg: integer (Default:0)- Debug print level for the fire module.
Levels greater than 0 will print extra messages at run time.
0: no extra prints
1: Extra prints
2: More extra prints
3: Even more extra prints
fire_atm_feedback: real (Default:1.0)- Multiplier for heat fluxes from the fire to the atmosphere.
0.0: one-way (atmosphere –> fire) coupling.
1.0: normal two-way coupling.
Intermediate values will vary the amount of forcing provided from the fire to the dynamical core.
fire_upwinding: integer (Default:9)- This option controls the type of upwinding scheme used for calculating the normal spread of the fire front. The choice of upwinding scheme significantly impacts the accuracy of fire spread simulations. Higher-order schemes, like WENO3 and WENO5, generally offer better accuracy but can be more computationally expensive.
0: Central Difference: Uses central differences for calculating gradients, combining left- and right-sided differences for both x- and y-directions to compute a central gradient approximation.
1: Standard: Employs an upwind scheme, selecting between left- and right-sided differences based on flow direction.
2: Godunov: The Godunov scheme is a first-order upwind scheme based on Osher & Fedkiw
3: ENO1: The First-Order Essentially Non-Oscillatory (ENO1) scheme uses the smoothest stencil to avoid sharp gradients, which can lead to underestimations of fire area and errors in the rate of spread.
4: Sethian scheme [Sethian, 1996]
5: 2nd-order: Calculates gradients using a second-order central difference.
6: WENO3: Third-Order Weighted Essentially Non-Oscillatory (WENO3) scheme.
7: WENO5: Fifth-Order Weighted Essentially Non-Oscillatory (WENO5) scheme.
8: Hybrid WENO3/ENO1: A hybrid scheme that combines WENO3 in a band surrounding the fire front interface with ENO1 for regions further away. This approach reduces computational cost while maintaining accuracy near the front.
9: Hybrid WENO5/ENO1 (default): Similar to option 8, but uses WENO5 instead of WENO3 in the band surrounding the fire front. This approach reduces computational cost while maintaining accuracy near the front.
fire_viscosity: real (Default:0.4)Artificial viscocity in level-set method away from the near-front region.
fire_lsm_reinit: logical (Default:.true.)Flag to activate reinitialization of the level-set method
fire_lsm_reinit_iter: integer (Default:1)Number of iterations for reinitialization PDE
fire_upwinding_reinit: integer (Default:4)- Numerical scheme (space) for reinitialization PDE.
1: WENO3
2: WENO5
3: hybrid WENO3-ENO1
4: hybrid WENO5-ENO1
fire_lsm_band_ngp: integer (Default:4)When using
fire_upwinding_reinit=3,4andfire_upwinding=8/9, the number of grid points around lfn=0 that WENO5/3 is usedfire_lsm_zcoupling: logical (Default:1)When true, uses
fire_lsm_zcoupling_refinstead offire_wind_heightas a reference height to calculate the logarithmic surface layer wind profilefire_lsm_zcoupling_ref: real (Default:50.0)[Units: m] Reference height from which the velocity at
fire_wind_heightis calculated using a logarithmic profilefire_viscosity_bg: real (Default:0.4)Artificial viscosity in the near-front region
fire_viscosity_band: real (Default:0.5)Number of times the hybrid advection band to transition from
fire_viscosity_bgtofire_viscosityfire_viscosity_ngp: integer (Default:2)Number of grid points around lfn=0 where
fire_viscosity_bgis usedfmoist_run: logical (Default:.false.)Runs moisture model on the atmospheric grid, outputting the result as a variable named
fmc_gcfmoist_freq: integer (Default:0)- Frequency to run moisture model.
0: use
fmoist_dtk>0: every “k” timesteps
fmoist_dt: real (Default:600.0)[Units: s] Time step of moisture model (only used if
fmoist_freq=0)fire_wind_height: integer (Default:6.096)[Units: m] Height of uah,vah wind in fire spread formula
fire_is_real_perim: logical (Default:.false.)- Determines if perimeter represents a real fire boundary.
.true. = observed perimeter
.false. = point/line ignition
frac_fburnt_to_smoke: real (Default:0.02)[Units: g/kg] Parts per unit of burned fuel converted to smoke, represented as grams of smoke per kilogram of air.
fuelmc_g: real (Default:0.08)Fuel moisture content ground (Dead FMC)
fuelmc_g_live: real (Default:0.30)Fuel moisture content ground (Live FMC). 30% Completely cured, treat as dead fuel
fuelmc_c: real (Default:1.00)Fuel moisture content of the canopy
fuel_opt: integer (Default:1)- Fuel type model.
1: Anderson fuel model (only option currently implemented)
ros_opt: integer (Default:0)- Rate of Spread (ROS) parameterization option.
0: Rothermel model (only option currently implemented)
fmc_opt: integer (Default:1)- FMC model
-1 = Constant fuel moisture (only option currently implemented)
fire_num_ignitions: integer (Default:1)Number of ignitions for fire initiation. Maximum of 5.
Note
For each additional fire ignition, you must specify an additional set of ignition parameters below, with increasing numerical suffixes ( e.g. fire_ignition_start_lon2, fire_ignition_start_lon3, etc. )
fire_ignition_start_lon1: real (Default:0.0)Longitude of first ignition start point.
fire_ignition_start_lat1: real (Default:0.0)Latitude of first ignition start point.
fire_ignition_end_lon1: real (Default:0.0)Longitude of first ignition end point.
fire_ignition_end_lat1: real (Default:0.0)Latitude of first ignition end point.
fire_ignition_ros1: real (Default:0.01)[Units: m/s] Rate of spread for first ignition (Rothermel parameterization).
fire_ignition_start_time1: real (Default:0.0)[Units: s] Start time of first ignition in seconds (counting from the beginning of the simulation)
fire_ignition_end_time1: real (Default:1)[Units: s] End time of first ignition in seconds (counting from the beginning of the simulation)
fire_ignition_radius1: real (Default:0.0)[Units: m] Radius of the ignition area for first ignition.
