# How to diagnose fuel properties in WRF-Fire

*Back to the WRF-SFIRE user guide*.

The fuel properties in WRF-Fire are given in file namelist.fire. These serve to create coefficients in the Rothermel's formula. To aid diagnostics it is useful to graph the resulting fire spread rate as a function of wind and slope. The graphs are available in metric units, as well as in English units for direct comparison with the published spread rate graphs for the Scott-Burgan fuel categories.

## Contents

## Prerequisites

## Location of the codes

- The fuel properties of every category of the fuel are defined in file
**namelist.fire**, which needs to be in the current directory. - The fire rate of spread computation is done in file
**WRFV3/phys/module_fr_sfire_phys.F**split between two subroutines. First, the coefficients at every point are precomputed from fuel properties in subroutine**set_fire_params**at initialization. In every time step, the spread rate at a point is then computed from those coefficients in subroutine**fire_ros**. - The computation has been duplicated in Matlab for diagnostic purposes in file
**other/Matlab/vis3d/fire_ros.m** - The plot routine to create graphs like shown here is
**other/Matlab/vis3d/plot_fuel.m**

## Heat flux diagnostics

**plot_fuel**also computes and prints several diagnostic quantities derived from fuel properties, including total heat density generated by the combustion of the fuel over time (J/m^{2}), and the maximal heat flux density (W/m^{2}) from the initial slope of the fuel loss curve. These can be compared with the statistic printed by WRF-Fire while running (maximal heat flux density, total heat flux of the fire, total heat generated by the fire).

## Step-by-step instructions

- Run WRF-Fire, you can kill it right after the the first time step. There will be file
**fuels.m**created in the current directory. - Start Matlab in the directory
**test/em_fire**to set up the search path properly. - In Matlab, navigate to the directory with the
**fuels.m**file - Type
**fuels**. This will create variable**fuel**. Say you want to check fuel 3. Type**fuel(3)**to see what is there. - Type
**plot_fuel(fuel(3))**to create the graph below in metric units, and**plot_fuel(fuel(3),'sb')**for the same units as in Scott and Burgan (2005). Type**help plot_fuel**for further information. Note: The graph data were created in WRF-Fire and stored in the**fuel**variable. - Type
**edit fire_ros**to see the fire spread rate calculation recreated in Matlab. - Type
**big(check_ros(fuel))**to make sure the calculation in Matlab is the same as in WRF-Fire, it should return rounding error only (less than 1e-4)

## Diagnostics provided

The **plot_fuel** command displays the graph of rate of spread as a function of wind speed, slope, and fuel moisture.

## Works with

- Version 16 Sep 2010
- Version 29 Aug 2012 and hopefully later
- Matlab R2010a, 2012a, and hopefully later

## References

- Anderson, H. E., 1982.
*Aids to determining fuel models for estimating fire behavior.*U.S. Forest Service General Technical Report INT-122. Ogden, UT. pdf (The original 13 fuel models) - Baughman, R. G. and F. A. Albini.,1980.
*Estimating midflame windspeeds.*In Sixth Conference on Fire and Forest Meteorology, Seattle, WA (Society of American Foresters) pdf (Windspeed reduction from 20ft) - Rothermel, R.C., 1972.
*A mathematical model for predicting fire spread in wildland fuels.*Res. Pap. INT-115. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station. 40 pp. pdf (The original for reference) - Rothermel, R. C., 1983.
*How to predict the spread and intensity of forest and range fires.*U.S. Forest Service General Technical Report INT-143. Ogden, UT. pdf (Updated version) - Scott, J. H. and Burgan, R. E., 2005.
*Standard fire behavior fuel models: a comprehensive set for use with Rothermel's surface fire spread model.*Gen. Tech. Rep. RMRS-GTR-153. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, 72 pp. pdf (More detailed fuel models, with rate of spread curves)