Overview

FARSITE-based wildfire spread prediction with directional ROS adjustment factors

The Fire module predicts wildfire spread in FARSITE using landscape, ignition, scenario inputs, and a directional ROS adjustment factor. It is prediction-oriented: the current service does not estimate or update that factor from satellite observations, but uses an already prepared factor to run wildfire spread simulation.

Primary objective

Run a prepared FARSITE case quickly and consistently with directional spread adjustments already defined.

Core engine

The workflow stages landscape, ignition, and scenario data, applies an adjustment factor, and launches the wildfire spread simulation.

Main outputs

Simulated wildfire spread products generated from the packaged FARSITE run, rather than a new estimate of the directional ROS factor itself.

Method

How it works

The current platform implements the prediction half of a larger wildfire-assimilation workflow.

In regional wildfire simulators such as FARSITE, one of the most important quantities controlling fire spread is the rate of spread (ROS). Because the raw ROS predicted from fuel, terrain, and weather data often differs from actual fire behavior, an adjustment factor is introduced to improve simulation accuracy:

$$ROS_i^{adj} = \eta_i \, ROS_i$$

Here, $ROS_i$ is the simulator's base rate of spread for fuel model $i$, and $\eta_i$ is the ROS adjustment factor. Once this factor is specified, FARSITE can generate wildfire perimeters, spread directions, and arrival-time outputs using the adjusted spread behavior.

Directional ROS adjustment factors can be updated in near real time from wildfire detections, but that estimation loop is not what this service runs. Instead, this module starts after those factors are already available and focuses on execution of the wildfire spread prediction itself.

Fire spread is not uniform in every direction. Forward, flank, and backing directions can have very different rates of spread, which is why directional adjustment factors matter.

Directional sectors organize the ROS adjustment factors around the ignition region so that head-fire and flank-fire behavior remain distinguishable during prediction.

The workflow in this platform prepares the FARSITE case, stages the landscape and ignition data, applies the provided ROS adjustment file, and runs wildfire spread simulation. If no custom adjustment file is supplied, a built-in adjustfactor.adj template is used as the fallback.

Important: this module does not derive directional ROS adjustment factors from satellite observations and does not run the near-real-time NNLSE updating procedure. It uses a prepared factor to produce spread predictions.

Inputs

What you need

Package the wildfire case as a single ZIP archive that contains the files FARSITE needs to run.

The Fire module requires one .zip file containing a complete FARSITE case package.
You need at least one .lcp landscape file and one .input scenario file.
You also need one ignition shapefile set: .shp, .shx, and .dbf.
An optional barrier shapefile set can be included if the case uses barriers.
An optional .adj file can be included for the directional ROS adjustment factor. If it is omitted, the module uses the built-in adjustfactor.adj.
Projection files such as .prj are optional but recommended for map alignment and result visualization.

In short: provide a prepared wildfire simulation case and, optionally, a pre-estimated directional ROS adjustment factor. The module uses that factor to run spread prediction in FARSITE.