FireBuster: a tool for fire management
| Authors: | John W. Benoit, Shyh-Chin Chen |
| Year: | 2019 |
| Type: | General Technical Report |
| Station: | Pacific Southwest Research Station |
| Source: | Gen. Tech. Rep. PSW-GTR-261 (English). Albany, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station: 25-37 |
Abstract
We developed an experimental high-resolution fire weather forecast system called FireBuster to help fire management in California. This system streamlines and automates many processes required to deliver timely fire weather intelligence. FireBuster provides routine twice-daily 72-hour weather forecasts in real-time for California at a 5-km grid resolution. Authorized users can select part of the domain and request a 1-km resolution 72-hour forecast with only a few clicks.
Forecast outputs include near surface values of temperature, relative humidity, wind speed and direction, precipitation, and several specialized variables. Each forecast run (5-km and 1-km resolution) takes from 1 to 2 hours to process on our in-house computing hardware.FireBuster delivers information, as each 6-hour increment completes, via a web page which visually depicts the forecast over an interactive map. The user can view future weather conditions hour-by-hour over the entire domain. When available, weather data from the MesoWest observational network can be displayed for post-evaluation. Additional features are in development, such as a format of the gridded output that can be downloaded for input into FARSITE, a popular fire spread model.
The Southern California Geographic Area Coordination Center (GACC), which manages all regional fire-fighting resources, is the intended user of this system. Collaboration with the GACC’s fire meteorologists has provided us ongoing feedback on how to best improve FireBuster. The next development phase of the system includes adding forecasted fire danger indices, observed fire perimeters and an economic data layer so that FireBuster can truly be part of an integrated fire management tool.
Also, we developed a related parallel system, called FireBusterSim, to examine the downscale simulation. FireBusterSim works similar to FireBuster, but produces downscaled weather given archived global analysis data. Both 5- and 1-km grid cell resolution model runs are made over the area of the fire. Increased spatial resolution is shown to indeed provide better accuracy in model data. Incorporating finer topographic details into the model improves weather prediction in complex terrain where fires often occur.