U.S. Forest Service BRASS
The U.S. Department of Agriculture Forest Service is a Federal agency that manages public lands in national forests and grasslands. Administering 193 million acres of land, an area equivalent to the size of Texas, the US Forest Service is divided into 9 regions, encompassing 155 National Forests and 20 National Grasslands. The natural resources on these lands are some of the Nation's greatest assets and have major economic, environmental, and social significance for all Americans.
The mission of the USDA Forest Service is to sustain the health, diversity, and productivity of the Nation's forests and grasslands to meet the needs of present and future generations. In keeping with US Forest Service guiding principles of using an ecological approach to multiple-use management, using the best scientific knowledge in making decisions, and selecting the most appropriate technologies in the management of resources, the BRASS (Burning Risk Advisory Support System) decision support tool provides a continuous means for forest and grassland managers to assess vegetation and weather to support decisions related to prescribed burning and/or the risk of wildfire.
The objective of the vegetation and fire monitoring system is to inventory, monitor, evaluate, and integrate land condition trends and capabilities with Forest Service management and public use goals to enhance, improve, repair, and sustain national forests and grasslands. Texas Agrilife Research has a continuing agreement with the US Forest Service to develop this system using a viable Phytomass Plant Growth model (PHYGROW) and a Burning Risk Advisory Support System (BRASS).
History
Texas Agrilife Research began its involvement with the US Forest Service in 2005 with a contract through the USDA Risk Management Agency (RMA) to develop BRASS for the Lincoln National Forest in New Mexico. An extension was received which also included the Prescott National Forest, and later in 2008, the Coconino National Forest in Arizona.
We are currently working on developing the BRASS model for other forests including the Kaibab, Carson, and SantaFe, as well as creating a new scalable technology stack of the automation system that we can tranfer to the Forest Service so that they can model their forests independent of CNRIT. Though our models and GIS applications will remain in C and C++, our handlers, web services, and other middleware applications are all being converted to PHP 5 and javascript for easy to maintain open source functionality.
Data Collection
Landscape Modeling
The BRASS decision support tool provides a continuous means for the US Forest Service resource managers to assess vegetation and weather to support decisions related to prescribed burning and/or the risk of wildfire. The BRASS system is composed of two main components, the PHYGROW growth model and the PHYRESIM burning model. PHYRESIM was developed from a software toolkit called Firelib, which is the same toolkit that drives the highly respected BEHAVE burning application. Firelib was developed by the US Forest Service as a toolkit to build custom BEHAVE type applications.
The PHYGROW model is a near-real time plant growth model that is updated daily utilizing current and forecasted weather conditions from the National Oceanic and Atmospheric Administration (NOAA). A PHYGROW model has been calibrated for each of the major plant communities and ecological sites within the base, which will continuously monitor vegetation production and fuel load conditions.
In order to distribute the modeled point data across the landscape, a methodology developed by the US Forest Service has been implemented called Most Similar Neighbor (MSN). First, a landscape map of plant communities is developed within a Geographic Information System (GIS) using available resources such as ecological site maps derived by the Natural Resource Conservation Service (NRCS), plant communities derived from classification of remotely sensed satellite imagery, and supplemental field collected data. Advanced image processing software (i.e. ERDAS, ENVI, and IDRISI) has facilitated the development of plant community polygons from multispectral satellite imagery. Next, the necessary PHYGROW and BRASS field sample data is collected for a minimum of one polygon within each unique plant community. The field collected dataset is then distributed across the landscape by matching similar non-sampled plant community polygons as determined by the MSN analysis with field sample data collected within the sampled polygons.
The PHYGROW output is integrated with the fire behavior model, PHYRESIM, to provide a continuously updated fire risk map for an area. PHYGROW outputs current live herbaceous moisture, live herbaceous production, 1-hr. fuel accumulation, live wood moisture, and live wood production to the PHYRESIM subsystem on a daily basis. PHYRESIM coordinates the fuel moisture stick model and PHYGROW outputs with NOAA current and forecasted weather data to produce a 7 day forecast updated at 6 hour intervals. Changing weather conditions and fluctuating plant communities create dynamic BRASS 30-minute burn area, flame length, spread rate, and fuel moisture outputs. This data can be used to select areas beforehand with adequate fuel-load and appropriate weather conditions for a prescribed burn, as well as, determine wildfire risk conditions.
Product Delivery
The final delivery for these multi-forest projects is the BRASS (Burning Risk Advisory Support System) software and configuation database. Fire conditions can be assessed for any point on a forest via the internet to assist controlled burn crews, fire fighters, and other groups associated with fire management in assessing conditions in the field. Additional range information such as vegetation production, drought prediction, and historical ranking is also delivered through the internet. The US Forest Service will establish it's own data center for running the BRASS software and will begin using their own system by the end of 2012.
Publications
Rhodes, E.C., D. Tolleson, W. Shaw, E. Twombly, J. Kava, and T. Brown. 2009. Comparing herbaceous vegetation sampling methods on the Coconino National Forest, AZ, USA. Society for Range Management 62nd Annual Meeting, Albuquerque, NM.