Improved models of forest and wildland-urban-interface fires

Revolutionary advances in computational power at ever lower costs and vastly improved computational algorithms, have changed in the past few decades the nature of modeling in science and engineering. These improvements have opened the door to development of a new generation of models for predicting the behavior of wildland and wildland urban interface (WUI) fires. These next generation, research models are fundamentally different from current operational models because they are physics based, fully three-dimensional, and solve the conservation equations (mass, momentum, energy, and species). While these models are both computationally and data intensive, they describe fire from first principles and promise to be better able to predict fire behavior under realistic 3D conditions of variable fuels, weather, and topography.

In contrast, current operational models for predicting wildland-fire behavior are based on concepts developed by Rothermel and others in excess of thirty years ago. They deal only with two-dimensional fuel distributions and attempt to predict fire-line behavior in wildland fuels using semi empirical relations between model variables. While these operational models, and other similar ones, have been very useful for land-management agencies in the past, they are unable to answer more detailed questions about fire behavior of concern today.

In this paper, a history will be given of the research, including experiments, utilized to develop, calibrate, and validate the current, highly successful mathematical model known as the Fire Dynamics Simulator (FDS). This model, which is well documented and has become very widely used by fire protection engineers around the world, predicts fire growth and spread in buildings. This history suggests that some concepts and relations developed for the Rothermel model may now be limiting the ideas needed to develop models of wildland and WUI fire of the type discussed above!

In April 2005, the Government Accountability Office (GAO) issued a report entitled, “TECHNOLOGY ASSESSMENT: Protecting Structures and Improving Communications during Wildland Fires.” This report identifies two favored methods to reduce the risk of damage to a structure from wildland fires. First, create and maintain a buffer, called the defensible space, from 30 to 100 feet around the structure, where vegetation and other flammable objects are reduced or eliminated. Second, during construction, use fire-resistant roofs and vents, and, perhaps, fire-resistant windows and building materials. Land-management agencies and other organizations are encouraged to adopt these suggestions and also to educate property owners to adopt them in order to reduce the shared risk.

Unfortunately, however, the basic physics of the fire growth and spread in wildland, as well as, in WUI areas is not understood, and, therefore, tools to evaluate fire threat and risk are not available! Research is needed to gain this understanding and to develop mathematical models with predictive capability for wildland and WUI fire spread. Without understanding and such predictive models for wildland and WUI fire behavior, quantitative evaluation cannot be carried out of defensible-space guidelines or of risk reduction from more fire-resistant building materials.