Spatial Decision Making: Using a Geographic Information System and the Analytic Hierarchy Process for Pre-Wildfire Management

Abstract

Strategic management of wildlands for fire is increasingly a mix of traditional firescience, climatology and human perceptions. Not only must managers be expert atmodeling fuels and fire behavior, they must also understand human behavior, and theeffects of climate on landscapes. We hypothiszed that areas in national forests differspatially in their importance to stakeholders, including both the public and landmanagers. That this difference is based upon the inclusion of factors not typically foundin wildland fire models. To test this hypothesis we used a multidimensional approach toassess the spatial variability several factors including recreation, property values and fuelmoisture. This approach combined a geographic information system with the analytichierarchy process to predict and test the current distribution of areas in national forestsimportant to stakeholders.Inclusion of stakeholders appears to improve the validity and useability of aspatial decision support system. Comparing the model created in this dissertation withseveral others demonstrates that it is important to strike the right balance betweenstakeholders and technical experts when designing and creating a model. It is alwaysbeneficial, however, to a significant level of stakeholder involvement.Areas important for fire mitigation efforts depended on the stakeholder oraudience rating the model. Raters from the U.S. Forest Service tended to favor areas withhigh fire probability scores, while those from the Park Service prefered recreation areasand places people value. In both cases, locations people had easy access to, such as alongroads and trails were favored.These results confirmed the hypothesis that areas of importance are differentbased on the individual rating the model. Further testing and refinement of the modelincludes expanding the study area beyond the southwestern United States as wells asobtaining better sources of data with finer spatial resolutions.