Year of Award


Document Type


Degree Type

Master of Science (MS)

Degree Name

Geography (Cartography and GIS Option)

Department or School/College

Department of Geography

Committee Chair

Anna E. Klene

Commitee Members

Christiane von Reichert, Tyron Venn


data uncertainty, error, risk assessment, scale, wildfire risk


University of Montana


Widespread use of geospatial data in environmental decision-making tools such as wildfire risk models has called attention to questions of availability, quality, and currency of input data layers. As wildfires are modeled with growing confidence and knowledge of how resources respond to fire is increasing, challenges must be addressed before geospatial data are acquired and used to represent resources of high value in wildfire risk assessments. Researchers at the Rocky Mountain Research Station and the Western Wildland Environmental Threat Assessment Center of the USDA Forest Service employ a framework for assessing wildfire risk to a range of human and ecological resources important in wildland fire management. This framework links spatially explicit fire behavior with potential fire effects and has been demonstrated to be scalable from national to project levels. Spatially identified resource “values” data are a necessary component to defining wildfire risk, and these data serve as baseline information useful in monitoring wildfire risk to resources of high value, as requested by various federal oversight agencies. Resources such as wildland-urban interface, critical habitat for plant and animal species, recreation infrastructure, and restoration of fire-adapted landscapes are important considerations in examining wildfire risk. A comparison study of “relative risk to resources” mapped at the national extent versus at the Deschutes National Forest extent provides a platform by which to discuss national data challenges of: (1) acquiring spatially explicit values data; (2) managing uncertainty surrounding these data; and (3) how use of these data for national assessments may alter or bias results. Relative patterns of wildfire risk to resources are demonstrated by plotting likelihood of burning against average simulated flame lengths for all pixels coincident with mapped values. Recommendations for describing spatial data uncertainty vary according to data type and associated metadata accounting for known errors. This research demonstrates a novel approach to exploring data uncertainties by comparing data developed for wildfire risk assessments at two different spatial scales.



© Copyright 2010 Julie Marie Gilbertson-Day