Perspectives on LANDFIRE Prototype Project accuracy assessment

The purpose of this chapter is to provide a general overview of the many aspects of accuracy assessment pertinent to the Landscape Fire and Resource Management Planning Tools Prototype Project (LANDFIRE Prototype Project). The LANDFIRE Prototype formed a large and complex research and development project with many broad-scale data sets and products developed throughout its various stages. The scope of the project was defined as mapping and modeling vegetation, wildland fuel, and fire regime characteristics (Rollins and others, Ch. 2). Because of the breadth of the investigation, it is important to base our expectations for accuracy on a clear understanding of the intricacies, interdependencies, and scope of mapping and modeling LANDFIRE products. Our goals in this chapter are to: 1) provide relevant background information regarding accuracies and what was realistically achievable in the LANDFIRE Prototype, 2) provide background regarding our strategies for LANDFIRE National, 3) describe our actual LANDFIRE Prototype accuracy results in broad terms, and 4) provide recommendations for the national implementation of LANDFIRE. This chapter is not intended to provide an exhaustive list and description of all of the various accuracy-related issues and conclusions resulting from the LANDFIRE Prototype (for specific details, the reader will be referred to the appropriate chapters). Rather, this chapter is intended to be broad in scope and to place the many accuracy components within the context of the LANDFIRE Prototype and LANDFIRE National projects. Please note that Lunetta and Lyon (2004) provide an in-depth discussion of the current state of accuracy assessment within the science community.