Modeling land cover change under conditions of multi-scaled spatial data: An application of landscape ecology in environmental planning

Spatial models of land cover change have proven to be important tools in regional environmental planning. These models provide insight into the magnitude and pattern of change in the landscape. By integrating remotely sensed satellite data of the earth's surface with analysis tools from a Geographic Information System (GIS), spatial models can provide a quantitative description of the changing landscape. However, such descriptions have traditionally been limited to a simple transition matrix and summary statistics of net change. The focus of this research is to expand landscape change analysis for environmental planning so that it addresses issues of multiple spatial scales, description of landscape structure, and the stochastic nature of land cover change. Within the framework of a GIS, the structural components of landscape elements are defined at various scales, and the relationship between indices of landscape structure and estimates of the probabilistic behavior of those elements is examined. A method for modeling the spatial structural of the landscape from classified remotely sensed data is combined with current theories of landscape ecology and a conceptual framework for modeling probabilistic landscape change. This approach provides a means to linking the probabilistic behavior of fine scale spatial dynamics to the patterns observed in the landscape. It is useful both for the development and refinement of new and broader theoretical models and for applied regional environmental planners who need accurate and pragmatic tools.