Keywords
biogeography, bayesian statistical modelling, gis, elicitation, mixture models, clustering
Start Date
1-7-2004 12:00 AM
Abstract
Ecological regions are increasingly used as a spatial unit for planning and environmentalmanagement. It is important to define these regions in a scientifically defensible way to justify any decisionsmade on the basis that they are representative of broad environmental assets. The paper describes amethodology and tool to identify cohesive bioregions. The methodology applies an elicitation process toobtain geographical descriptions for bioregions, each of these is transformed into a Normal density estimateon environmental variables within that region. This prior information is balanced with data classification ofenvironmental datasets using a Bayesian statistical modelling approach to objectively map ecological regions.The method is called model-based clustering as it fits a Normal mixture model to the clusters associated withregions, and it addresses issues of uncertainty in environmental datasets due to overlapping clusters.
Ecoregion Classification Using a Bayesian Approach and Model-based Clustering
Ecological regions are increasingly used as a spatial unit for planning and environmentalmanagement. It is important to define these regions in a scientifically defensible way to justify any decisionsmade on the basis that they are representative of broad environmental assets. The paper describes amethodology and tool to identify cohesive bioregions. The methodology applies an elicitation process toobtain geographical descriptions for bioregions, each of these is transformed into a Normal density estimateon environmental variables within that region. This prior information is balanced with data classification ofenvironmental datasets using a Bayesian statistical modelling approach to objectively map ecological regions.The method is called model-based clustering as it fits a Normal mixture model to the clusters associated withregions, and it addresses issues of uncertainty in environmental datasets due to overlapping clusters.
