Quantification of uncertainty in spatial data for ecological applications

Land cover maps are a primary data source for persons designing reserves, working on management plans for sensitive species, and performing ecological risk assessments. Maps are generalizations of reality and contain spatial uncertainty from the generalization which should be quantified for ecological applications. We propose to develop a spatial statistical approach that can deal comprehensively with the problem of error propagation and provide a bridge between the distinct traditions of remote sensing and Geographic Information Systems. One or more case studies illustrating the approach would be undertaken by an interdisciplinary team, and a specialists meeting would provide input during the first year of work.