GIS Uncertainty Modeling Using a Probabilistic Polygon Intersection Method
Author(s): David Geoffrey Goode
Assimilating map data as polygons and finding relationships among them using intersection is a common GIS analysis technique. Normally, polygon and their values are assumed to be nonvarying; however, in practice they often do vary. In such cases, defining confidence intervals about polygon intersections may be useful. This paper presents a polygon intersection method that accounts for this uncertainty by allowing polygon size and value to vary. Uncertainties are obtained from known distributions or by bootstrapping. Confidence intervals are generated using a Monte Carlo algorithm. A forestry economical analysis and an ecological assessment are used to illustrate the method.
David Geoffrey Goode
15375 SE 30th Place
Bellevue, WA 98007