Representing uncertainty of area class maps with a correlated inter-map cell swapping heuristic

Abstract

Decision makers are often forced to rely on maps too imprecise for their GIS applications. Monte Carlo simulation, a technique that generates many versions of possible application results, is one method for representing uncertainty for applications using overly generalized maps. For the purposes of this paper, an overly-generalized map contains an area class map too coarse to be useful for a particular application. A conflation technique combining the generalized map and samples of application quality data defines a probability distribution of map classes and the spatial statistics desired in the set of Monte Carlo input maps. This paper introduces a combination of model components to better construct area class maps for Monte Carlo simulation, including a correlated inter-map cell swapping algorithm, a class probability model, and a new spatial statistic. Together, these components allow for the generation of area class maps for the purpose of representing spatial application uncertainty. This paper focuses its discussion on the cell swapping heuristic. This heuristic allows for a modular approach to simulation modeling, where each model component can be chosen for its particular benefits. A modular approach potentially allows spatial data uncertainty models to be more easily constructed.