A qualitative spatial model of hardwood rangeland state- and-transition dynamics

Abstract

We present a method for computerizing the transition rules of a state-and-transition model and then linking this model to a geographic information system. The resulting simulation characterizes rangeland vegetation dynamics in space and time. The method makes use of an expert system, a computer program that forms logical chains of transition rules. Simulation using state-and-transition rules, sometimes called qualitative simulation, has the disadvantage that it is less precise than traditional numerical simulation. However, it may have the advantage of being able to generate more robust simulation of complex vegetation communities. We demonstrate the application of the method by constructing a model of hardwood rangeland in the western foothills of the Sierra Nevada. The model is tested by comparison with historic black-and-white aerial photographs. The model is found to agree generally with the observed data but to differ substantially in some locations. Implications of this difference are discussed.