Four Fundamental Questions to Evaluate Land Change Models with an Illustration of a Cellular Automata–Markov Model

Abstract

Numerous models exist for users to simulate land change to communicate with an audience concerning future land change. This article raises four fundamental questions to help model users decide whether to use any model: (1) Can the user understand the model? (2) Can the audience understand the model? (3) Can the user control the model? (4) Does the model address the goals of the specific application? This article applies these questions to the popular cellular automata–Markov (CA–Markov) model as IDRISI’s CA–Markov module expresses. Sensitivity analysis examines 120 ways to set the module’s parameters. Verification compares the module’s behavior to the software’s documentation. Results show that the cellular automata’s allocation fails to follow the quantity of change that the Markov module computes. The module’s behavior is likely to cause users to misinterpret the validation metrics and to miscommunicate with audiences. Thus, the answers to the four questions were not satisfactory for this article’s case study. This article’s framework helps users to judge a model’s appropriateness for a specific application by combining sensitivity analysis with verification in a manner that helps to interpret validation. Users should answer the four questions as they decide whether to use any software’s modules.