Generalizing surficial geological maps for scale change: ArcGIS tools vs. cellular automata model

Map generalization is rapidly becoming an important issue in surficial geology. One of the most complicated and least defined steps in this procedure is polygon generalization. Here we outline and compare two different approaches to a particular generalization problem, combining four large-scale maps of unconsolidated material from the Chibougamau region, Québec, Canada into a single map for the entire area. First we provide a working multi-step recipe for the ArcGIS software and then show how this procedure can be automated and outperformed through using the cellular automata (CA) model previously proposed for treating satellite images. We finally chose the CA as the preferred method as its results better meet the requirements imposed on the final product. Along the way, we also demonstrate a few ways to control the final level of generalization and show how large maps can be processed on ordinary computer equipment.