Summary

This lesson revisits "spatial overlay" and "land use allocation" and discusses these two key concepts of spatial analysis in more depth. First, spatial overlay is extended by fuzzy logic, complementing Boolean overlay ("true" or "false") by concepts partial truth. Fuzzy overlay allows vagueness and uncertainty because nature often can't be modeled by crisp points, lines and polygons having crisp attribute classes. The fuzzy membership function (FMF) describes the degree of membership of an entity to a class that does not have sharply defined boundaries (fuzzy set). The degree of membership (d.o.m.) takes values from 1 to 0 with 1 representing complete certainty of membership and 0 representing non membership. For example, crisp slope classes (medium slope is between 20º and 30°) could be replaced by fuzzy slope classes, modeled by FMFs. This would allowing a site with, for example, 26.8° to have only a membership of 0.9 to the fuzzy class "medium slope", but also a small membership of 0.1 to "high slope". Fuzzy overlay then combines fuzzy spatial layers by use of logical combinations of fuzzy sets. The second unit in this lesson is about multi-objective analysis, where GIS adopts a decision support system (DSS) function. The problem at hand is to divide up land into use classes according to its suitability for different objects that may or may not be conflicting. Typical examples include areas suited for industrial development vs. wildlife preservation or industrial development vs. agriculture. Such problems require a multi-objective evaluation (MOE), most commonly performed through the Multi Objective Land Allocation (MOLA) approach. MOLA is a hierarchical extension of multi-criteria evaluation in which potentially conflicting objectives are weighted and thereafter combined: a set of suitability maps, each derived as a single objective, are themselves weighted and combined. A prominent feature of the MOLA approach is that unequal weighting of objectives is possible and very intuitive. Allocation conflicts are solved by allocating the respective cells to the classes they are most suited for (decision heuristic). MOLA is relatively easy to understand and hence well suited for a participatory decision making process.