Validation of spatial multicriteria decision analysis results using public participation GIS

Abstract

Spatial or GIS-based Multicriteria Decision Analysis (GIS-MCDA) incorporates a mixed set of theories, methods and tools in order to combine geospatial data with decision makers' judgments for the purpose of generating information required in spatial decision making. However, given the complex nature of spatial decision making processes, contradicting preferences of experts regarding the proper evaluation criteria to be used as well as uncertainty of inputs, the GIS-MCDA results may not be perfectly matched with the community preferences. Consequently, ensuring that the outcomes of decision-making processes are indeed consistent with the needs, values and preferences of the public is of critical importance. Accordingly, this study proceeds to employ Public Participation GIS (PPGIS) or citizen-generated information as reference data to validate the results of GIS-MCDA processes. This validation is carried out by means of certain spatial indices including total coverage, geometric intersection, central distances and statistical indices. To this end, the issue of selecting the optimal sites for the establishment of restaurants in Babolsar, Mazandaran province, was selected as the case study. The study proceeded by initially determining the optimal locations of restaurants based on expert opinion and via the GIS-based Best-Worst multicriteria method. Next, a Web-based PPGIS tool was designed and implemented, enabling citizens to draw the optimal sites for restaurant establishment. Finally, the optimal locations determined via the multicriteria method (in three suitable, moderate and unsuitable classes) were compared and validated with PPGIS in accordance with the corresponding spatial indices. Although the suitable class of the GIS-MCDA output map (expert-derived restaurant sites), indicating preferences of experts, had the highest level of complete coverage (0.1%) and intersection (2.6%) with PPGIS (citizen-derived restaurant sites) as opposed to the other two classes, the validation results indicate a rather low level of consistency between experts’ solution and citizen opinions in the designation of proper sites for the establishment of restaurants.