A GIS-Based Sensitivity Analysis of Multi-Criteria Weights

Sensitivity analysis (SA) is a prerequisite for model building since it determines the reliability of the model through assessment of uncertainties in the simulation results. With growing interest in extending GIS to support multi-criteria decision-making (MCDM) methods, enhancing GIS-based MCDM with sensitivity analysis procedures is crucial. SA should be involved in GIS-MCDM model evaluation that tests the robustness of a model and the extent of output variation when parameters are systematically varied over a range of interest. The most common approach is based on varying criteria or their weights which represents input parameters in order to understand the model behavior and its limitations. This paper presents a novel approach of examining multi-criteria weight sensitivity of a GIS-based MCDM model. The objectives of this study are to explore the dependency of model output on input parameters, identifying criteria that are especially sensitive to weight changes and to show the impacts of changing criteria weights on the model outcomes in spatial dimension. A methodology was developed to perform simulations where the decision weights associated with all criteria used for suitability modelling were varied to investigate their relative impacts on the final results of the evaluation. A tool which incorporates SA with Analytical Hierarchy Process (AHP) within the ArcGIS environment was implemented. It permits a range of user defined simulations to be performed to quantitatively evaluate model dynamic changes, measures the stability of results with respect to the variation of different parameter weights, and displays spatial change dynamics. A comprehensive case study of irrigated cropland suitability assessment addressing the application of the new GIS-based AHP-SA tool is described. After an initial evaluation which is assumed to be the base run according to best available knowledge, the original weights for the five different criteria from the base run were varied within a range of 20% provided that a 1% step was used for simulations. Summary tables for 40 simulation runs of each criterion and the resultant evaluation maps were generated and displayed. The tables demonstrate the change in the output suitability classification relative to changes in input criteria weights, identify criteria that are especially sensitive to weight variations, and visualise the spatial dimension of change dynamics. The results indicate that the evaluation map derived from base run could be used for analysing potential expansion of irrigated cropland because, when water is available, the highest possibility of such expansion will occur on highly suitable and moderately suitable lands which are relatively stable from the SA investigation in this study. The integration of SA with AHP using GIS has enhanced the conventional AHP module, improved the reliability of MCDM output, and extended existing GIS functionalities. This GIS-based AHP-SA tool supplies more immediate feedback to evaluators/modellers. It is easier for non-experts to understand, and provides a mechanism to explore the decision problem while learning how changes in criteria weights affect evaluation outcomes spatially and quantitatively. It enables decision makers to follow/conduct a comprehensive yet easy-to-use procedure to examine weight sensitivity in both criteria and geographic space. It can give better insights for improving the capabilities of current GIS-based AHP-MCDM models to create more realistic output scenarios. Continued advances in this research area will allow GIS and MCDM to be applied to practical land-management issues with greater success.