A GIS-Based Sensitivity Analysis of Multi-Criteria Weights

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