Comparative Study about Performance of Multi-Criteria Analysis in Study of Land Suitability (Case Study: Site Selection of Municipal Waste Landfill in Shiraz City)

Introduction
Using multi-criteria models and techniques in conjunction with GIS capabilities can be considered as one of the most outstanding aspects in application of decision support systems (DSS) in the decision process. Using decision rules, we can classify alternatives according to priority in the process of site selection. There are several decision rules in the field of multi-criteria decision making (MCDM) techniques for determining the appropriateness of land use, such as WLC SAW, AHP, and TOPSIS. Today, it is necessary both to consider the circumstances of carrying out these techniques with principles in the particulate applied fields and to make efforts on the comparative examination on performance of several methods in a particular subject. Continuous consideration to these points can greatly influence the application of multi-criteria analysis techniques and rules in the decision process. Accordingly, in this research, there is an intention to test the operational capabilities of TOPSIS and WLC models as two prominent examples of the multi-criteria analysis techniques in the experimental field of site selection for municipal waste landfill in Shiraz. In the next step, we can provide the pattern for prioritizing the use of these methods in a particular applied field.
 
Methodology
Data and tools that are used in this test are maps and archives of information. The data are collected to define the criteria and constraints that can be applied to determine the desirability of lands in the locating Municipal waste landfill in Shiraz. In this study, some software applications have been employed for data entry, data storage, data management, data processing, data analysis, etc. These are Excel 2007, ArcGIS 9.3, ARCView 3.3 Kilimanjaro IDRISI, and ILWIS 3.3. The main steps in the process of the study as the research methodology are:

Preparing criterion and constraint maps for locating landfills, which led to definition of 26 criteria and 21constraints.
Valuation and standardization of criteria maps: the process of valuation and standardization was performed based on value of membership in fuzzy set.  Standardization was performed using the possibilities that exist in the FUZZY function of IDRISI Kilimanjaro application.
Weighting of criteria maps: in this step we have tried to determine weights of criteria and criterion significance coefficient by using CRITIC method.
Operational use of multi-criteria decision rules and modeling the choices of multi-criteria analysis techniques: in this step, there is the intention to test the operational capabilities of TOPSIS and WLC models as two prominent examples of multi-criteria analysis techniques; in the experimental field of site selection for municipal waste landfill in Shiraz. Thus, we have provided the pattern for prioritizing the use of these methods in a particular applied field. In the framework of presented pattern it is assumed that higher pixels values screening in the output of each model can be viewed as a positive point for applying the model.  

 
Results and discussion
In this research, classified maps that involve suitability of locations for municipal waste landfill sites are obtained by the operational procedures and guidelines that can be considered in the process of using TOPSIS and WLC methods. In the obtained maps, as the score of each pixel value approaches to 250it indicates favorable conditions of that pixel as a suitable site for municipal waste landfill. With the constraints, these pixels can be used proportionally for the time that there are the need for allocation of that area to the land use type in a particular area. Therefore, the obtained maps can be used as guidance by the decision makers in the selection of appropriate locations for municipal waste landfill. Comparison on the rate of screening in the outputs of TOPSIS and WLC models shows that, according to the given utility thresholds in site selection of landfill in study area, the rate of the screening in the output of TOPSIS model is more than that of WLC model. Therefore, the situation in the selection of preferable pixels can reveal that application of TOPSIS model in this particular applied field is recommended.
 
Conclusion
In this paper, with the specified area of the Shiraz as a case study, capabilities and operational mechanisms of TOPSIS and WLC models has been tested in the site selection of municipal waste landfill.  There was the interest in testing the operational capabilities of TOPSIS and WLC models as two prominent examples of the multi-criteria analysis techniques; This was in the experimental field of the site selection for municipal waste landfill in Shiraz. In the next step, we have provided the pattern for prioritizing the use of these methods in a particular applied field. In the framework of presented pattern it is assumed that the higher pixel values screening in the output of each model, as can be investigated in the each level of utility threshold, is considered as a positive point for applying the model at the selected area, with the utility thresholds.
For further documentation of the validity of land use suitability map that has been acquired in the process of using TOPSIS, we have tried to investigate the characteristic of two sample pixels that are selected as preferable and are located on the area with no constraints, and qualified with the defined  criteria. Results of the investigations indicate that the preferable pixels in the output map have the optimal conditions in terms of defined criteria.  For example, one of the pixels that have been selected as a preferable pixel has suitability score more than 250, in the 19 criteria and is located in the acceptable condition in terms of degree of membership in fuzzy function.