Multi-Criteria Evaluation of Land use Effects on Environment using Quantitative Methods in GIS: The Talighan Basin (Iran)
Third, Talighan is only a 60 minute drive from Iran's populous capital, Tehran. Considering its proximity to the Talighan region, many of Tehran's residents fatigued by work and pollution, not only would plan to spend their weekends there, but may even make long range commitments to settle in the region and commute to the capital. All together this is leading to higher land prices in the region which in turn encourages the natives to eventually sell their lands to either speculators or the new settlers. As for now, it is neither politically nor economically prudent for the government to intervene in the market and forbid transactions. Therefore, one can expect an alarming pressure being built up on the environment, leading to haphazard development of land use and a long-term destruction of the region with national significance. To intervene in this destruction process, a method is first needed to seek suitable land for various expected or planned uses. In the following sections, a quantitative method used in the GIS environment is discussed for evaluation of the expected land use effects in the Talighan region.
A Case Made for using GIS Methodology
For the region described above, it is necessary to perform a multi-faceted evaluation of resource capacities. A common method to achieve this evaluation, is the mutli-critera evaluation in which land suitability is examined based on thinking through a multiple of objectives. In this method, overlaying is often used and factors such as slope and elevation are considered but without regards to their continuos nature. Instead, data is categorized as if it were discrete data, based merely on personal judgment. Also, in the overlaying technique, assigning different weights for different variables is a difficult and inefficient task when one intends to take into account the qualitative-ness of issues i.e. whether a use is compatible with other land use or not (Carver, 1991). The users and experts alike, however, prefer the overlaying technique over others and tend to ignore its limitation mostly due to its arithmetic simplicity.
On the other hand, GIS as a powerful tool in decision support, can reduce the time and cost of evaluation process and can assist the users in selection of an appropriate strategy (Leung, 1992). Although computer and GIS application in studying the natural resources are expanding, the methods of data integration and analysis have seen a paucity in discussion.
Minimum Distance to Ideal Points
Minimum distance to ideal point (MDIP) is a common technique in pattern recognition and clustering of data when using satellite data and raster GIS. This method is effective in evaluation of both continuos and discrete variables (Duckstien and Opricovic 1980). This is based on calculating the Euclidean distance of each alternative land use to the ideal point. The process includes pinpointing of ideal points, developing alternative metrics, standardizing data, determining relative significance of each criteria, and finally, analyzing data (Pereria and Stein,1993).
Briefly said, in multi-criteria evaluation, user is faced with a set of alternative and a number of different factors. In this set, each alternative has a number of characteristics, related to land suitability factors attached to it. Through this, the scores for each alternative is determined, within which, at least one MDIP is included. The land use alternatives with higher scores are more important than others.
In such cases, resources can be allocated to not one but several land use according to longitudinal or location considerations. In such situation and specially due to high volume and complexity of data, the planners cannot reach a valid conclusion. Here, MDIP can be used to drive the alternatives. In this method, each of the suitability maps, which are produced in the multi-objective evaluation process, are placed in a vector space, along one axis. In such a space, each alternative is within the reach of the decision maker based on its level of suitability. Selection of best alternative for each land use takes place by using the decision making line. This line is the best fit on the vector space.
Methodology
In this study, ARC/info and IDRISI are used as GIS softwares. The land use map is produced by using digital data from Landsat-5 TM through hybrid classification. The greenness and wetness maps are created by using tasseled cap technique. Digital Terrain Modeling (DTM) of the Talighan region including height, aspect and slope is created through digitizing topography map of the region. The existing soil and geological maps are also digitized and are used in determining land suitability for agricultural, pasture, orchards, urban and tourism land use.
As discussed above, multi-objective evaluation in MDIP method requires determination of ideal points for each factor. The factor used in this study have different attributes (Including continuos and discrete). Here , two different methods are used to define the ideal points for each variable taking into account the fact that variables are multi-faceted. First, variables such as soil type and aspect are considered as discrete. Thus, for each land use, the frequencies for these are calculated.
By extracting information on variables to create the evaluation metrics, the matrix elements are valued from 0 to 255. This information is used in data analysis. Weights for variables are assigned through pariwise comparison and expert judgment. Here by using the weights, the land suitability map for each land use is created.
Also, for comparison purposes, the common Boolean (Overlay) method is used to find suitable land use for the region. Here again, land suitability maps are created which show the relative suitability of alternatives for each land use. Through integration of such maps, the multi-objective land use map is also produced. Also, the present land use situation of the region is mapped and based on this, the significance of each land use in the region is determined and priorities are listed. Using different weights for each land use, the multi-objective land use map is create.
