Sapanca Drinking Water Basin has been chosen as the study area. The fact that the natural areas and water resources of the basin were under threat due to the approximately double increase in the population in the basin from 2000 to 2017 and the uncontrolled urbanization effect in order to fulfil the housing and social demands of the increasing population, was effective in determining the area as a study area (Yıldız Dönmez, 2019). The Sapanca Basin is located in the Çatalca-Kocaeli plateau in the east of the Marmara Region. It is in the hinterland of İstanbul-Ankara metropolises. The basin is administratively divided between Kocaeli and Sakarya provinces, 26 km of the 39 km lake coastline is within Sakarya and 13 km is within Kocaeli provincial borders (SASKİ, 2013). Sapanca Lake, which gives its name to the basin, is a crucial natural area and water resource for the region (Fig. 1).
As the study area is a crucial drinking water basin, the sustainability of its natural structure and water resources becomes very important. In recent years, serious changes have been detected in land use in the basin. The main factor in the change of landscapes within the basin is the urbanization pressure. The forest areas in the basin have converted into agricultural and residential areas over the years with the effect of urbanization, therefore, suitability analyzes have been carried out to determine settlement and agricultural areas within the scope of the study in order to direct the development of settlement and agricultural areas in the future (Yıldız Dönmez, 2019).
The existing land use map satellite images, aerial photographs, forest stand maps, Kocaeli and Sakarya Metropolitan construction plans and existing maps and area researches related to the study area have been formed by transferring 1/25.000 scale forest stand maps obtained from Sakarya Regional Directorate of Forestry to the GIS environment. All data have been brought to the same coordinate system by georefarating with the UTM coordinate system in the GIS environment, and overlapped with the basin boundary received from SASKI and accepted as the study boundary. After the coordinating process of satellite and aerial photographs, controlled classification have been conducted to manual digitizing structures. The basin is divided into 6 classes as forest, agricultural land, degraded forest, pasture, meadow and settlement. On the other hand, residential areas have also been classified as industrial, commercial and residential areas based upon the researches and determinations made in the area and a map of the current area uses has been obtained.
The overlay method developed by McHarg has been used in the suitability analysis. McHarg defines suitability analysis as preparation and interpretation of ecological inventory, forming and mapping an inventory of natural and cultural resources, grouping suitable and unsuitable area uses, synthesizing data for suitability map, overlaying these maps according to the purpose and forming suitability maps (Cengiz et al., 2009). Ecologically based suitability maps have been prepared for residential and agricultural areas in the study area. Criteria and sub-criteria have been created for each area to prepare suitability maps. The studies of Akten (2008), Cengiz et al. (2009), Konaklı (2011) have been used to evaluate these criteria and sub-criteria. For settlement and agriculture suitability analysis, potential erosion risk analysis values, water function analysis values and basin protection zones specified in the Drinking Water Protection Regulation have also taken in addition to ecological criteria such as slope, aspect, land use capability, erosion, geological classes. Sub-criteria for each area are scored from 1 to 4 according to their suitability. In this scoring, 1 represents the lowest, 2 the low, 3 the suitable, and 4 the most suitable (Table 1). The sub criteria for settlement and agriculture are evaluated in terms of potential uses and given values between 1 and 4. For instance, when the criteria of suitable topographic structure are evaluated, the slope between 0–6% is evaluated as the most suitable with 4 points, 6–12% as suitable for settlement value with 3 points 12–24% less suitable with 2 points and areas larger than 24% as not suitable areas with 1 point. Similar to the slope, 1 and 4 suitability points are given for the landscape, geological structure, erosion, land use capabilities, water function analysis, potential erosion analysis, and synagogue protection zones, which are determined as sub-criteria for settlement. The same method has been used in agricultural areas. For example, 1st and 2nd class lands have been determined as 4 points as the most suitable areas for agriculture in land use capabilities, 3rd class lands have been determined as suitable areas with 3 points, 4th and 5th class students have been determined as less suitable with 2 points and 6th, 7th and 8th class have been determined as unsuitable with 1 point. This scoring system has been calculated in the same way among all criteria determined in the suitability analysis for agriculture.
By overlapping the criteria in GIS programs, a cumulative structure has been formed and optimal area uses have been formed by classifying them as "very suitable", "moderately suitable", "less suitable" and "not suitable" according to the FAO classification system (FAO, 1977). Optimal land uses have been compared with the existing land use map and Environmental Plan.
Table 1: Criteria for optimum land-use settlement and agricultural analysis