Land suitability evaluation for changing spatial organization in Urmia County towards conservation of Urmia Lake

Applied Geography 81 (2017) 1e12

Contents lists available at ScienceDirect

Applied Geography journal homepage: www.elsevier.com/locate/apgeog

Land suitability evaluation for changing spatial organization in Urmia County towards conservation of Urmia Lake Hedayat Nouri a, *, Robert J. Mason b, Nosrat Moradi a a b

University of Isfahan, Isfahan, Iran Temple University, Philadelphia, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 February 2016 Received in revised form 7 January 2017 Accepted 7 February 2017

One of the most effective ways to reverse the decline of Iran's iconic Urmia Lake is to directly confront the development patterns that have contributed to the current crisis. This study's objective is to create a model for conservation of Urmia Lake that identifies suitable lands for agricultural and residential development in Urmia County that are distant from Urmia Lake. This was accomplished through a Geographic Information System-based multi-stage process. The first step involved production of maps based on an initial assessment of the region's geography, geomorphology, landforms, and hazards potential. In the next step, all of the parameters were overlaid and land suitability maps were generated by using determinant maps. In developing the final map, the lands were divided into four classes for future development potential: highly suitable, suitable, marginally suitable and not suitable. The results showed that well away from highly populated regions adjacent to Urmia Lake, there are highly suitable and suitable lands that presently contain 14 and 8 percent of total settlements, respectively. The highly suitable and suitable lands, which cover 5.6 and 6.7% of the total area, may serve as appropriate axes for changing the traditional spatial organization of the region, redirecting future development and consequently decreasing ecological pressure on lands close to Urmia Lake as well as preventing further excessive usage of water resources in the region. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Urmia Lake Land suitability Spatial organization Geographic information system

1. Introduction Urmia Lake, located in the northwestern part of Iran, has an area of approximately 6000 km2 and contains 102 islands (in its pre1999 condition). It is one of the great hyper-saline lakes of the world and the largest lake in the Middle East (Ghaheri, Baghal, & Naziri, 1999; Hassanzadeh, Zarghami, & Hassanzadeh, 2012; UNEP & GEAS, 2012). The lake, which is a national park, was added to the Ramsar List of Wetlands of International Importance in 1971 and designated a UNESCO Biosphere Reserve in 1976 (AGH, 2014; Manaffar et al., 2011; Marjani & Jamali, 2014; UNEP & GEAS, 2012). Urmia Lake is home to many faunal species, including a unique brine shrimp species (Artemia Urmiana), flamingo, ducks, pelicans, and mammals; it is also habitat for an array of vegetative species. In addition, the lake is a natural asset that has been considered unique in terms of ecological, environmental, cultural, economic, aesthetic, recreational, scientific and conservation values for many years (Abbaspour, Javid, Mirbagheri, Ahmadi Givi, &

* Corresponding author. E-mail address: [email protected] (H. Nouri). http://dx.doi.org/10.1016/j.apgeog.2017.02.006 0143-6228/© 2017 Elsevier Ltd. All rights reserved.

Moghimi, 2012; Tourian et al., 2015; UNEP & GEAS, 2012). In recent years, Urmia Lake has experienced a rapid shrinkage. Satellite images reveal that the lake's area was 6100 km2 in 1995 and has declined to 2366 km2 in August 2011(UNEP & GEAS, 2012), and then to 953 km2 in August 2013 (AGH, 2014). Fig. 1 shows the changes in Urmia Lake's area between 1999 and 2014. Various factors explain the drying up of Urmia Lake. Demographic developments, human activities and especially demand for water in farming lands near the lake are among the main factors causing the lake's shrinkage (Abbaspour et al., 2012; AGH, 2014; Dehghanzadeh, Safavy Hir, Shamsy Sis, & Taghipour, 2015; Fathian, Morid, & Kahya, 2014; Hassanzadeh et al., 2012; Kakahaji, Banadaki, Kakahaji, & Kakahaji, 2013; Madani, 2014; Tourian et al., 2015; UNEP & GEAS, 2012). This phenomenon is quite comparable to the case of the Aral Sea during the Soviet and post-Soviet eras ((Cai, McKinney, & Rosegrant, 2003; Conte, 1995; Gaybullaev, Chen, & Gaybullaev, 2014; Glantz, 2005; Kravtsova & Tarasenko, 2010; Levintanus, 1992; Lioubimtseva, 2015; Morimoto, Natuhara, Morimura, & Horikawa, 2005; Rafikov & Gulnora, 2014; Rifikov, Rifikova, & Mamadganova, 2014; UNEP & GEAS, 2014). In 1980, the total cultivated area within the Urmia Lake basin was 150,000 ha, but by 2007, it had increased to

2

H. Nouri et al. / Applied Geography 81 (2017) 1e12

Fig. 1. Urmia Lake in 3 July 1999, 20 July 2014.

400,000 ha (TMU & CIWP, 2012). In short, the total area of cultivated lands has increased dramatically, nearly tripling in less than three decades. Consequently, during this period, the usage of water by the agricultural sector has increased from 1.8 billion cubic meters (BCM) to 5.5 BCM (Ibid). The Urmia Lake watershed, with a total area of 51,876 sq2 and a population of approximately 6.4 million, is an important agricultural region (UNEP & GEAS, 2012). One of the highly populated areas in this region is the eastern part of Urmia County. This area, which is located in the western portion of the Urmia Lake watershed, contains around 15 percent of the total cultivated lands in the watershed (TMU & CIWP, 2012). The population density of this area is approximately 621 persons per sq2, while it is less than 125 persons per sq2 for the entire Urmia Lake watershed. In addition, approximately 511 million cubic meters (26 percent) out of 1959 million cubic meters of total groundwater withdrawals in the Urmia Lake watershed takes place in this area (UNDP/GEF, 2008). Also worth noting is that approximately 17 percent of surface water in the Urmia Lake watershed historically has flowed to Urmia Lake from sub-basins of Urmia County, but these waters no longer enter the Lake because of damming and creation of reservoirs (Marjani &

Jamali, 2014). This area is also considered the center of economic activity and agricultural development for Urmia County and the Urmia Lake watershed. The overall aim of this study is to explore the changing spatial organization of Urmia County and examine alternative spatial organization strategies for conservation of Urmia Lake, such as proposing new areas for agriculture and housing development far from Urmia Lake, which would decrease the population concentrations in areas close to the lake. In order to propose changes in the spatial organization of Urmia County, initial assessment of the region's geography needs to address this critical question: is there suitable land for future development outside of coastal areas of Urmia Lake that could attract people away from densely populated coastal areas? The present study assumes that there are suitable lands for development in such areas. In addition, it aims to identify and propose suitable lands outside of coastal areas as an alternative setting for the region's future development. Regarding the above assumptions, the issue of land is considered as the most important limitation and effective natural factor in reconfiguring spatial organization processes in the region.

H. Nouri et al. / Applied Geography 81 (2017) 1e12

3

The phenomena of drying and drastic reduction in area and storage capacity of Urmia Lake in less than two decades have been addressed in recent years by a number of researchers as an environmental crisis (AGH, 2014; Abbaspour & Nazaridoust, 2007; Abbaspour et al., 2012; Azarnivand, Hashemi-Madani, & Banihabib, 2015; Fathian et al., 2014; Ghorbani-Aghdam, Dinpashoh, & Mostafaeipour, 2013; Hassanzadeh et al., 2012; Kakahaji et al., 2013; Karbassi, Bidhendi, Pejman, & Bidhendi, 2010; Noury, Sedghi, Babazedeh, & Fahmi, 2014; Tourian et al., 2015; UNEP & GEAS, 2012; Zarghami, 2011; Zeinoddini, Bakhtiari, & Ehteshami, 2014); these studies focused on specific aspects of the issue, mainly the causes of the disaster and its natural and environmental consequences.

elevation is 1762 m above mean sea level and average slope is 22 percent. Most of the township is located in a mountainous area (more than 75%) and the remaining area is occupied by a vast alluvial plain in the western part of the Urmia Lake. Because of the mountainous character of Urmia County, the majority of residents have settled in the plains and flatlands adjacent to Urmia Lake. In 2011, More than 85 percent of Urmia County's population, nearly 830,709 residents (Iran, 2011), resides along the alluvial plains close to Urmia Lake, indicating the high population density in this area. In the absence of purposeful land-use planning for the region, it has become a very highly populated area with intensive urban and rural land uses; this current trend continues to accelerate.

2. Materials and methods

2.2. Preparation of data

2.1. Study area

Spatial planning is based on the physical geographical conditions and especially terrain characteristics within a region (Fontes & Pejon, 2008; Lein, 2003; Marsh, 1998). Due to high levels of relief within the study area, we have chosen to focus on the land and its characteristics as the major subjects of the current study. In this paper, employing a set of primary criteria for selecting the most pertinent parameters and determining their scores for land

The study region, Urmia County, with a total area of 5312 km2, is located in the northwestern part of Iran, positioned between 44
240 -45
250 East longitude and 37
70 -38
100 North latitude. It is situated within the Urmia Lake watershed (Fig. 2). The elevation of the area ranges from 1270 to 3596 m above mean sea level. Average

Fig. 2. Standardized layers.

4

H. Nouri et al. / Applied Geography 81 (2017) 1e12

suitability constitutes the basis of the geographical and environmental theoretical framework, which incorporates relevant literature and the regional conditions simultaneously. Next, all of the required parameters for analysis were defined and classified. After gathering all of the data from relevant organizations, a digital database was created in the GIS environment. The layers used in the study process are shown in Table 1.

and then combining the weights and standardized suitability maps to obtain an overall suitability score. Unlike the Boolean operations, weighted linear combination is a compensatory method in the sense that a low score on one suitability criterion can be compensated by a high suitability score on another (K. Chen, Blong, & -Riveira, CrecenteJacobson, 2001; Malczewski, 2004; Sante  s, 2008). WLC can be defined as follows: Maseda, & Miranda-Barro

2.3. Methods

S¼

2.3.1. Multi criteria evaluation (MCE) There are three major groups of approaches to GIS-based landuse suitability analysis: (i) computer-assisted overlay mapping, (ii) multi-criteria evaluation (MCE) methods, and (iii) soft computing or geocomputation methods (Malczewski, 2004). The GIS-based multi-criteria evaluation method as the foundation for decisionmaking has been widely applied in environmental planning (Azizi, Malekmohammadi, Jafari, Nasiri, & Amini Parsa, 2014; J. Chen, Zhang, & Zhu, 2011). The integration of multi-criteria evaluation with GIS has considerably advanced the conventional map overlay techniques used in land-use suitability analysis (Malczewski, 1999; X.; Zhang, Fang, Wang, & Ma, 2013). “Multicriteria decision analysis has two critically important characteristics: One is the GIS capabilities of data acquisition, storage, retrieval, manipulation and analysis, and other is the multi-criteria decision making capabilities for combining the geographical data and decision maker's preferences into unidimensional values of alternative decisions” (Malczewski, 2004). Land use suitability evaluation techniques based on the MCE-GIS can be expressed by a normal equation:

S ¼ f ðxÞ

(1)

where S is the suitability level; x is the evaluation index (criteria)which has a strong influence on the suitability level; f(x) denotes the decision rule (X. Zhang et al., 2013). 2.3.2. Weighted linear combination (WLC) Over the past several decades, a number of multi-criteria evaluation methods have been proposed for GIS-based land suitability analysis (Lafortezza, Chen, Sanesi, & Crow, 2008). Among these methods, the Boolean overlay operations and the weighted linear combination (WLC) are the most common methods in land-use suitability evaluation. These two fundamental classes of multicriteria evaluation methods are the most frequently used decision rules for combining a set of criterion maps (Beinat & Nijkamp, 1998; Boroushaki & Malczewski, 2008; J.; Chen et al., 2011; JelokhaniNiaraki & Malczewski, 2015; Lafortezza et al., 2008; Malczewski, 2000, 2004, 2006; Malczewski & Rinner, 2005). The weighted linear combination approach, which is easily implemented in a raster GIS model, involves standardization of the suitability maps, assigning weights of relative importance to the suitability maps,

n X

wi xi

(2)

i¼1

where S is the suitability level; xi is the value of the index i; wi is the weight of the index i; and n represents the number of indexes. There are four steps in WLC based on Equation (2), which are selecting the index (i), providing index value (xi), determining index weight (wi) and using the overlaying rule (C. Zhang, Li, & Day, 2006; X. Zhang et al., 2013).

2.4. Standardization of parameters In order to perform the weighted linear combination method, each attribute must be converted into a comparable scale. By applying a standardization function, all parameters will have commensurate values, allowing for their subsequent aggregation. There are various standardization methods, including deterministic, probabilistic and fuzzy methods (Anagnostopoulos, Vavatsikos, Spiropoulos, & Kraias, 2010; Fischer & Nijkamp, 1993; Malczewski, 1999; Qiu, Chastain, Zhou, Zhang, & Sridharan, 2014). The map layers were standardized using the fuzzy procedure. The fuzzy set was introduced by Zadeh in 1965. “Fuzzy logic is a superset of conventional (Boolean) logic that has been extended to handle the concept of partial truth” (Malczewski, 2002). “A fuzzy set is a class of objects with a continuum of grades of membership. Such a set is characterized by a membership function which assigns to each object a grade of membership ranging between zero and one (Zadeh, 1965, p. 338), as opposed to classical set where each element must have either 0 or 1 as the membership degree” (Malczewski, 2004; Zadeh, 1965). “The capability of fuzzy sets to articulate gradual transitions from membership to nonmembership has a broad utility not only for representing geographical entities with imprecise boundaries, but also for GISbased operations and analyses including multi-criteria decision analysis. The concept of membership function can be used as a tool for standardizing criterion maps” (Malczewski & Rinner, 2015, p. 197). “The fuzzy procedure involves specifying a fuzzy set membership function, which can take one of the following forms: sigmoidal, J-shaped, linear, or user-defined function” (Malczewski & Rinner, 2015). In ArcGIS 10, the fuzzy memberships are provided for standardization of parameters. In this study, the linear fuzzy membership was applied for standardizing of parameters. In layers such as the land type, land

Table 1 Source and the type of the data. Parameter

Type

Source

Elevation Slope Aspect Land type Hazards (digitized) Land cover (digitized) Sub basin Population density Distance from Urmia Lake

Raster Raster Raster Vector Vector Vector Vector Vector Raster

Extracted from ASTER GDEM* Extracted from ASTER GDEM Extracted from ASTER GDEM Department of Natural Resources Geological Survey of Iran Ministry of Roads & Urban Development Forest, Range & Watershed Management Organization of Iran This study This study

H. Nouri et al. / Applied Geography 81 (2017) 1e12

5

cover and sub basins, which included thematic attributes, raw scores ranging from 1 to 9 were assigned in accordance with their relative importance, and then these layers were standardized using the linear fuzzy membership. In the case of continuous attributes, such as elevation, slope, aspect, distance from hazards extents, distance from the lake and population density, the standardization was also done using the linear fuzzy membership. Thus all of the parameters were standardized based on the linear fuzzy membership and were given scores ranging between 1 and 0, with 0 as the least and 1 as the maximum suitability level for each layer (Table 2 & Fig. 2).

information and knowledge” (Malczewski & Rinner, 2015, p. 196, p. 196). Based on the first rule, regarding the importance of slope in the human geographical landscape (Marsh, 1998; Saxena, 2004), which assessment of regional population and settlement distribution confirms as the leading parameter, the slope layer was given the highest weight value of 0.5, while the elevation and aspect were given relatively lower weights of 0.3 and 0.2 respectively (Table 3).

2.5. Parameter weights

After the scores of the sub-parameters were assigned and the parameters were weighted, all of the vector layers were converted to raster format with a 30 m  30 m pixel size. Because the Digital Elevation Model (DEM) was prepared with a pixel size of 30 m, all of the layers were resampled in 30 m pixel size. The standardized parameters then must be overlaid. In order to do this, the land suitability model was constructed using ArcGIS Model Builder and all of the parameters were input into Model Builder. Then, the overlay operation was conducted by using weighted sum overlay analysis in the Model Builder environment. The overlay operations process is illustrated in Fig. 3. As shown in Fig. 3, the multi-staged overlay operations used in this study consists of the following steps:

Regarding the processes of overlaying the layers and the multistaged evaluation, the weights of the parameters were assigned based on two general rules: if the output layers were more than two layersdfor example, slope, elevation and aspectdweights ranging from 1 to 0 were assigned in accordance with relative importance to the layers and the region's conditions, while if the input layers were two layers, including input and output layers, the output layer was weighted at 0.75 and the new parameter at 0.25. Applying this rule can decrease criterion weight uncertainty which “in some situations the decision maker may be unable to exactly specify his/her preferences due to limited or imprecise

2.6. Overlay operations

Table 2 The classes, values and fuzzy memberships of the parameters (The column shown in bold represent the standardized values). Parameters

Class

Attribute\class value range

Value

Fuzzy membership

Elevation

1 2 3 4 5 1 2 3 4 1 2 3 4 1 2 3 4 5 6 7 8 1 2 1 2 3 4 5 1 2 1 2 1 2 3 4 5 6 7 8 9

1217e1500 1500e1800 1800e2100 2100e2500 2500e3596 <3 3e10 10e25 >25 Aspects in slope less than 3% North, Northwest Northeast, East, Southeast, South Southwest, West Mountains Hills Plateau and Upper Terraces Piedmont Alluvial Plain River Alluvial Plain Low Lands Flood Plain Gravelly Colluvial Fans <200 >200 S1, S2, W1D1, W1D2, W1D3, W2, R1, R2, F2, F3, U1, U2 R2, R2R3, RA O, R3, R3R4 A2, A3 R4, AO, A1 Urmia, Rashakan Ziveh, Serow, Silvana 0e171 171e166,874 0e6 6e13 13e20 20e27 27e33 33e40 40e47 47e54 54e60

1343 1625 1933 2264 2786 1.82 6.16 16.25 44.13 9 3 7 5 1 3 7 7 9 1 1 3 100 200 1 3 5 7 9 1 9 0e171 171e166,874 3 9.5 16.5 23.5 30 36.5 43.5 50.5 57

1 0.80 0.59 0.36 0 1 0.75 0.35 0 1 0 0.67 0.33 0 0.25 0.75 0.75 1 0 0 0.25 0 1 0 0.25 0.5 0.75 1 0 1 0 0.12 0.25 0.38 0.5 0.62 0.75 0.88 1

Slope

Aspect

Land Type

Hazards (floods and earthquakes) Land Cover

Sub Basins Population Density* (Sq KM) Distance from Urmia Lake (Azizi et al.)

6

H. Nouri et al. / Applied Geography 81 (2017) 1e12 Table 3 The weights of the parameters and overlaid maps. Parameters

Weight

Output Map

Elevation Slope Aspect Land Type and Hazards Landform Land Cover Geomorphology Determinant Map (Distant from Urmia Lake, Population Density, Sub Basins) Potential Map

0.3 0.5 0.2 0.25 0.75 0.25 0.75 0.25 0.75

Landform

Table 4 The Areas and percentage of the classes of final land suitability map. Suitability classes

Area (Km2)

Area (%)

Highly suitable Suitable Marginally suitable Not suitable Total

289.50 594.36 666.63 3662.29 5212.8

5.55 11.40 12.79 70.26 100

Geomorphology Potential Map Suitability Map

necessary to assign their weights with respect to each other. After weights were assigned, the parameters were overlaid and the landform map was created. (ii) Adding and testing of additional environmental parameters was facilitated, in part, by producing the base map. So, the geomorphology map was produced by combining land type, hazards and landform layers. To do this, all of the land type units were scored based on the literature review and distri-

Fig. 3. Land suitability evaluation model for changing spatial organization in Urmia County.

(i) In the first step of the process, the landform map was provided as the base map for the next part of the analysis. In order to prepare the landform map, the elevation map was divided into 5 classes using the Natural Breaks classification method. Then, based on the literature review, the slope map was divided into 4 classes which are suitable for settlement and activity (Marsh, 1998). The aspect parameter also was classified based on the regional conditions. Due to the differing influences of the parameters in the region, it was

bution of settlements in the region. Based on the regional conditions and similar studies, a 200 m buffer zone was created around the flood and fault layers and the hazards map was produced. (iii) In the next step, the land cover layer was combined with the geomorphology map. The land cover layer was classified into 5 groups in terms of land capability for settlement, development, and agriculture and non-agricultural activities. The first group (R4, AO, and A1) contains the irrigated and dry

H. Nouri et al. / Applied Geography 81 (2017) 1e12

farming lands and rangelands with 5e25 percent canopy cover, suitable for housing development. These land covers have no limitation for development and were assigned the highest score. The second group (W1D1, W1D2, W1D3, W2, R1, R2, F2, F3, S1, S2) is mainly comprised of lands with high soil salinity, wetlands, and forest and high-quality rangelandsdall of which are not suitable for development due to their physical limitations or conservation values. This group was given the lowest score. The third and fourth groups (A2, A3, O, R3, R3R4) consist of the lands that are suitable and marginally suitable. The fifth group (R2, R2R3, RA) also consists of lands that are not suitable for development and were subsequently assigned the lowest scores. (iv) In the final step, to produce the suitability map, the determinant maps that include distance from Urmia Lake, population density and sub basins layers were overlaid with the potential map. In overlay operations introduced above, the fuzzy values of overlaid layers were reclassified into four suitability classes, namely, highly suitable, suitable, marginally suitable and not suitabledbased on Jenk's scheme for Natural Breaks grading in GIS environment. This method, which has been used by various researchers in different fields (e.g.Bathrellos, Gaki-Papanastassiou, Skilodimou, Skianis, & Chousianitis, 2013; Karakas¸, 2013; Shi, Wang, & Yin, 2013; Swetnam et al., 2011), creates classes according to clusters and gaps in the data (Ormsby, Napoleon, Burke, Groessl, & Bowden, 2010). 3. Results and discussion As mentioned previously, the land and its management always play an important role in the larger environmental picture. In this

7

study, regional physical characteristics are given higher priority than is the case in typical regional approaches. An accepted geographical principle is that whenever a parameter is limited in geographical space, that parameter becomes the more important one. This is true for Urmia County, where the land is the main constraint in the region. Based on common procedure in environmental planning, this stage starts with the study of landforms (Beer & Higgins, 2004; Fontes & Pejon, 2008; Golley & Bellot, 1999; Marsh, 1998). From the statistics which were extracted from the landform layer (Fig. 4) by Zonal Statistics as Tableda subset tool provided in Spatial Analyst tools of ArcGIS Desktopdit was found that the lands classified as highly suitable cover approximately 1539.59 km2, which is 29.23% of the total land area. The other classesdsuitable, marginally suitable and not suitabledcover 30.84, 22.20 and 17.73% of the total land area, respectively. In this step, the land type units, which included various attributes, were added to the base map to facilitate the evaluation of land suitability in the region. The land type units are comprised of the main units, including the most valuable lands such as the river alluvial plains, plateau, upper terraces and piedmont alluvial plains, and the least valuable lands, such as hills, gravelly colluvial fans, and also unusable lands such as mountains and lowlands. In order to identify the land characteristics comprehensively, it was necessary to add the hazards map to the new output layer. Studies at the regional level show that floods and earthquakes are considered to be the main hazards in the region. Therefore, we focused exclusively on these two parameters to produce the hazards map. By combining these two layers (Fig. 5) and overlaying them with the previous output map, a new map was obtained which is called the geomorphology map (Fig. 6). By overlaying the land type-hazards layer with the landform layer and producing the geomorphology map, the percentage of

Fig. 4. Landform map.

8

H. Nouri et al. / Applied Geography 81 (2017) 1e12

Fig. 5. Land type and Hazards maps.

Fig. 6. Geomorphology map.

highly suitable and suitable classes have decreased to about 20.83 and 25.54%, respectively. Of the remaining areas, marginally suitable area has increased to 28.99% and unsuitable areas to 24.63%. The previous sections presented the main characteristics of the

terrain. Considering the purpose of the land suitability evaluation for development, with the emphasis on agriculture, it seems that land cover plays an important role in identification of quality of available lands, especially in the flatlands, and in our initial

H. Nouri et al. / Applied Geography 81 (2017) 1e12

assessment of the lands. The land cover map, as the last partial map, was overlaid with the geomorphology map and, encompassing all of the relevant basic environmental parameters, named the “potential map” for the region (Fig. 7). This map can be regarded as the potential map for Urmia County for the development, before taking the purpose of this study into consideration. The results of the potential map revealed that the area of the highly suitable class has decreased once again. Based on the results of this map, therefore, it was determined that highly suitable areas for agricultural and residential development occupy 968 km2, i.e. 18.6%, of the total area of Urmia. In the first phase of the study, the region's development potential was evaluated and land suitabilities were identified. It was obvious that the plains close to Urmia Lake are the most suitable lands for development, whereas the main objective of the present study is to identify the possible alternatives outside of these fertile lands for development. To do this, in the next phase, all of the required parameters were selected and overlaid. In the new map, which is named “determinant map” (Fig. 8), three main factors were considered: First, based on the spatial extension of the sub-basins in Urmia County, the sub-basins of the region were classified into two groups/regions. The first group contains the Urmia and Rashakan sub-basins, whose waters enter Urmia Lake directly. This region was given the lowest score. The second group includes Serow, Silvana and Zive sub-basins, whose waters enter Urmia Lake via the first region. This region was assigned the highest score. The second factor involves identification of suitable lands for settlement and agricultural development by considering distance from the lake. In order to achieve this, a Euclidean distance map for distance from the lake was generated. Then, the distance from the lake map was classified into 9 groups using the Natural Breaks classification method and, as with other parameters, was standardized by using the fuzzy linear method. The resultant map

9

shows that the land values for agricultural activity and residential development increase in the areas further away from the lake and vice versa. The third and final factor is population density, which because of its general importance, was implied in the study. This factor plays a critical dual role. On the one hand, due to the effect of population density on water usage, it works against our overall study objective and thus is assigned a negative score. The population density of 621 persons per square kilometer in the flat lands closer to the lake, compared with the population density of 171 per sq2 for all of Urmia County, can be a significant and clear sign of the water usage problem. On the other hand, the high population density can be regarded as a sign of the land's potential for development. Thus high and low population densities can be regarded as positive and negative factors, respectively. In order to take into account these two characteristics, the population density map was scored based on triangular distribution. Considering this, we have used the Urmia County population density (171 per square kilometers) as the basis for scoring. Thus, in low and high population density areas, with respect to the average population density of 171, the land suitability score increases and vice versa. The population density map, as the last parameter of the determinant map, was thus produced. The primary goal of the determinant map was to modify the final map of the first phase (i.e., the potential map) towards achieving the final objective of this study. The final map is the result of overlaying the potential and determinant maps. The resultant map is represented as Fig. 8; this land suitability classification provides possible alternatives for the development in the area beyond the fertile lands close to Urmia Lake. Table 4 presents statistics which relate to Fig. 9. With respect to the objective of this study, the suitability classes of the final map showed that about 289 km2, which covers 5.6% of the total area, is

Fig. 7. Potential map.

10

H. Nouri et al. / Applied Geography 81 (2017) 1e12

Fig. 8. Determinant map.

Fig. 9. Development suitability map for Urmia County.

highly suitable and 594.36 km2, i.e. 11.4% of the total area, is suitable for agricultural and residential development. In broader perspective, then, a multi-stage process was applied to develop a land-suitability approach, with the broad objective of

changing the spatial organization of Urmia County in order to promote conservation of Urmia Lake. To do this, a set of evaluation criteria were selected and compared. By producing landform, geomorphology and potential maps, current spatial organization

H. Nouri et al. / Applied Geography 81 (2017) 1e12

was depicted. Given their critical role, landform parameters played a key part in generating the results and developing suitability maps. Development of the landform map, which incorporates the viewpoints of Marsh (1998) and Saxena (2004), demonstrated that geomorphology significantly impacts a majority of the region's lands and land use potential. Indeed, the region's mountainous character makes terrain the leading and determinant factor among those analyzed. The complementary factors of land type and hazards, incorporated into the geomorphology map, also act as important constraining factors with respect to suitability for development. In the resultant final maps, the most important threats are in the form of environmental hazards, which limit land potentials; thus hazards were added to land characteristics. Due to the importance of land cover, it was used in the analysis to propose suitable alternative lands for agriculture and residential development far from Urmia Lake. In spite of the fact that this parameter is not naturally a constraint, but it acted as a constraint parameter in the analysis. As determined from the “potential map” (Fig. 7), the preponderance of highly suitable and suitable lands are located in the eastern part of Urmia County close to Urmia Lake. To spatially analyze the issue and address the situation of current spatial organization, the suitability classes of the potential map were overlaid with settlement and sub basins layers (Fig. 7). As can be seen from this map, the greatest concentrations of population, which contain 360 settlements with a population of approximately 851,221, i.e. 88 percent of the total population, are within the highly suitable and suitable classes. These lands, which are the most intensive agricultural areas, are located in the Urmia and Rashakan sub-basins. As mentioned previously, the drying of Urmia Lake is linked to the preventing of flowing water surfaces of these subbasins to Urmia Lake. While, the excessive usage of groundwater is also extracted in this area. Thus, this area can be regarded as the ecological pressure zone-and deconcentration of population and activities within it could very significantly assist in the preservation of Urmia Lake. In producing the current spatial organization map and recognizing the most significant natural factors that affect development suitability, it has become cleardbased on our initial assumptiondthat land is the most important limitation and effective natural factor in possible alternative spatial structures for the region. In order to achieve the overall objective of this studydby incorporating parameters of population density, sub basins and distance from Urmia Lake, and producing the land suitability mapdthe current spatial organization was reconfigured, and suitable lands for development were identified in the western part of Urmia County, relatively far from the ecological pressure zone. In this reconfigured spatial organization, the highly suitable and suitable lands–located in the Serow, Silvana and Ziveh sub-basins (Fig. 9)d cover 5.6 and 6.7 percent of the total area, respectively, and contain 14 and 8 percent of the total settlements. These lands nearly include 7 percent of the total population. In the suitable lands identified by this study, the population density is approximately 117 persons per sq2, while it reaches more than 482 persons per sq2 for the ecological pressure zone. Despite the natural limitations of the region for development, per our initial assumption, land-use alternatives for the area close to Urmia Lake can be identified and regional planners can use this information to promote the appropriate utilization of these lands. 4. Conclusions This paper is aimed at promoting conservation of Urmia Lake through the evaluation of the terrain and lands outside of the

11

environmentally problematic area of Urmia County. It builds upon previous work regarding the Urmia Lake ecological crisis by developing a model for We suggest possible alternatives and new areas for future development by reconfiguring the spatial organization of Urmia County, which has presently concentrated about 85 percent of human activities around Urmia Lake. In this study, we have sought to identify land potential in a multi-stage process, based on the geographic characteristics of Urmia County. First, various parameters were used and the potential of the region was identified. Then, the determinant map, incorporating sub basins, distance from the lake and population density factors, was created to achieve the final objective of the study. Finally, the suitability map was produced. This study shows that there are some areas located further away from Urmia Lake that present possible alternatives for future development. The study demonstrates that there are highly suitable and suitable lands in the western part of Urmia County, which cover 5.6 and 6.7 percent of the total area, respectively. These lands contain only 14 and 8 percent of the total settlements and a small proportion of the region's economic activities. These possible alternatives can be viewed not only as an essential priorities for future regional development and a reconfigured spatial organization of the region, but also as an effective alternative for decreasing the ecological pressure on the lands close to Urmia Lake and preventing excessive usage of water resources in such areas. These goals can be achieved through (1) promoting new activities such as developing agricultural plans, establishing new towns and building residential complexes in proposed lands distant from Urmia Lake; and (2) encouraging and supporting residents to relocate their activities, especially those activities that greatly affect Urmia Lake with respect to water usage. References Abbaspour, M., Javid, A. H., Mirbagheri, S. A., Ahmadi Givi, F., & Moghimi, P. (2012). Investigation of lake drying attributed to climate change. International Journal of Environmental Science and Technology, 9(2), 257e266. http://dx.doi.org/10.1007/ s13762-012-0031-0. Abbaspour, M., & Nazaridoust, A. (2007). Determination of environmental water requirements of lake Urmia, Iran: An ecological approach. International Journal of Environmental Studies, 64(2), 161e169. http://dx.doi.org/10.1080/ 00207230701238416. AGH, N. (2014). How to save the dying lake Urmia. Acta Geologica Sinica (English Edition), 88(s1), 178e179. Anagnostopoulos, K. P., Vavatsikos, A. P., Spiropoulos, N., & Kraias, I. (2010). Land suitability analysis for natural wastewater treatment systems using a new GIS add-in for supporting criterion weight elicitation methods. Operational Research, 10(1), 91e108. http://dx.doi.org/10.1007/s12351-009-0055-5. Azarnivand, A., Hashemi-Madani, F. S., & Banihabib, M. E. (2015). Extended fuzzy analytic hierarchy process approach in water and environmental management (case study: Lake Urmia Basin, Iran). Environmental Earth Sciences, 73(1), 13e26. http://dx.doi.org/10.1007/s12665-014-3391-6. Azizi, A., Malekmohammadi, B., Jafari, H. R., Nasiri, H., & Amini Parsa, V. (2014). Land suitability assessment for wind power plant site selection using ANP-DEMATEL in a GIS environment: Case study of Ardabil province,Iran. Environmental Monitoring and Assessment, 186(10), 6695e6709. http://dx.doi.org/10.1007/ s10661-014-3883-6. Bathrellos, G. D., Gaki-Papanastassiou, K., Skilodimou, H. D., Skianis, G. A., & Chousianitis, K. G. (2013). Assessment of rural community and agricultural development using geomorphologicalegeological factors and GIS in the Trikala prefecture (Central Greece). Stochastic Environmental Research and Risk Assessment, 27(2), 573e588. Beer, A., & Higgins, C. (2004). Environmental planning for site development: A manual for sustainable local planning and design. Routledge. Beinat, E., & Nijkamp, P. (1998). Multicriteria analysis for land-use management (Vol. 9). Springer. Boroushaki, S., & Malczewski, J. (2008). Implementing an extension of the analytical hierarchy process using ordered weighted averaging operators with fuzzy quantifiers in ArcGIS. Computers & Geosciences, 34(4), 399e410. http:// dx.doi.org/10.1016/j.cageo.2007.04.003. Cai, X., McKinney, D. C., & Rosegrant, M. W. (2003). Sustainability analysis for irrigation water management in the Aral Sea region. Agricultural Systems, 76(3), 1043e1066. Chen, K., Blong, R., & Jacobson, C. (2001). MCE-RISK: Integrating multicriteria evaluation and GIS for risk decision-making in natural hazards. Environmental Modelling & Software, 16(4), 387e397.

12

H. Nouri et al. / Applied Geography 81 (2017) 1e12

Chen, J., Zhang, X., & Zhu, Q. (2011). Multi-objective decision making for land use planning with ordered weighted averaging method. Systems Engineering Procedia, 2, 434e440. http://dx.doi.org/10.1016/j.sepro.2011.10.063. Conte, F. P. (1995). The aral sea: A saline lake in transition. GeoJournal, 35(1), 33e34. Dehghanzadeh, R., Safavy Hir, N., Shamsy Sis, J., & Taghipour, H. (2015). Integrated assessment of spatial and temporal variations of groundwater quality in the eastern area of Urmia salt Lake basin using multivariate statistical analysis. Water Resources Management, 29(4), 1351e1364. http://dx.doi.org/10.1007/ s11269-014-0877-7. Fathian, F., Morid, S., & Kahya, E. (2014). Identification of trends in hydrological and climatic variables in Urmia Lake basin, Iran. Theoretical and Applied Climatology, 119(3e4), 443e464. http://dx.doi.org/10.1007/s00704-014-1120-4. Fischer, M. M., & Nijkamp, P. (1993). Geographic information systems, spatial modelling and policy evaluation. Springer. Fontes, S. B., & Pejon, O. J. (2008). Proposal of a geo-environmental zoning method based on Ottobasin compartmentalization. Bulletin of Engineering Geology and the Environment, 67(4), 555e563. Gaybullaev, B., Chen, S.-C., & Gaybullaev, G. (2014). The large aral sea water balance: A future prospective of the large aral sea depending on water volume alteration. Carbonates and evaporites, 29(2), 211e219. Ghaheri, M., Baghal, M., & Naziri, J. (1999). Lake Urmia, Iran: A summary review. International Journal of Salt Lake Research, 8(1), 19e22. Ghorbani-Aghdam, M., Dinpashoh, Y., & Mostafaeipour, A. (2013). Application of factor analysis in defining drought prone areas in Lake Urmia Basin. Natural Hazards, 69(1), 267e277. http://dx.doi.org/10.1007/s11069-013-0706-y. Glantz, M. H. (2005). Water, climate, and development issues in the Amu darya basin. Mitigation and Adaptation Strategies for Global Change, 10(1), 23e50. Golley, F. B., & Bellot, J. (1999). Rural planning from an environmental systems perspective. Springer. Hassanzadeh, E., Zarghami, M., & Hassanzadeh, Y. (2012). Determining the main factors in declining the Urmia Lake level by using system dynamics modeling. Water Resources Management, 26(1), 129e145. http://dx.doi.org/10.1007/ s11269-011-9909-8. Iran, S. C. o. (2011). National census of 2011: West Azarbaijan Province. Tehran: SCI Press. Jelokhani-Niaraki, M., & Malczewski, J. (2015). A group multicriteria spatial decision support system for parking site selection problem: A case study. Land Use Policy, 42, 492e508. http://dx.doi.org/10.1016/j.landusepol.2014.09.003. Kakahaji, H., Banadaki, H., Kakahaji, A., & Kakahaji, A. (2013). Prediction of Urmia Lake water-level fluctuations by using analytical, linear statistic and intelligent methods. Water Resources Management, 27(13), 4469e4492. http://dx.doi.org/ 10.1007/s11269-013-0420-2. Karakas¸, A. (2013). Defining the suitability of new crushed rock aggregate source areas in the North of Kocaeli Province using GIS. Bulletin of Engineering Geology and the Environment, 73(4), 1183e1197. http://dx.doi.org/10.1007/s10064-0130557-5. Karbassi, A., Bidhendi, G. N., Pejman, A., & Bidhendi, M. E. (2010). Environmental impacts of desalination on the ecology of Lake Urmia. Journal of Great Lakes Research, 36(3), 419e424. http://dx.doi.org/10.1016/j.jglr.2010.06.004. Kravtsova, V., & Tarasenko, T. (2010). Space monitoring of aral sea degradation. Water Resources, 37(3), 285e296. Lafortezza, R., Chen, J., Sanesi, G., & Crow, T. R. (2008). In R. Lafortezza, J. Chen, G. Sanesi, & T. R. Crow (Eds.), Patterns and processes in forest landscapes. Patterns and processes in forest landscapes. Berlin: Springer, 2008. ISBN: 978-1-40208503-1, 1. Lein, J. K. (2003). Integrated environmental planning. Blackwell Science Ltd. Levintanus, A. (1992). Saving the aral sea. Environmentalist, 12(2), 85e91. Lioubimtseva, E. (2015). A multi-scale assessment of human vulnerability to climate change in the Aral Sea Basin. Environmental Earth Sciences, 73(2), 719e729. Madani, K. (2014). Water management in Iran: What is causing the looming crisis? Journal of Environmental Studies and Sciences, 4(4), 315e328. http://dx.doi.org/ 10.1007/s13412-014-0182-z. Malczewski, J. (1999). GIS and multicriteria decision analysis. John Wiley & Sons. Malczewski, J. (2000). On the use of weighted linear combination method in GIS: Common and best practice approaches. Transactions in GIS, 4(1), 5e22. Malczewski, J. (2002). Fuzzy screening for land suitability analysis. Geographical and Environmental Modelling, 6(1), 27e39. http://dx.doi.org/10.1080/ 13615930220127279. Malczewski, J. (2004). GIS-based land-use suitability analysis: A critical overview. Progress in planning, 62(1), 3e65. Malczewski, J. (2006). Ordered weighted averaging with fuzzy quantifiers: GISbased multicriteria evaluation for land-use suitability analysis. International Journal of Applied Earth Observation and Geoinformation, 8(4), 270e277. http://

dx.doi.org/10.1016/j.jag.2006.01.003. Malczewski, J., & Rinner, C. (2005). Exploring multicriteria decision strategies in GIS with linguistic quantifiers: A case study of residential quality evaluation. Journal of Geographical Systems, 7(2), 249e268. http://dx.doi.org/10.1007/s10109-0050159-2. Malczewski, J., & Rinner, C. (2015). Multicriteria decision analysis in geographic information science. Springer. Manaffar, R., Zare, S., Agh, N., Siyabgodsi, A., Soltanian, S., Mees, F., … Van Stappen, G. (2011). Sediment cores from Lake Urmia (Iran) suggest the inhabitation by parthenogenetic Artemia around 5,000 years ago. Hydrobiologia, 671(1), 65e74. http://dx.doi.org/10.1007/s10750-011-0704-6. Marjani, A., & Jamali, M. (2014). Role of exchange flow in salt water balance of Urmia Lake. Dynamics of Atmospheres and Oceans, 65, 1e16. http://dx.doi.org/ 10.1016/j.dynatmoce.2013.10.001. Marsh, W. M. (1998). Landscape Planning: Environmental applications. Wiley. Morimoto, Y., Natuhara, Y., Morimura, A., & Horikawa, M. (2005). The pelican scenario for nature restoration of Aral Sea wetland ecosystems. Landscape and Ecological Engineering, 1(1), 85e92. Noury, M., Sedghi, H., Babazedeh, H., & Fahmi, H. (2014). Urmia lake water level fluctuation hydro informatics modeling using support vector machine and conjunction of wavelet and neural network. Water Resources, 41(3), 261e269. http://dx.doi.org/10.1134/s0097807814030129. Ormsby, T., Napoleon, E., Burke, R., Groessl, C., & Bowden, L. (2010). Getting to know ArcGIS desktop. Esri Press Redlands. Qiu, F., Chastain, B., Zhou, Y., Zhang, C., & Sridharan, H. (2014). Modeling land suitability/capability using fuzzy evaluation. GeoJournal, 79(2), 167e182. Rafikov, V., & Gulnora, M. (2014). Forecasting changes of hydrological and hydrochemical conditions in the Aral Sea. Geodesy and Geodynamics, 5(3), 55e58. Rifikov, V., Rifikova, N., & Mamadganova, G. (2014). Forecasting changes of arid geosystems under ecological destabilizing conditions in the Aral Sea region. Geodesy and Geodynamics, 5(2), 16e23. -Riveira, I., Crecente-Maseda, R., & Miranda-Barro s, D. (2008). GIS-based Sante planning support system for rural land-use allocation. Computers and Electronics in Agriculture, 63(2), 257e273. http://dx.doi.org/10.1016/ j.compag.2008.03.007. Saxena, H. (2004). Environmental geography (2nd ed.). New Delhi: RAWAT. Shi, Y., Wang, H., & Yin, C. (2013). Evaluation method of urban land population carrying capacity based on GISda case of Shanghai, China. Computers, Environment and Urban Systems, 39, 27e38. http://dx.doi.org/10.1016/ j.compenvurbsys.2013.02.002. Swetnam, R. D., Fisher, B., Mbilinyi, B. P., Munishi, P. K., Willcock, S., Ricketts, T., … Lewis, S. L. (2011). Mapping socio-economic scenarios of land cover change: A GIS method to enable ecosystem service modelling. Journal of Environmental Management, 92(3), 563e574. http://dx.doi.org/10.1016/ j.jenvman.2010.09.007. TMU & CIWP. (2012). Agriculture and agricultural water allocation of lake Urmia basin in drought situations. Tourian, M. J., Elmi, O., Chen, Q., Devaraju, B., Roohi, S., & Sneeuw, N. (2015). A spaceborne multisensor approach to monitor the desiccation of Lake Urmia in Iran. Remote Sensing of Environment, 156, 349e360. http://dx.doi.org/10.1016/ j.rse.2014.10.006. UNDP/GEF. (2008). An Independent Review : The Status of Water Resources in the Lake Urmia Basin. Retrieved from:. UNEP & GEAS. (2012). The drying of Iran's lake Urmia and its environmental consequences. Environmental Development, 2, 128e137. http://dx.doi.org/10.1016/ j.envdev.2012.03.011. UNEP & GEAS. (2014). The future of the aral sea lies in transboundary co-operation Article reproduced from united nations environment programme (UNEP) global environmental Alert service (GEAS). Environmental Development, 10, 120e128. Zadeh, L. A. (1965). Fuzzy sets. Information and control, 8(3), 338e353. Zarghami, M. (2011). Effective watershed management; case study of Urmia Lake, Iran. Lake and Reservoir Management, 27(1), 87e94. http://dx.doi.org/10.1080/ 07438141.2010.541327. Zeinoddini, M., Bakhtiari, A., & Ehteshami, M. (2014). Long-term impacts from damming and water level manipulation on flow and salinity regimes in Lake Urmia, Iran. Water and Environment Journal. http://dx.doi.org/10.1111/wej.12087. Zhang, X., Fang, C., Wang, Z., & Ma, H. (2013). Urban construction land suitability evaluation based on improved multi-criteria evaluation based on GIS (MCEGIS): Case of New Hefei City, China. Chinese Geographical Science, 23(6), 740e753. Zhang, C., Li, W., & Day, M. (2006). Towards rationalizing protected-area designation in China using a web-based spatial decision support system. Journal of Spatial Science, 51(2), 33e46.