Open access peer-reviewed chapter

Mapping the Land-Use Suitability for Urban Sprawl Using Remote Sensing and GIS Under Different Scenarios

Written By

Onur Şatir

Submitted: November 12th, 2015 Reviewed: March 14th, 2016 Published: September 28th, 2016

DOI: 10.5772/63051

Chapter metrics overview

2,437 Chapter Downloads

View Full Metrics


Urbanization is one of the important issues in fast developing countries, such as China, Turkey, Brazil, and South Africa. Therefore, sustainable urbanization strategies come into question while designing the cities. In this point, land-use suitability mapping for urban areas is of importance. Spatial information sciences, such as geographical information systems (GIS) and remote sensing are applied widely for mapping land-use suitability. In this study, Van City, which is the most crowded city in eastern Turkey, was evaluated by applying three different scenarios called ecological, economic, and sustainable. The multi-criteria evaluation technique was used in GIS environment in the mapping stage. Distance from roads, distance from urban boundary, hillshade, slope, elevation, land-use cover, and land-use ability factors were used as inputs in the analysis stage. The weights of each input factor were calculated according to urban change dynamics between 2002 and 2015. As a result of the study, the weighting approach using the natural change dynamics of Van City has a great potential to define objective weights. In addition, Van City was developed orderly on agricultural lands and grasslands, and it was not a sustainable development for the region because the main income is still agriculture and animal production, so a new strategy was designed in a sustainable scenario to prevent agriculture and grassland area loss in a mutual benefit between nature and human.


  • multi-criteria evaluation
  • urban land-use suitability
  • fuzzy standardization
  • ideal point-based weighting
  • remote sensing
  • geographical information system
  • Van City
  • Turkey

1. Introduction

Human and nature are always in a paradox after the industrial revolution. Human requirements, such as food and accommodation, have increased in time because of raised human population and fast industrial developments. Particularly, fast developing countries, such as China, Turkey, Brazil, and South Africa, needed new lands for urban sprawl and industrial facilities recently. Landscape planning is important for decision makers in this point, because available lands for agriculture, urbanization, and industrial activities are limited and land-use planning and decision-making process are vital for a sustainable development.

In this content, land suitability analysis is one of the oldest forms of decision-making support systems in the field of landscape planning [1]. Nowadays, geographical information system (GIS) and remote sensing (RS) based techniques are used frequently in land-use planning to obtain land-use decisions and alternatives geographically. RS science is generally used to create layouts for spatial decision support systems (SDSS) in the analysis stage. RS provides time energy and cost savings, and huge areas can be identified using satellite images, radar and lidar technologies, and aerial photos from different platforms, such as unmanned air vehicles, towers, balloons, or planes [2]. RS can provide many data on the ecosystem physical, biological, and social dynamics to land-use planning studies. For example, topographical dynamics [3, 4], vegetation dynamics [5, 6], LUC dynamics [2, 7], soil physical, chemical, and biological dynamics [8, 9], and hydrological dynamics can be quantified using several modeling or classification techniques by RS.

RS outputs can be integrated with other raster or vector data in GIS interface to derive land-use decisions. When the decision support systems are integrated with GIS, it is called SDSS. SDSS is allowed stakeholder participation, iterative analysis, and integration with external spatial or non-spatial data sources [10]. Scientific literature has shown that SDSS is mostly used in multi-criteria evaluation (MCE) analyses. MCE is a multiple data assessment technique, and it is used not only in land-use evaluation studies but also in environmental risk probability mapping, such as landslide [11], forest fire risk [12, 13], and flood risk [14], in landscape planning science. GIS-based MCE techniques have three essential stages: criteria selection (factor definition), data standardization, and weighting [15].

“Criteria selection” is defined with respect to the study goal. All factors may be called as inputs. In this extent, this stage is important to get an accurate result. These data sets must identify the essentials of the research. For example, in an agricultural land-use suitability mapping for wheat crop, inputs are related on wheat growth environment, such as climatic variables (temperature, precipitation, solar radiation, humidity), physical variables (elevation, aspect, slope, hillshade, soil depth, soil texture), chemical variables (soil pH, soil nutrients, soil salinity), accessibility (road network, settlement location), and bio-environmental variables (chlorophyll content, crop yield). Additionally, environmental risk factors, protection areas (natural and historical sites), and restriction areas (roads, security regions, built-up areas) must be defined spatially.

Figure 1.

Sample slope standardizations using various techniques for urban sprawl suitability.

“Standardization” is necessary because of input data unit and range differences. Inputs will multiply with detected weights and standardization is a vital part for input data comparison. However, some non-parametric multiple data assessment techniques, such as artificial neural network (ANN) or regression tree, do not need standardization process because of internal weighting abilities [13]. There are many standardization techniques, such as fuzzy standardization, linear ordered standardization, and Boolean approach. Introducing these techniques with a sample is better for understanding. Imagine a slope map of a land in urban built-up suitability study. The slope unit is degree and data range is between 0 and 90°. In the fuzzy approach, flat fields defined the most suitable (1) and steep fields defined the opposite (0). Other fields are valued between 0 and 1 according to the suitability values, such as 0.1, 0.2, 0.25, etc. So that there will not be any absolute transition zone in the standardized map, all transitions will be soft surface. This technique is appropriate to avoid the missing data effect in transition zone. Ordered standardization needs categorical data. In this technique, slope data must be reclassified based on slope degrees, such as 0-5, 5-15, 15-25, 25-35, etc. Each slope range is reclassified as 1, 2, and 3 based on slope suitability for urban built-up. In this method, there is no transition zone and the boundaries of the categories are hard. In Boolean approach, all inputs are standardized as only 1 and 0 or suitable or not. There is a threshold value to detect the suitable land for each factor. For example, if slope degree less than 15 is suitable, other areas are not suitable for urban built-up. Slope data standardization using these techniques is shown in Figure 1.

“Weighting” is the hardest part of an MCE analysis. Here, expert-based weighting, literature-based weighting, and ideal point-based linear weighting are discussed. In addition, there are some non-linear ideal data-dependent techniques, such as ANN, logistic regression, ant colony algorithm, and regression tree.

Expert-based technique is preferred by those who have not any ideal point data, because, in this technique, all factors are evaluated by the experts, applying a survey that includes questions about the priority of the factors. Analytical hierarchical process (AHP) can be used because of pairwise comparison abilities to detect the weights of factors after the expert’s decisions. Pairwise comparison matrix is defined as weights using binary priority definition ability [16, 17]. Although expert-based weighting is easy, all experts may be given various answers to the same question. Thus, the subjectivity of this method is high. However, this method is still used widely [18].

Literature-based weighting is another approach to define factor weights. It is completely based on similar studies in the literature and factor weights are adapted to the new research from previous studies. This method is easy for weight definition; however, regional environmental and social differences are ignored in this approach, so the reliability of the technique is a question.

Ideal data-based weighting is the most reliable approach in MCE studies. The problem in this technique is that what is the ideal data. For example, in a crop-based agricultural land-use suitability detection study, crop productivity can be used to define ideal crop suitability areas and all factors can be weighted relating to high productive areas to find potential suitable areas. Crop productivity data can be obtained from farmer surveys or using RS [13]. There is another technique using ideal point for weighting called the weight of evidence (WOE). This approach is also effective to define factor weights in land-use suitability or environmental hazard probability studies [19, 20].

This chapter discusses land-use suitability for urban growth using ideal point-based technique. Ideal urban sprawl is defined using urban growth dynamics from past to current. In this extent, Van City, which is located in eastern Turkey, was modeled by applying three different scenarios: economic, ecological, and sustainable. Distance from road and central city area, elevation, slope, hillshade, land-use ability (LUA), and land-use cover (LUC) were used as factors in the study. Weights were defined according to urban change and were used in all scenarios. On the contrary, restriction areas and fuzzy suitability degrees of LUC and LUA data were modified separately for the scenarios.


2. Short background: GIS-based land-use suitability studies

While factor and objectives are spatial, a GIS interface is necessary to analyze geographic data, and it requires a combination of multi-criteria methods with a GIS interface [21, 22]. In the literature, there are many land-use decision studies using GIS and other spatial information science, such as RS. Malczewski [22] completed a detailed study on the survey of the GIS-based MCE techniques. Results showed that this method is used widely (top three subjects) and orderly in environment and ecology, transportation, and urban-rural planning studies. In addition, these techniques are used in waste management, agriculture and forestry, facility area suitability, rangeland management, recreation and tourism, natural hazard studies, hydrological studies, and real estate-housing studies. Some techniques, data, and criteria used for land-use suitability analyses are found in Table 1.

Source Technique Factors Suitability
[31] GIS-based ANN Soil depth, moisture,
fertility, texture, salinity,
aeration, temperature, accessibility
Agriculture 1994
[32] Boolean overlay
Slope, soil mechanics,
flood-erosion hazard, water level
and drainage,
toxicities, rooting conditions
Crop growth
[33] AHP Soil depth, LUA, erosion
hazard, slope,
elevation, distance from
water sources,
distance from road,
limiting soil factors
Land use 2009
[34] GIS-AHP Population density,
available land, land values
Public parks 2011
[35] GIS-AHP expert-
based weighting
LUC, DEM, slope,
tourist map, attraction places,
road map, population, protected areas,
wildlife areas
Tourism and
Accessibility, slope,
interactions with other
facilities, population density
[15] RS integrated
fuzzy, GIS-AHP,
Temperature, precipitation,
soil texture, soil salinity (EC), soil
depth, soil porosity, GDD,
crop productivity
Crop growth
[18] AHP, expert-
based weighting
Soil groups, soil depth,
erosion, slope,
aspect, elevation, soil
parameters, LUA
Agriculture 2013
[37] RS integrated
GIS, expert-
based weighting
using AHP
Slope, soil depth,
texture, moisture, organic C,
pH, EC, primary
nutrients (N, P, K)
Agriculture 2015
[30] GIS-MCE-
fuzzy AHP
Residential, extractive industry,
marine industry,
recreation subfactors
Urban land

[38] RS integrated
GIS-based MCE,
NDVI, LUC, climate data set,
DEM, economic and
social data
[39] Fuzzy logic,
Distance from river, LUC,
urban areas,
crop pattern, distance from
residential areas, roads,
proximity to interested
Waste water
area suitability

Table 1.

Sample techniques and factors used in land-use suitability studies.

EC, electrical conductivity; GDD, growing degree days; NDVI, normalized difference vegetation index.

According to Table 1, even if the suitability target is same, there might be small differences between factors. Some of the important essential factors are the same in agricultural suitability studies, such as soil depth, slope, and soil texture. However, there are different factors: erosion rate, soil nutrients, and crop productivity. Data accessibility, database availability, and method can be affected factors despite the same purpose. Also, these factors may be changed with respect to regional differences. If there is flood risk in an area, we have to assess flood risk rate in land-use suitability analyses.


3. Study area

Van City is located in eastern Turkey, and it is the most crowded city of the Eastern Anatolian Region of Turkey (Figure 2). Van Province’s population was 1,096,397 in 2015, and almost 472,000 of them lived in Van central city area. Population has increased almost 5.5% during the last 5 years [23]. The main incomes in the region are animal production, agriculture, border trading, and tourism. Mainly continental climate is dominant, but alpine and sub-alpine climate effects are observed in high regions. Van central city’s area climate is warmer than in those around it because of Van Lake. This lake is the biggest lake in Turkey and its water contains CaCO3. The central city area of Van was selected as the study area due to the fast development in last 10 years. The city was damaged by two big earthquakes (7.2 and 6.1) in 2011. After these tragic events, urban area started to sprawl to the far regions from the central city area.

Figure 2.

Location of the study area and boundary of the case area (UTM WGS 84 Projection Zone 38N).


4. Data characteristics

The research data set contains (i) satellite images and LUC data, (ii) digital elevation model (DEM) data, (iii) LUA data, (iv) distance from road and urban built-up area data, and (v) restrictive areas or constrained areas (Figure 3).

4.1. Satellite images

Landsat 5 TM and Landsat 8 OLI data sets were used to detect urban area change between 2002 and 2015 in Van central city area. Landsat imageries have a great potential for monitoring the land-use/cover change because of large time series database and available spatial (30 m) and spectral (VIS, NIR, SWIR, and TIR) resolutions [24]. Two Landsat images were used for LUC and change detection. An earlier image was taken on August 2002, and the later one was recorded on August 2015.

Figure 3.

Input data set (criteria or factors) used in the urban sprawl suitability of Van City.

4.2. DEM data

Elevation data were obtained from ASTER Global DEM data set in 30 m spatial resolution. Slope and hillshade images were produced from DEM data in GIS interface. Additionally, DEM data were used as ancillary data in LUC classification stage to detect the LUC change based on elevation. Also, such data sets may be improved for the accuracy of the LUC classification [2].

4.3. LUA data

The LUA of Turkey was mapped by Soil Survey Staff of Turkey. This map shows capable lands and capability degrees in eight categories (1 refers to very capable and 8 refers to not capable) based on slope, soil depth, soil type, geological type, etc. Highly available lands can be used for agriculture, urban built-up, and industrial facilities. However, land-use suitability is variable according to the scenarios. For example, agricultural areas are important because available lands are limited so, in ecological or sustainable scenarios, highly available lands must be protected for agriculture.

4.4. Distance images

Distance from roads and urban built-up areas are produced from road maps and urban area maps, which are produced from classified LUC map. Road and settlement distance is significant to evaluate the infrastructure and superstructure availability of a land for urban sprawl.

4.5. Restrictive areas

Urban cannot sprawl to regions, such as water bodies, existing built-up areas, security areas, and historical protection areas. Also, restrictive areas are modified according to the scenarios. Some wetlands are closed for urban sprawl in the ecological scenario.


5. Methodology

The study was performed in two substages: (i) LUC classification, change detection, and accuracy assessment and (ii) MCE process in GIS environment (Figure 4).

Figure 4.

Summary of the methodology.

5.1. Object-based classification

Many complex land covers exhibit similar spectral characteristics, making separation in feature space by simple per-pixel classifiers difficult, leading to inaccurate classification. Therefore, an object-based classification is a potential solution for the classification of such regions. The specific benefits are an increase in accuracy, a decrease in classification time, and that it helps to eliminate within-field spectral mixing [25]. Basically, there are three steps in object-based classification: segmentation, classification, and per field integration. An image is divided into segments dependent on pixel spectral similarities, structure of the image, and surface texture characteristics. This progress is up to variables such as scaling factor, smoothness versus compactness, and shape factors. Each segment contained a group of pixels, and scaling factor is defined as the minimum pixel counts that have similar spectral characteristics in a segment. Compactness and smoothness are important for creating pixel groups. Shape factor deals with the boundary of a segment. Scale factor is variable according to the study scale and ideal scale can be found trying different scale factors [2].

5.2. Accuracy assessment of LUC classifications

The accuracy of the LUC classifications is tested by applying error matrix and κ statistic. The error matrix approach is the most widely used in accuracy assessment [26]. After the generation of an error matrix, other important accuracy assessment elements, such as overall accuracy, user accuracy, producer accuracy, and κ coefficient, can be derived. κ is the difference between the observed accuracy and the chance agreement divided by 1 minus that chance agreement [27]. Ground truth data collected from field trip and old topographical and forest maps are used for the accuracy assessment of 2002 and 2015 LUC.

5.3. GIS-based MCE process

The general information on this process is discussed in Section 1. However, criteria selection, standardization, weighting, and mapping progress and methods are introduced in urban sprawl suitability extent in this study. Factors are defined based on similar literature, data accessibility, and regional variations. Seven factors are defined: elevation, slope, hillshade, LUC, LUA, distance from roads, and distance from built-up areas.

Weighted linear combination (WLC) is one of the most used MCE techniques to define suitability degrees for continuous factors. With WLC, factors are combined by applying a weight to each followed by a summation of the results to produce a suitability map [28]:

S = W i X i , E1

where S is suitability, Wi is the weight of factor i, and Xi is the criteria score of factor i.

5.3.1. Fuzzy approach and weighting

Nothing is exact in nature. Fuzzy mentality is based on this simple rule. In a landscape, boundaries are flexible and natural characteristics of a land do not change suddenly, such as slope and soil texture. A fuzzy set can define a factor with transition zones according to the suitability degree of the factor [29].

Factors are of different data ranges and units. Weighting a factor linearly needs a standard input data set to define the priority degree of the factors. Standardization process is applied to the factors separately. However, ideal and non-ideal values of the each factor have to be defined primarily. In this study, ideal regions for urban sprawl are defined based on urban development between 2002 and 2015. Changed regions are accepted ideal areas for urban sprawl. The ideal and non-ideal values of each factor are defined considering the existing urban growth. It does not mean that the existing urban growth is ideal for sustainable development, but, finally, Urban was developed to these regions and the spatial characteristics of the changed areas are good indicators for the future urban sprawl. All factors are reclassified based on ideal intervals, which are defined according to data frequencies, and ideal data ranges are found for fuzzy standardization.

The weights of each factor are obtained using ideal data diversity in the categories of the factors. For example, slope is reclassified to five intervals: very high, high, medium, low, and very low. If urban change is located in all areas homogeneously, it means that the slope is not important for urban sprawl; if urban change is monitored heterogeneously, it means that the slope is important because only one or two slope ranges are suitable for urban growth. The standard deviation (SD) of the each factor diversity is shown as data homogeneity with a standard result. If the SD value is high, weight should be high. The weights of each factor are defined according to the SD values.

5.3.2. Scenario development

Three scenarios were evaluated in the study: economic, ecologic and sustainable. In the economic scenario, only water bodies, existing city areas, security areas, and historical protection areas were analyzed and defined as restrictive regions. There was not any limitation on LUC or LUA. Therefore, the trend of the last 13 years was considered to define urban suitability in the economic scenario. Ecologically important areas, such as wetlands, were ignored.

In the ecologic scenario, important areas, such as wetlands (bulrush) and highly available lands for agriculture and grasslands, were protected strictly from urban sprawl.

In the sustainable scenario, urbanization demands were considered using urban change in time. According to urban change results, agricultural and wetland areas were protected partly in addition to other restrictive areas, so that a balanced urban spread was provided for human life and nature in the future.


6. Results

In this study, urban change in time has the key role for standardization and weighting of the factors. Therefore, RS science is into play in this point to create LUC layouts and define urban change. Then, a GIS-based MCE process is applied according to the changed areas.

6.1. Urban change detection by RS

Landsat imageries taken on August 2002 and 2015 were classified using object-based classification approach. A post-classification change detection technique was applied and atmospheric or radiometric correction is not necessary in this technique. Also, Landsat imageries have already been corrected geometrically. Segmentation parameters were modified after a small experimental application. Seven LUCs were classified, except Van lake area: agriculture, grassland, bareground, settlement, bulrush, woodland, and inland water.

August 2002 and 2015 overall κ classification coefficiencies were obtained as 0.89 and 0.92, respectively. The urban built-up area was 2066 ha in 2002 and increased to 7694 ha in 2015. Then, 3841 ha agriculture, 1265 ha grassland, 328 ha bareground, 89 ha woodland, and 18 ha bulrush areas were transformed to the urban. Particularly, after the 2011 earthquakes, urban growth system was changed and new residential areas were established on the far regions from the city center. This was affected urban built-up area change fast (Figure 5).

Figure 5.

LUC of Van central city area (2015–2002) and changes in urban built-up areas.

6.2. Fuzzy standardization and factor weighting

Factors were reclassified based on natural breaks in ideal intervals. LUC and LUA data sets were already categorical. However, other inputs were continuous data format. Elevation data were reclassified into eight categories with 100 m intervals. Slope was divided into five categories with 15° intervals as very low (flat), low, medium, high, and very high. Hillshade was reclassified into five categories based on sun effect as very high, high, medium, low, and very low. Distance from roads and distance from urban built-up areas were classified as 10 intervals for each 1000 m. Urban area change was recorded as 5628 ha and all changed areas were separated according to urban change areal diversity to each category for each factor (Table 2).

Categories 1 2 3 4 5 6 7 8 9 10
Elevation* 4169 1188 161 76 26 19 13 0
Slope* 5631 20 0 0 0
Hillshade* 1 9 496 5109 37
Distance from road* 4293 863 177 124 6 39 91 14 17 4
Distance from built-up areas* 4367 532 257 168 137 62 39 33 28 5
LUA** 1783 973 1675 258 21 523 257 136
LUC*** 3841 1265 328 89 74

Table 2.

Areal diversity of urban change areas to each category inside the factors.

*(1) Refers lowest value range of the factor.

**(1) Refers the highest value of the factor.

***(1) Agriculture, (2) grassland, (3) bareground, (4) woodland, and (5) bulrush.

All factors must be evaluated alone based on the suitability degree for fuzzy standardization. In this frame, following fuzzy standardization functions were applied to the factors and all inputs were standardized between 0 and 1 (Table 3; Figure 6).

Categories Technique Function Explanation
Elevation Monotonically decreasing Almost linear When elevation is increased, urban suitability is decreased
Slope Monotonically decreasing User defined When slope degree between 0 and 15 suitability is decreasing polynomial, urban growth is not possible after the 35th degree
Hillshade Optimal value Gaussian or symmetric in narrow range Sun effect should be high but not very high or very low
Distance from road Monotonically decreasing Almost linear If an area far from the road network, suitability of urban sprawl is decreased
Distance from built-up areas Monotonically decreasing Almost linear If an area far from the road network, suitability of urban sprawl is decreased
LUA Optimal values User defined Suitability degree of each category was defined based on urban growth distribution from past to current

Table 3.

Fuzzy standardization rules of the factor for urban built-up suitability.

Figure 6.

Standardized factors by fuzzy approach for urban sprawl suitability.

Weighting the factors was performed according to the SD of urban change in each factor. Areal diversity of urban change was already obtained for each category of factors. SDs were used to see the heterogeneity of urban change diversity in the factors. The SD values and weights are shown in Table 4.

Categories SD Weights
Elevation 1456 0.131
Slope 2516 0.227
Hillshade 2234 0.2
Distance from road 1336 0.12
Distance from built-up areas 1346 0.121
LUA 699 0.063
LUC 1498 0.135

Table 4.

Weights of factors.

As a result of the weights, slope and hillshade were the most significant factors for urban sprawl. The lowest effect was recorded in LUA because the urbanization aspects from past to current were not the care of LUA for urban sprawl. However, these results showed that urban sprawl in time was not sustainable if this situation continued, because city growth on the agricultural areas that are limited in the region and there is no too much option for income, except agriculture and tourism in the region.

6.3. Scenario applications

Scenarios are described in Section 5.3.2. In the application stage, there are several approaches to apply a scenario. Some of these approaches were applied modifying the weights of each factor for each scenario. However, in this study, all factor weights have been already obtained based on urban change, and physical variables, such as hillshade, elevation, slope, and distance factors, are not changed according to the our scenarios. Therefore, the weights of these data were used the same in all scenarios. However, LUC and LUA suitability can be changed according to the economic, ecological, and sustainable scenarios.

In the economic scenario, the fuzzy standardization of the LUC and LUA was defined without any limitation. For example, agricultural areas were transformed to urban areas in 13 years dominantly. In the economic scenario, this situation was continued; in the ecological scenario, some of the agricultural areas were protected, which were located in the first, second, and third zones of the LUA. Also, in the ecological scenario, wetland, coastal line, and nearby and natural grassland usage was limited for urban sprawl.

In the sustainable scenario, only the first zone of the LUA was protected because, if first three zones were protected, there would not be enough urban sprawl area in the future and this situation would not be sustainable.

6.3.1. Urban sprawl suitability in the economic scenario

Restrictive areas were defined as historical protection areas, current built-up areas, water surfaces (lake and rivers), and security zones (military security areas). There was no limitation on LUC and LUA or wetland usage in the area. In this scenario, urban change was presumed continuous in the future like in the last 13 years.

Suitability degrees were classified into five categories and constrained areas. In the economic scenario, particularly, medium and high suitability regions covered 35% and 32%, respectively. Restrictive areas only covered 12% of the total area (Figure 7).

Figure 7.

Distribution of the urban sprawl suitability under the economic scenario.

6.3.2. Urban sprawl suitability in the ecological scenario

This scenario was protected in all important natural lands and high capable lands for agricultural usage strictly. In this extent, the LUC map of 2015 and LUA maps were re-standardized and the lowest suitability degree was assigned to agriculture after the bulrush (wetlands) areas in the LUC map. In the LUA map, the first, second, and third zones were extracted from the result map, because these lands are available for agriculture.

Medium suitable lands and constrained areas covered 41% and 36%, respectively. High suitable lands were located in the north and east sides of the city mainly (Figure 8).

Figure 8.

Distribution of the urban sprawl suitability under the ecological scenario.

6.3.3. Urban sprawl suitability in the sustainable scenario

This scenario is a mixture of the economic and ecological scenarios. Restrictive areas were less than the ecological scenario, because only the ecological scenario is not applicable and sustainable for a fast developing city like Van City. However, natural lands and coastal regions that have a good recreational ability were ignored in the economic scenario. Additionally, most of the agricultural areas were suitable for urban growth under the economic point of view because of construction expenses (less filing process and infrastructure necessity). Therefore, we have to both protect important areas for high-quality life in the future and answer the future urban built-up demand. Therefore, the sustainable scenario was considered in both variables.

Medium and high suitable areas covered 43% and 28% of the total area, respectively, in the sustainable scenario. Restrictive areas covered 19%. Agricultural areas were protected partly according to the suitability degree in the LUA map. Only the first zone was extracted from the result map, and the fuzzy suitability map of the LUA was defined based on LUA stages orderly in a negative way. Urban sprawl was planned on less productive agricultural areas without ignoring the urban sprawl on agricultural areas (Figure 9).

Figure 9.

Distribution of the urban sprawl suitability under the sustainable scenario.


7. Discussion and conclusions

Urban sprawl suitability under the various scenarios was applied to Van City, which developed fast in the last decade in Turkey. The purpose of the study was not only to focus on the Van City growth options but also to present how we can apply various scenarios in GIS interface for a land-use decision-making process using an objective weighting technique based on an indicator data set such as urban change in time.

Suitability degree Economic scenario Ecological scenario Sustainable scenario
Area (ha) % Area (ha) % Area (ha) %
Constrained 8910 12 25,814 36 13,597 19
Very low 425 1 125 <1 62 <1
Low 4220 6 4125 6 2617 4
Medium 25,032 35 29,434 41 31,356 43
High 23,047 32 11,243 16 19,860 28
Very high 10,505 15 1406 2 4613 6

Table 5.

Areal diversity of the urban suitability categories of scenarios.

Three different scenarios were cross-checked as economic, ecologic, and sustainable using seven factors, which have significant effects on urban sprawl. These factors were diversified according to the regional differences or data accessibilities for urban growth suitability [30]. As a result of the study, restrictive areas covered too much places in the ecological scenario. The highest suitability areal cover was recorded in the economic scenario. On the contrary, in the sustainable scenario, development areas were defined ideally for the future urban sprawl of Van City. Van City changed almost 5700 ha in 13 years, and it needs a minimum of 5000 ha area for the next 10 years according to the population increase. In the sustainable scenario, very high suitable lands almost satisfied this demand. Also, high suitable areas were enough for the long future development (Table 5).

In general, very high and high suitable areas were located in north, east, and south parts of the city. In the city development plan, a belt highway is under construction from the north to south parts of the city shaped as a half-circle. In this extent, heavy industrial places are located at the north side of the city and it will improve to near the belt highway in the northern side. These regions are available for industrial development. The south part of the city started to develop after the earthquakes in 2011. According to the ground surveys (ground stability map could not be used in this study because of data accessibility problem), this is available for urban sprawl without taking any additional construction preventions for built-up area growth. The east part of the city developed so fast after the 2000s. Particularly, immigrants from the rural areas and other cities settled in the east part of the city. Unfortunately, the east part is a good sample for unplanned urbanization, so this part will probably be within the scope of urban transformation in future. Therefore, urban sprawl in this region will be less than expected.

In conclusion, GIS-based MCE techniques are efficient to evaluate the land-use options because of geographic interface and scenario application abilities. Also, this visual ability can be integrated with public participation in weighting or decision-making stages. In this study, RS data were used to define fuzzy rules and weights of factors using a categorical diversity of changed areas in the factors. This approach was provided objective weighting by an ideal data set. This ideal data set was not an ideal city development, but finally the city developed on these areas in time and priority of the factors can be organized according to past urban development. For the ideal urban growth, scenarios were tested. Sustainable urban sprawl areas were detected comparing the ecological and economic scenarios and urbanization demand in the future. As a result of the decisions, a sustainable scenario was developed and applied to Van City considering sustainable life quality, urbanization demands, and nature protection.



  1. 1. Malczewski J: GIS-based land-use suitability analysis: A critical overview. Progress in Planning. 2004; 62(1): 3–65.
  2. 2. Şatır O, Berberoğlu S: Land use/cover classification techniques using optical remotely sensed data in landscape planning. In Özyavuz M. (Ed.) Landscape Planning. Rijeka: InTech; 2012. pp. 21–54. DOI: 10.5772/31351.
  3. 3. Guarnieri AM: SAR interferometry and statistical topography. IEEE Transactions on Geoscience and Remote Sensing. 2002; 40(12): 2567–2581.
  4. 4. Zhanyu W, Arrowsmith JR, Honglin H: Evaluating fluvial terrace riser degradation using LIDAR-derived topography: An example from the northern Tian Shan, China. Journal of Asian Earth Sciences. 2015; DOI: 10.1016/j.jseaes.2015.02.016.
  5. 5. Berberoglu S, Satir O, Atkinson PM: Mapping percentage tree cover from Envisat MERIS data using linear and non-linear techniques. International Journal of Remote Sensing. 2009; 30(18): 4747–4766.
  6. 6. Donmez C, Berberoglu S, Curran P: Modelling the current and future spatial distribution of NPP in a Mediterranean watershed. International Journal of Applied Earth Observation and Geoinformation. 2011; 13: 336–345.
  7. 7. Akin A, Sunar F, Berberoğlu S: Urban change analysis and future growth of Istanbul. Environmental Monitoring and Assessment. 2015; 187: 506.
  8. 8. Satir O, Berberoglu S, Kapur S, Nagano T, Akça E, Erdogan MA, Donmez C, Satir NY, Tanaka K: Soil salinity mapping using Chris-Proba hyperspectral data. In Proceedings of ESA Hyperspectral Workshop 2010; 17–19 March 2010; Frascati, SP-683.
  9. 9. Lakshmi V, James J, Soundaria S, Vishalini T, Pandian KP: A comparison of soil texture distribution and soil moisture mapping of Chennai coast using Landsat ETM+ and IKONOS data. Aquatic Procedia. 2015; 4: 1452–1460.
  10. 10. Kim HY, Choi Y, Kim H, Oh SH: Planning for the sustainable? Land use suitability and social and ecological factors for locating a new hazardous facility. KSCE Journal of Civil Engineering. 2016; 20(1): 359–366.
  11. 11. Chalkias C, Ferentinou M, Polykretis C: GIS-based landslide susceptibility mapping on the Peloponnese Peninsula, Greece. Geosciences. 2014; 4: 176–190.
  12. 12. Jaiswal RK, Mukherjee S, Raju KD, Saxena R: Forest fire risk zone mapping from satellite imagery and GIS. International Journal of Applied Earth Observation Geoinformation. 2002; 4: 1–10.
  13. 13. Satir O, Berberoglu S, Donmez C: Mapping regional forest fire probability using artificial neural network model in a Mediterranean forest ecosystem. Geomatics Natural Hazards and Risk. 2015; DOI: 10.1080/19475705.2015.1084541.
  14. 14. Papaioannou G, Vasiliades L, Loukas A: Multi-criteria analysis framework for potential flood prone areas mapping. Water Resources Management 2014; 29(2): 399–418.
  15. 15. Şatır O: Determining the agricultural land use suitability using remote sensing and geographical information system in Lower Seyhan Plane (Ph.D. thesis). Cukurova University Natural and Applied Sciences Institute; 2013.
  16. 16. Saaty T. The Analytical Hierarchy Process. New York: John Wiley; 1980.
  17. 17. Saaty T: Relative measurement and its generalization in decision making: why pairwise comparisons are central in mathematics for the measurement of intangible factors e the analytic hierarchy/network process. Review of the Royal Spanish Academy of Sciences Series A Mathematics. 2008; 102(2): 251–318.
  18. 18. Akıncı H, Özalp AY, Turgut B: Agricultural land use suitability analysis using GIS and AHP technique. Computers and Electronics in Agriculture. 2013; 97: 71–82.
  19. 19. Dickson BG, Prather JW, Xu Y, Hampton HM, Aumack EN, Sisk TD: Mapping the probability of large fire occurrence in northern Arizona, USA. Landscape Ecology. 2006; 21: 747–761.
  20. 20. Zheng X, Lv L: A WOE method for urban growth boundary delination and its applications to land use planning. International Journal of Geographical Information Science. 2016; 30(4): 691–707.
  21. 21. Stewart TJ, Janssen R: A multi objective GIS-based land use planning algorithm. Computers, Environment and Urban Systems. 2014; 46: 25–34.
  22. 22. Malczewski J: GIS-based multicriteria decision analysis: A survey of the literature. International Journal of Geographical Information Science. 2006; 20: 703–726.
  23. 23. TSS. Turkey Statistical Service. Population records of the Van Province. 2016.
  24. 24. Özyavuz M, Şatır O, Bilgili BC: A change vector analysis technique to monitor land use/land cover in Yıldız Mountains, Turkey. Fresenius Environmental Bulletin. 2011; 20(5): 1190–1199.
  25. 25. Berberoglu S, Lloyd CD, Atkinson PM, Curran PJ: The integration of spectral & texture information using neural networks for land cover mapping in the Mediterranean. Computers and Geosciences. 2000; 26: 385–396.
  26. 26. Foody GM: Status of land cover classification accuracy assessment. Remote Sensing of Environment. 2002; 80: 185–201.
  27. 27. Lillesand TM, Kiefer RW. Remote Sensing and Image Interpretation. New York: Oxford University Press; 1994.
  28. 28. Voogd H. Multicriteria Evaluation for Urban and Regional Planning. London Pion, Ltd. 1983.
  29. 29. Zadeh LA. Fuzzy sets. Information and Control. 1965; 8: 338–353.
  30. 30. Mosadeghi R, Warnken J, Tomlinson R, Mirfenderesk H: Comparison of fuzzy-AHP and AHP in spatial multi-criteria decision making model for urban land-use planning. Computers, Environment and Urban Systems. 2015; 49: 54–65.
  31. 31. Wang F: The use of artificial neural networks in a geographical information system for agricultural land-suitability assessment. Environment and Planning A. 1994; 26(2): 265–284.
  32. 32. Kalogirou S: Expert systems and GIS: an application of land suitability evaluation. Computers, Environment and Urban Systems. 2002; 26: 89–112.
  33. 33. Cengiz T, Akbulak C: Application of analytical hierarchy process and geographic information systems in land-use suitability evaluation: a case study of Dumrek village. International Journal of Sustainable Development and World Ecology 2009; 16(4): 286–294.
  34. 34. Chandio IA, Matori AN, Lawal DU, Sabri S: GIS-based land suitability analysis using AHP for public parks planning in Larkana City. Modern Applied Science 2011; 5(4): 177–189.
  35. 35. Bunruamkaew K, Murayama Y: Site suitability evaluation for ecotourism using GIS&AHP: A case study of Surat Thani province, Thailand. Procedia Social and Behavioral Sciences. 2011; 21: 269–278.
  36. 36. Javadian M, Shamskooshki H, Momeni M: Application of sustainable urban development in environmental suitability analysis of educational land use by using AHP and GIS in Tehran. Procedia Engineering. 2011; 21: 72–80.
  37. 37. Zolekar RB, Bhagat VS: Multi-criteria land suitability analysis for agriculture in hilly zone: Remote sensing and GIS approach. Computers and Electronics in Agriculture. 2015; 118: 300–321.
  38. 38. Wang Y, Ding Q, Zhuang D: An eco-city evaluation method based on spatial analysis technology: A case study of Jiangsu province China. Ecological Indicators. 2015; 58: 37–46.
  39. 39. Aydi A, Abichou T, Nasr IM, Louati M, Zairi M: Assessment of land suitability for olive mill wastewater disposal site selection by integrating fuzzy logic, AHP and WLC in a GIS. Environmental Monitoring and Assessment. 2016; 188: 59.

Written By

Onur Şatir

Submitted: November 12th, 2015 Reviewed: March 14th, 2016 Published: September 28th, 2016