Environ Monit Assess (2010) 160:215227

DOI 10.1007/s10661-008-0689-4
Groundwater quality mapping in urban
groundwater using GIS
Bilgehan Nas Ali Berktay
Received: 19 September 2008 / Accepted: 20 November 2008 / Published online: 19 December 2008
Springer Science + Business Media B.V. 2008
Abstract Konya City, located in the central part
of Turkey, has grown and urbanized rapidly. A
large amount of the water requirement of Konya
City is supplied from groundwater. The quality
of this groundwater was determined by taking
samples from 177 of the wells within the study
area. The purposes of this investigation were (1)
to provide an overview of present groundwater
quality and (2) to determine spatial distribution
of groundwater quality parameters such as pH,
electrical conductivity, Cl

, SO
4
2
, hardness, and
NO
3

concentrations, and (3) to map groundwa-

ter quality in the study area by using GIS and
Geostatistics techniques. ArcGIS 9.0 and ArcGIS
Geostatistical Analyst were used for generation
of various thematic maps and ArcGIS Spatial
Analyst to produce the nal groundwater quality
map. An interpolation technique, ordinary krig-
ing, was used to obtain the spatial distribution of
groundwater quality parameters. The nal map
shows that the southwest of the city has optimum
groundwater quality, and, in general, the ground-
water quality decreases south to north of the city;
5.03% (21.51 km
2
) of the total study area is clas-
sied to be at the optimum groundwater quality
level.
B. Nas (B) A. Berktay
Department of Environmental Engineering,
Selcuk University, 42075, Konya, Turkey
e-mail: bnas@selcuk.edu.tr
Keywords GIS Geostatistics
Groundwater quality Kriging
Introduction
Groundwater is an important source of drinking
water for many people around the world, espe-
cially in rural areas. Groundwater can become
contaminated from natural sources or numerous
types of human activities. Residential, municipal,
commercial, industrial, and agricultural activities
can all affect groundwater quality. Contamination
of groundwater can result in poor drinking water
quality, loss of water supply, high cleanup costs,
high costs for alternative water supplies, and/or
potential health problems.
Natural resources and environmental con-
cerns, including groundwater, have beneted
greatly from the use of GIS. Typical examples
of GIS applications in groundwater studies are
site suitability analyses, managing site inventory
data, estimating vulnerability of groundwater to
pollution potential from nonpoint sources of pol-
lution, modeling groundwater movement, model-
ing solute transport and leaching, and integrating
groundwater quality assessment models with spa-
tial data to create spatial decision support systems
(Engel et al. 1999). Hudak (1999, 2000, 2001)
and Hudak and Sanmanee (2003) have reported
a number of studies about Texas groundwater
216 Environ Monit Assess (2010) 160:215227
quality. ArcView GIS was used to map, query,
and analyze the data in these studies. Vinten and
Dunn (2001) studied the effects of land use on
temporal changes in well water quality. Levallois
et al. (1998) studied groundwater contamination
through nitrates associated with intense potato
culturing in Qubec, Canada. The data analysis
was carried out by combining GIS and statisti-
cal methods. Ahn and Chon (1999) investigated
groundwater contamination and spatial relation-
ships among groundwater quality, topography, ge-
ology, land use, and pollution sources using GIS in
Seoul, Korea. Ducci (1999) produced groundwa-
ter contamination risk and quality maps by using
GIS in Italy. Fritch et al. (2000) developed an
approach to evaluate the susceptibility of ground-
water in north-central Texas to contamination.
Interpolation is the estimation of Z values of a
surface at an unsampled point based on the known
Z values of surrounding points. There are two
main groupings of interpolation techniques: deter-
ministic and geostatistical. Deterministic interpo-
lation techniques create surfaces from measured
points, based on either the extent of similarity
(e.g., inverse distance weighted (IDW)) or the
degree of smoothing (e.g., radial basis functions).
Geostatistical interpolation techniques (e.g., krig-
ing) utilize the statistical properties of the mea-
sured points (ESRI (Environmental Systems
Research Institute) 2001).
Applications of geostatistics can be found in
very different disciplines ranging from the classi-
cal elds of mining and geology to soil science,
hydrology, meteorology, environmental sciences,
agriculture, and even structural engineering. Krig-
ing is used widely in geology, hydrology, environ-
mental monitoring, and other elds to interpolate
spatial data (Stein 1999). With the recent advances
in computation facilities and the availability of
geostatistical software, the use of kriging in the
spatial analysis of environmental data has become
increasingly popular. Today, a number of variants
of kriging are in general use, and these are simple
kriging, ordinary kriging (OK), universal kriging,
block kriging, cokriging, and disjunctive kriging.
Among the various forms of kriging, OK has
been used widely as a reliable estimation method
(Yamamoto 2000). Kriging is distinguished from
IDW and other interpolation methods by tak-
ing into consideration the variance of estimated
parameters (Buttner et al. 1998). OK is most
commonly adopted for environmental studies
(Poon et al. 2000; Kravchenko and Bullock 1999;
Lin et al. 2001; Tranchant and Vincent 2000;
Gringarten and Deutsch 2001). A more detailed
explanation of the kriging method is given by
Stein (1999), Yamamoto (2000), Gringarten and
Deutsch (2001), Mcgrath and Zhang (2003), and
Cressie (1990). Zimmerman et al. (1999) com-
pared the accuracy of OK, Universal Kriging, and
IDW methods based on an analysis of synthetic
data. Pozdnyakova and Zhang (1999) used the
geostatistical methods of kriging and cokriging
to estimate the sodium adsorption ratio in an
agricultural eld. Zhu et al. (2001) produced a
radon distribution map using the kriging and GIS
techniques in Belgium. Dagostino et al. (1998)
investigated the spatial distribution of nitrate con-
centration in the aquifer of central Italy and com-
pared cokriging and OK techniques.
The purposes of this investigation are (1) to
provide an overview of present groundwater qual-
ity and (2) to determine spatial distribution of
groundwater quality parameters such as pH, elec-
trical conductivity, chloride, sulfate, hardness, and
nitrate concentrations, and (3) to map groundwa-
ter quality in the Konya City area by using GIS
and geostatistics techniques.
Description of study area
The city of Konya is located at between 36.5

and
39.5

north latitude and 31.534.5

east longitude
and is the largest province of Turkey with a sur-
face area of 38,183 km
2
. The population of the city
is about 850,000. Figure 1 shows the location of
Konya City. Study area has about 17.1 km wide
from east to west and 25 km long from north to
south, which yields a total area of 427.5 km
2
.
A large proportion of water requirements for
the City of Konya are supplied from 198 ground-
water wells. Presently, new deep wells are still be-
ing drilled and operated by the Water Authority
of Konya City Municipality (WAKCM), as the
water requirements of the city constantly increase.
Depth of the wells varies between 25 m (mini-
mum) and 206 m (maximum), with an average of
Environ Monit Assess (2010) 160:215227 217
Fig. 1 Study area
location
128 m. In 1995, a drinking water treatment plant
(WTP) was put into operation. The amount of
water drawn from wells and the amount of water
supplied from WTP are shown in Fig. 2 for the
years between 1998 and 2005. As indicated by
Fig. 2, the amount of water drawn from wells was
44 10
6
m
3
in 1998 and was increased up to
89.3 10
6
m
3
in 2001 (Water Authority, Works
Report 2000 2001).
Geology and hydrogeology
A simplied geological map of the study area,
the location, the number of wells, and the direc-
tion of groundwater ow are shown in Fig. 3.
It was digitized in a Universal Transverse Mer-
cator (UTM) projection system by applying the
digitizing method on-screen using the ArcGIS
software. The deep wells in and around the city
center are located generally in Plio-Quaternary
aged soft and partially cemented sediments.
The basement rocks of the western high parts
of the investigated eld, which includes Konya
City and its vicinity, are made of Paleozoic,
Mesozoic, and Tertiary aged different lithologic
units. Plio-Quaternary aged detritic sediments
cover the plain. The principal aquifer formations
of the Konya plain are Paleozoic, Mesozoic, and
Fig. 2 Water supplies
from wells and WTP to
the city between 1998
and 2005
218 Environ Monit Assess (2010) 160:215227
Fig. 3 A simplied
geological map showing
sampling points and
direction of groundwater
ow
Neogene aged limestones and sandygravelly lev-
els of Plio-Quaternary aged detritus.
Land use pattern
Figure 4 shows the land use patterns of study area.
Konya City is divided into ve (development ar-
eas, residential areas, industrial areas, agricultural
areas, and cemeteries) districts. The locations of
the 177 groundwater wells were classied accord-
ing to land use patterns: 82 samples in residen-
tial district, 30 samples in development district,
13 samples in industrial district, 27 samples in
agricultural district, one sample in cemetery, and
23 samples other areas. There are 60 cemeter-
ies actively used within the city area. Location
and distances of the cemeteries from water wells
are illustrated in the generalized land uses map
Environ Monit Assess (2010) 160:215227 219
Fig. 4 Land use pattern
of Konya City
in Fig. 4. Twenty-seven water wells are located
south of the city. Some vegetables such as carrot,
tomato, green pepper, cucumber, etc. are culti-
vated, and nitrogen-based fertilizers are used by
farmers in this area. Eighty-two water wells are
located in the residential area of the city. In this
area, new residential and industrial activities have
been rapidly increasing over the years.
Materials and methods
A GIS software package ArcGIS 9.0 and Arc-
GIS Geostatistical Analyst extension were used
to map, query, and analyze the data in this study
for the assessment of groundwater quality. The
paper map of the city has a 1:25,000 scale and
was digitized to UTM coordinate system (6

sec-
tion width) by applying the on-screen digitizing
method. The well locations were obtained for 177
wells spreading all over the region by using a
Magellan Spor Trak hand held Global Positioning
System (GPS) receiver. In addition, attribute in-
formation of wells was also input to a digital map
using the ArcGIS 9.0 software.
Chemical analysis
Groundwater samples were taken directly from
177 wells in April and May 2001 by the WAKCM.
220 Environ Monit Assess (2010) 160:215227
The wells were pumped until the temperature-,
conductivity-, and pH-stabilized. Glass containers
were used for the collection of water samples for
the analyses and delivered to the WAKCM labo-
ratory within 2 h. Analyses were normally carried
out as soon as the samples reach the laboratory.
Water quality parameters (chloride, sulfate,
hardness) were then analyzed in the laboratory
according to the methods given in the Standard
Methods (APHA, AWWA, WPCF 1985). Sam-
ple pH was measured using a glass electrode pH
meter. Electrical conductivity was measured using
a platinum electrode conductivity meter. Nitrate
(NO
3
) concentrations were measured with a UV
VIS spectrophotometer by Brucine colorimetric
method (APHA, AWWA, WPCF 1976).
Spatial interpolation of groundwater
quality parameters
In this study, a geostatistical software package
called ArcGIS Geostatistical Analyst Extension
was used for the ordinary kriging estimations.
Ordinary kriging was used to obtain the spatial
distribution of groundwater quality parameters
over the area. The groundwater quality data has
been checked by a histogram tool and normal
QQPlots to see if it shows a normal distribution
pattern. For each water quality parameter, an
analysis trend was made and the 11 different semi-
variogram models were tested. Prediction perfor-
mances were assessed by cross-validation.
Results and discussion
Groundwater quality gives a clear picture about
the usability of the water for different purposes.
The standard quality for drinking water has
been specied by the World Health Organization
(WHO) and the Turkish Standard Institute (TSE;
(WHO) World Health Organization 2004; TSE
1997). It has given the permissible and desir-
able limits for the presence of various elements
in groundwater (Table 1). Statistical evaluation
of groundwater quality parameters can be seen
in Tables 2 and 3 shows chemical composi-
tions of groundwater relating to land use in the
Konya City.
Examining the distribution of the data
Kriging methods work best if the data are ap-
proximately normally distributed. Transforma-
tions were used to make data normally distributed
and satisfy the assumption of equal variability for
the data. In the ArcGIS Geostatistical Analyst,
the histogram and normal QQPlots were used to
see what transformations, if any, are needed to
make the data more normally distributed.
Normal QQPlots provide an indication of uni-
variate normality. If the data is asymmetric (i.e.,
far from normal), the points will deviate from the
line. Histogram tool and normal QQPlots analysis
were applied for each water quality parameter,
and it was found that only the pH parameter
showed a normal distribution. It was determined
that electrical conductivity, chloride, sulfate, hard-
ness, and nitrate concentrations do not show nor-
mal distributions. For those parameters, a log
transformation has been applied to make the dis-
tribution closer to normal.
Examining the global trend through
trend analysis
The trend tool raises the points above a plot of
the study site to the height of the values of the
attribute of interest in a three-dimensional plot of
the study area. The points are then projected in
Table 1 Standards for
quality of drinking
water (TSE Turkish
Standar Institute 1997;
WHO World Health
Organization 1985)
MCL maximum
contaminant level
Parameter WHO (2004) TSE (1997) TSE (1997)
MCL recommend MCL
pH 6.58.5 6.58.5 6.59.2
Conductivity (S/cm) 400 2,000
Chloride (mg/L) 250 25 600
Sulfate (mg/L) 250 25 250
Hardness (

F) 50
Nitrate (mg/L) 50 25 50
Environ Monit Assess (2010) 160:215227 221
Table 2 Statistical evaluation of groundwater quality parameters (n = number of studied samples)
pH n Conductivity n Chloride n Sulfate n Hardness n Nitrate n
(S/cm) (mg/L) (mg/L) (

F) (mg/L)
<7.5 64 < 1000 159 < 50 135 < 50 102 < 30 41 < 10 48
7.58 100 1,0002,000 17 50100 40 50100 70 3050 117 1050 106
>8 13 > 2, 000 1 > 100 2 > 100 5 > 50 19 > 50 3
Parameter Min. Max. Mean Median SD Skewness Kurtosis
pH 7.0 8.6 7.59 7.6 0.28 0.64 3.73
Conductivity (S/cm) 332 2,892 680.4 608 305.32 3.08 18.61
Chloride (mg/L) 10 520 44.7 28 49.17 5.85 52.03
Sulfate (mg/L) 7 550 64.6 37 75.01 3.54 19.04
Hardness (

F) 15 86 36.4 34 11.72 1.42 6.15

Nitrate (mg/L) 3 110 16.1 14 13.66 4.27 26.07
two directions onto planes that are perpendicular
to the map plane. A polynomial curve is t to
each projection. If the curve through the projected
points is at, no global trend exists. In this situa-
tion, it is not necessary to remove the trend before
modeling the semivariogram/covariance for krig-
ing. For each water quality parameter, an analysis
trend was made, and it was determined that there
is no global trend for all parameters.
Semivariogram models
In this study, the semivariogram models (Cir-
cular, Spherical, Tetraspherical, Pentaspherical,
Exponential, Gaussian, Rational Quadratic, Hole
effect, K-Bessel, J-Bessel, Stable) were tested
for each parameter data set. Prediction perfor-
mances were assessed by cross-validation. Cross-
validation allows determination of which model
provides the best predictions.
For a model that provides accurate predictions,
the standardized mean error should be close to 0,
the root-mean-square error and average standard
error should be as small as possible (this is use-
ful when comparing models), and the root-mean
square standardized error should be close to 1.
When the average estimated prediction standard
errors are close to the root-mean-square predic-
tion errors from cross-validation, then you can
be condent that the prediction standard errors
are appropriate (ESRI (Environmental Systems
Research Institute) 2001).
Table 3 Chemical compositions of groundwater relating to land use
Land use pH Conductivity Chloride Sulfate Hardness Nitrate
(S/cm) (mg/L) (mg/L) (

F) (mg/L)
Development areas (30) Ave. 7.7 530 27.8 48 28.9 13.5
Max. 8.6 1,448 115 210 59 26
Min. 7.1 332 12 7 15 3.7
Residential areas (82) Ave. 7.6 725.6 50.3 78 37.9 18.9
Max. 8.3 1,612 238 550 77 110
Min. 7.9 418 15 16 15 3
Cemetery areas (1) 7.3 1,542 160 110 72 84
Industrial areas (13) Ave. 7.4 1,009 106.5 111 45.5 12.2
Max. 8.1 2,892 520 220 86 22
Min. 7.3 452 25 26 22 8
Agricultural areas (27) Ave. 7.7 532.4 23.2 21.3 32.9 9.84
Max. 8.1 938 50 50 55 23
Min. 7 377 13 8 22 3
Others (23) Ave. 7.6 678 32.8 62.3 37.7 15.3
Max. 8.1 1,700 140 450 78 28
Min. 7 412 10 8 20 5
222 Environ Monit Assess (2010) 160:215227
After applying different models for each water
quality parameter examined in this study, the er-
ror was calculated using cross-validation and mod-
els giving best results were determined. Table 4
shows the most suitable models and their pre-
diction error values for each parameter. Table 4
also shows that for different parameters, different
models may give better results.
The groundwater quality map
Figure 5 shows the spatial distribution of pH, con-
ductivity, chloride, sulfate, hardness, and nitrate
concentrations in study area, respectively.
Agroundwater quality map (Fig. 6) was created
following the classication shown in Table 5. The
construction of the groundwater quality map was
carried out through the overlapping of the the-
matic maps (Fig. 5af), which are produced as a
result of kriging interpolations. The spatial inte-
gration for nal water quality mapping was carried
out using ArcGIS Spatial Analyst extension.
pH
No health-based guideline value is proposed for
pH. Although pH usually has no direct impact on
consumers, it is one of the most important opera-
tional water quality parameters, with the optimum
pH required often being in the range of 6.59.5
(WHO (World Health Organization) 2004).
The maximum contaminant level (MCL) for
pHin drinking water is given as to be 6.58.5 mg/L
by the WHO but 6.59.2 by the TSE. In addi-
tion, Turkish Standards recommend the value of
6.58.5 for pH (WHO (World Health Organiza-
tion) 2004; TSE 1997).
The minimumand maximumvalues of pHwere
measured as 7.0 and 8.6, respectively. There was
no well in which the pH exceeds the MCL of 9.2
given in the Turkish Standards. Spatial distribu-
tions of pH concentrations are shown in Fig. 5a.
It is shown that the low pH concentrations (7.0
7.5) occur north-east of the city and within the city
center.
Conductivity
Electrical conductivity (EC) is a parameter re-
lated to total dissolved solids (TDS). The impor-
tance of EC and TDS lies in their effect on the
corrrosivity of a water sample and in their effect
on the solubility of slightly soluble compounds
such as CaCO
3
. TDS is comprised of inorganic
salts (principally calcium, magnesium, potassium,
sodium, bicarbonates, chlorides, and sulfates) and
small amounts of organic matter that are dissolved
in water. Concentrations of TDS in water vary
considerably in different geological regions ow-
ing to differences in the solubilities of minerals
(WHO (World Health Organization) 2004).
For EC, the value of 400 S/cm is rec-
ommended with the MCL of 2,000 S/cm by
Turkish Standards (1997). The recommended
value of 400 S/cm was obtained from 11 out
of 177 wells. The MCL of 2,000 S/cm was
only exceeded in well No. 54 with the value of
2,892 S/cm. As shown in Fig. 5b, the EC value in-
creases from south to north with the upper ranges
being greater than 1,000 S/cm.
Table 4 Fitted parameters of the theoretical variogram model for groundwater quality parameters
Parameters Models Prediction errors
Mean Root-mean Average standard Mean Root-mean-square
square error standardized standardized
pH Stable 0.00207 0.2239 0.2252 0.01019 1.0
Conductivity Tetraspherical 11.67 258 227.6 0.03142 1.004
Chloride Stable 2.325 42.75 33.4 0.04598 1.163
Sulfate Rational quadratic 2.876 61.13 91.95 0.02767 0.8174
Hardness Gaussian 0.2076 9.439 9.21 0.008576 0.9892
Nitrate K-Bessel 0.3677 12.18 9.844 0.02927 1.093
Environ Monit Assess (2010) 160:215227 223
a
d e f
b c
Fig. 5 Spatial distribution of a pH, b electrical conductivity, c chloride, d sulfate, e hardness, and f nitrate concentrations
Chloride
Chlorides occur in all natural waters in widely
varying concentrations. The chloride content nor-
mally increases as the mineral content increases
(Sawyer and Mccarty 1978). The chloride ion oc-
curs in natural waters in fairly low concentrations,
usually less than 100 mg/L, unless the water is
brackish or saline (Fetter 1999). No health-based
guideline value is proposed for chloride in drink-
ing water. High concentrations of chloride give a
salty taste to water and beverages (WHO (World
Health Organization) 2004). However, chloride
concentrations in excess of about 250 mg/L can
give rise to a detectable taste in water (WHO
(World Health Organization) 2004; Sawyer and
224 Environ Monit Assess (2010) 160:215227
Fig. 6 Final groundwater quality map
Mccarty 1978). Chloride in drinking water orig-
inates from natural sources, sewage and indus-
trial efuents, urban runoff containing de-icing
salt, and saline intrusion (WHO (World Health
Organization) 2004).
The MCL for chloride in drinking water is
given as 250 mg/L by the WHO but 600 mg/L by
the TSE. In addition, Turkish standards recom-
mend the value of 25 mg/L for chloride (WHO
(World Health Organization) 2004; TSE 1997).
Chloride concentration was complied with a value
of 25 mg/L for 52 out of 177 wells. There was
no water well in which the chloride concentration
exceeds the MCL given in Turkish Standards.
As indicated by Fig. 5c, chloride concentration
increased from southwest to northeast. In a wide
area around the south and west part of the city,
less than 50 mg/L chloride concentration occurs.
Sulfate
Sulfates occur naturally in numerous minerals and
are used commercially, principally in the chem-
ical industry. They are discharged into water in
industrial wastes and through atmospheric depo-
sition; however, the highest levels usually occur
in groundwater and are from natural sources. No
health-based guideline is proposed for sulfate.
Sulfur is widely present in reduced forms in ig-
neous, sedimentary, and metamorphic rocks as
metallic suldes. This sulfur turns to sulfate when
weathered in contact with aerated water. Water
in igneous or metamorphic rocks generally con-
tains less than 100 mg/L sulfate, but sedimentary
rocks can contain much higher levels (Fytianos
and Christophoridis 2004). The presence of sulfate
in drinking water can cause a noticeable taste,
and very high levels might cause a laxative effect
in unaccustomed consumers. Taste impairment
varies with the nature of the associated cation;
taste thresholds have been found to range from
250 mg/L for sodium sulfate to 1,000 mg/L for cal-
cium sulfate (WHO (World Health Organization)
2004).
TSE and WHO indicate the MCL of 250 mg/L
for sulfate. A concentration of 25 mg/L is rec-
ommended in Turkish Standards (WHO (World
Health Organization) 2004; TSE 1997). The sul-
fate concentration of 25 mg/L can be seen in 36 of
the 177 wells. But, the MCL of sulfate (250 mg/L)
was exceeded in wells No. 72, No. 81, No. 64,
No. 34, and No. 37. It can be seen in Fig. 5d that
the sulfate concentration increases from south to
Table 5 Groundwater quality classication (Ducci 1999)
Quality Class Chloride (mg/L) Sulfate (mg/L) Hardness (

F) Conductivity (S/cm) Nitrate (mg/L)

Optimum A < 50 < 50 < 30 < 1,000 < 10
Medium B 50200 50250 3050 1,0002,000 1050
Poor C > 200 > 250 > 50 > 2,000 > 50
Environ Monit Assess (2010) 160:215227 225
north of the city and it exceeds MCL of 250 mg/L
northwest of the city.
Hardness
Hardness in water is caused by dissolved calcium
and, to a lesser extent, magnesium. It is usually
expressed as the equivalent quantity of calcium
carbonate (WHO (World Health Organization)
2004). The hardness of water reects the nature
of the geological formations with which it has
been in contact (Sawyer and Mccarty 1978). A
number of ecological and analytical epidemio-
logical studies have shown a statistically signi-
cant inverse relationship between the hardness of
drinking water and cardiovascular disease. There
is some indication that very soft waters may have
an adverse effect on mineral balance, but detailed
studies were not available for evaluation. Public
acceptability of the degree of hardness may vary
considerably from one community to another de-
pending on local conditions, and the taste of water
with hardness in excess of 500 mg/L is tolerated by
consumers in some instances. Soft water, with a
hardness of less than 100 mg/L, may, on the other
hand, has a low buffering capacity and so can
be more corrosive for water pipes (WHO (World
Health Organization) 2004).
There is no limitation of water hardness by
Turkish Standards but the WHO gives a value of
50

F as the maximum value for water hardness

(WHO(World Health Organization) 2004). There
is no water hardness value at all between 7.5
15

F and 07.5

F, which are classied as mod-

erately hard and soft water in the study area,
respectively. The minimum, maximum, and mean
values of hardness are 15

F, 86

F, and 36.4

F,
respectively.
The north of the study area has an aquifer
lithology of limestone and the hardness value in
this area reaches up to 50

F. Insoluble bicarbon-
ates are converted to soluble carbonates because
of the existence of carbon dioxide in the soil.
Since limestone is not pure carbonate but includes
impurities such as sulfates and chlorides, these
materials become exposed to the solvent action of
the water as the carbonates are dissolved and they
pass into solution, too. Therefore, chlorides and
sulfate concentrations as well as water hardness
were very high in such areas. There was no water
well on the aquifer lithology of the limestone
(Neogene), limestone (Paleozoic), and clay (Plio-
Quaternary) in the study area (Fig. 3).
The south of the study area (Alakova region)
has sandy, gravelly (Plio-Quaternary) and sandy
clay (Plio-Quaternary) aquifer lithology, and the
hardness value in this area was observed to be
between 30

F and 50

F. From the north of the

Alakova region to southwest, west and northwest
of the city, the water hardness was estimated to
be less than 30

F. The value of water hardness

increases from south to northeast of the city area
(Fig. 5e).
Nitrate
The nitrate concentration in groundwater and sur-
face water is normally low but can reach high
levels as a result of leaching or runoff from agri-
cultural land or contamination from human or
animal wastes as a consequence of the oxidation
of ammonia and similar sources (WHO (World
Health Organization) 2004).
During recent years, the problem of ground-
water contamination by nitrates has been stud-
ied thoroughly all over the world (Hudak 1999,
2000; Vinten and Dunn 2001; Levallois et al.
1998; TSE 1997; Nas and Berktay 2006; Fytianos
and Christophoridis 2004). The primary health
concern regarding nitrate and nitrite is the for-
mation of methemoglobinemia, the so-called blue-
baby syndrome. Several studies document adverse
effects of higher nitrate levels, most notably
methemoglobinemia (Hudak 1999, 2000; Levallois
et al. 1998; WHO (World Health Organization)
1985, 2004; EPA (U.S. Environmental Protection
Agency) 1993).
The MCL of nitrate is given as 50 mg/L by the
TSE and WHO for drinking water. On the other
hand, the TSE describes the limit of concentra-
tion of nitrate as 25 mg/L (WHO (World Health
Organization) 2004; TSE 1997).
Spatial distributions of nitrate concentrations
are shown in Fig. 5f. The nitrate concentrations of
84, 95, and 110 mg/L were measured in the wells
No. 31, No. 2, and No. 41, respectively. Nitrate
concentrations for these three wells exceed the
226 Environ Monit Assess (2010) 160:215227
MCL of 50 mg/L indicated by the TSE. Ceme-
teries contribute to nitrate pollution into ground-
water (EPA (U.S. Environmental Protection
Agency) 1993). Well No. 31 was placed at the
cemetery, and this high level of nitrate could be
attributed to its location. On the other hand, well
No. 41 was placed at a park in the city center
where fertilizers are often applied to the lawns.
Fertilizers may cause this high level of nitrate
concentration. In the same area, a 46 mg/L nitrate
concentration was also measured on the samples
taken from well No. 33. This groundwater well
was also taken out of operation. As a result of
this investigation, wells No. 31, No. 2, and No. 41
had been taken out of operation. Nitrate contents
for 12 groundwater wells do not meet the stan-
dard of 25 mg/L indicated by the TSE, whereas
average nitrate concentration at residential ar-
eas is 18.9 mg/L and in the agricultural areas is
9.84 mg/L.
Figure 6 shows the nal water quality map
that was produced by overlapping of the thematic
maps as a result of geostatistical analysis. The
spatial integration for groundwater quality map-
ping was carried out using ArcGIS Spatial Analyst
extension. The area for optimum water quality
could be found on a large area at the southwest
and small areas at city center and northwest of
the city. These areas for optimum water quality
cover 5.03%(about 21.51 km
2
) of total study area.
The rest of the study area, which is about 94.97%
(405.99 km
2
), has water classied as medium and
poor quality levels. In addition, some water wells
have poor water quality in terms of nitrate con-
centration especially at the old occupied part of
the city center (Fig. 5f).
Conclusions
The primary objective of this study was to map
and evaluate the groundwater quality in Konya
City. Spatial distribution of groundwater quality
parameters were carried out through GIS and
geostatistical techniques. These techniques have
successfully demonstrated its capability in ground-
water quality mapping of Konya City.
Geostatistical techniques create surfaces incor-
porating the statistical properties of the measured
data. Because geostatistics is based on statistics,
these techniques produce not only prediction sur-
faces but also errors or uncertainty surfaces, giving
you an indication of how good the predictions are.
The disadvantage of kriging is the need to make
many decisions regarding transformations, trends,
models, and parameters. However, the advantages
of this system are that it is very exible, allows
assessment of spatial autocorrelation, and can ob-
tain prediction standard errors.
The nal map, which is a groundwater quality
map, shows that the southwest of the city has
optimumgroundwater quality, and, in general, the
groundwater quality decreases south to north of
the city. The deep wells in and around the city
center are located generally in Plio-Quaternary
aged soft and partially cemented sediments. On
the other hand, the south of the city is a low-
density residential area and a mainly agricultural
area. Industrial activities are located northeast of
the city. Forty-six percent of wells (82 wells) are
located in the residential areas of the city. In this
area, new residential and industrial activities have
been rapidly increasing over the years.
Prior to new well drilling, the groundwater
quality map as a result of this research has to
be taken into account by WAKCM as a decision
support system.
Acknowledgements This study was supported by
Selcuk University Scientic Research Fund with project
No: 2001/145. The authors would like to thank Water
Authority of Konya City Municipality (WAKCM).
References
Ahn, H., & Chon, H. (1999). Assessment of groundwater
contamination using geographic information systems.
Environmental Geochemistry and Health, 21, 273289.
doi:10.1023/A:1006697512090.
APHA, AWWA, WPCF (1976). Standard methods for the
examination of water and wastewater (14th edition).
New York: APHA, AWWA, WPCF.
APHA, AWWA, WPCF (1985). Standard methods for the
examination of water and wastewater (16th edition).
Washington: APHA, AWWA, WPCF.
Buttner, O., Becker, A., Kellner, S., Kuehn, S., Wendt-
Potthoff, K., Zachmann, D. W., et al. (1998). Geo-
statistical analysis of surface sediments in an acidic
mining lake. Water, Air, and Soil Pollution, 108, 297
316. doi:10.1023/A:1005145029916.
Environ Monit Assess (2010) 160:215227 227
Cressie, N. (1990). The origins of kriging. Mathematical
Geology, 22, 239252.
Dagostino, V., Greene, E. A., Passarella, G., Vurro, M.
(1998). Spatial and temporal study of nitrate con-
centration in groundwater by means of coregion-
alization. Environmental Geology, 36(34), 285295.
doi:10.1007/s002540050344.
Ducci, D. (1999). GIS techniques for mapping groundwa-
ter contamination risk. Natural Hazards, 20, 279294.
doi:10.1023/A:1008192919933.
Engel, B. A., & Navulur, K. C. S. (1999). The role of
geographical information systems in groundwater en-
gineering. In J. W. Delleur (Ed.), The handbook of
groundwater engineering (pp. 21, 116). Boca Raton:
CRC.
EPA (U.S. Environmental Protection Agency) (1993).
Wellhead protection: a guide for small communities.
Ofce of Research and Development Ofce of Water,
Washington, DC., EPA/625/R-93/002, 144 p.
ESRI (Environmental Systems Research Institute) (2001).
Using ArcGIS geostatistical analyst (300 p.), USA.
Fetter, C. W. (1999). Contaminant hydrogeology (p. 500).
USA: Prentice Hall.
Fritch, T. G., Yelderman, J. C., Dworkin, S. I., & Arnold,
J. G. (2000). A predictive modeling approach to as-
sessing the groundwater pollution susceptibility of the
Paluxy Aquifer, Central Texas, using a geographic
information system. Environmental Geology, 39(9),
10631069. doi:10.1007/s002549900078.
Fytianos, K., & Christophoridis, C. (2004). Nitrate, ar-
senic and chloride pollution of drinking water in
Northern Greece: Elaboration by applying GIS.
Environmental Monitoring and Assessment, 93, 5567.
doi:10.1023/B:EMAS.0000016791.73493.aa.
Gringarten, E., & Deutsch, C. V. (2001). Teachers
aide variogram interpretation and modeling. Math-
ematical Geology, 33(4), 507534. doi:10.1023/A:
1011093014141.
Hudak, P. F. (1999). Chloride and nitrate distributions in
the Hickory Aquifer, Central Texas, USA. Environ-
ment International, 25(4), 393401. doi:10.1016/S0160-
4120(99)00016-1.
Hudak, P. F. (2000). Regional trends in nitrate con-
tent of Texas groundwater. Journal of Hydrol-
ogy (Amsterdam), 228(12), 3747. doi:10.1016/S0022-
1694(99)00206-1.
Hudak, P. F. (2001). Water hardness and sodium trends in
Texas aquifers. Environmental Monitoring and Assess-
ment, 68, 177185. doi:10.1023/A:1010760413010.
Hudak, P. F., & Sanmanee, S. (2003). Spatial patterns
of nitrate, chloride, sulfate, and uoride concentra-
tions in the woodbine aquifer of North-Central Texas.
Environmental Monitoring and Assessment, 82, 311
320. doi:10.1023/A:1021946402095.
Kravchenko, A., & Bullock, D. G. (1999). A comparative
study of interpolation methods for mapping soil prop-
erties. Agronomy Journal, 91(3), 393400.
Levallois, P., Thriault, M., Roufgnat, J., Tessier, S.,
Landry, R., Ayotte, P., et al. (1998). Groundwater
contamination by nitrates associated with intensive
potato culture in Qubec. The Science of the Total
Environment, 217, 91101. doi:10.1016/S0048-9697(98)
00191-0.
Lin, Y., Tan, Y., & Rouhani, S. (2001). Identifying spatial
characteristics of transmissivity using simulated an-
nealing and kriging methods. Environmental Geology,
41(12), 189199. doi:10.1007/s002540100382.
Mcgrath, D., &Zhang, C. (2003). Spatial distribution of soil
organic carbon concentrations in grassland of Ireland.
Applied Geochemistry, 18, 16291639. doi:10.1016/
S0883-2927(03)00045-3.
Nas, B., &Berktay, A. (2006). Groundwater contamination
by nitrates in the City of Konya, (Turkey): A GIS
perspective. Journal of Environmental Management,
79, 3037. doi:10.1016/j.jenvman.2005.05.010.
Poon, K., Wong, R. W., Lam, M. H., Yeung, H., & Chiu,
T. K. (2000). Geostatistical modelling of the spatial
distribution of sewage pollution in coastal sedi-
ments. Water Research, 32, 99108. doi:10.1016/S0043-
1354(99)00119-0.
Pozdnyakova, L., &Zhang, R. (1999). Geostatistical analy-
ses of soil salinity in a large eld. Precision Agriculture,
1, 153165. doi:10.1023/A:1009947506264.
Sawyer, C. N., & Mccarty, P. L. (1978). Chemistry for envi-
ronmental engineering (p. 532). NY: Mc-Graw Hill.
Stein, M. L. (1999). Interpolation of spatial data: Some the-
ory for kriging (p. 264). Berlin: Springer.
Tranchant, B. J. S., & Vincent, A. P. (2000). Statistical
interpolation of ozone measurements from satellite
data (TOMS, SBUV and SAGE II) using the krig-
ing method. Annales Geophysicae, 18, 666678. doi:10.
1007/s00585-000-0666-x.
TSE (1997). Drinking water Turkish standards, Turkish
standards (TSE-266). Ankara.
Vinten, A. J. A., & Dunn, S. M. (2001). Assessing the
effects of land use on temporal change in well water
quality in a designated nitrate vulnerable zone. The
Science of the Total Environment, 265, 253268. doi:
10.1016/S0048-9697(00)00662-8.
Water Authority. Works Report 2000 (2001). Water au-
thority (143 p.). Municipality of Konya City, Konya,
Turkey (In Turkish).
WHO (World Health Organization) (1985). Health haz-
ards from nitrates in drinking water. WHO, Regional
Ofce for Europe.
WHO (World Health Organization) (2004). Guidelines
for drinking water quality, third edition. Geneva:
WHO.
Yamamoto, J. K. (2000). An alternative measure of
the reliability of ordinary kriging estimates. Math-
ematical Geology, 32(4), 489509. doi:10.1023/A:
1007577916868.
Zhu, H. C., Charlet, J. M., & Pofjn, A. (2001). Radon
risk mapping in Southern Belgium: An application of
geostatistical and GIS techniques. The Science of the
Total Environment, 272, 203210. doi:10.1016/S0048-
9697(01)00693-3.
Zimmerman, D., Pavlik, C., Ruggles, A., & Armstrong,
M. P. (1999). An experimental comparison of ordinary
and universal kriging and inverse distance weighting.
Mathematical Geology, 31(4), 375390. doi:10.1023/
A:1007586507433.