In this study, the fuzzy analytic hierarchy process (AHP) is used
to study the relationship between drinking water quality based on content of
inorganic components and landform classes in the south of Firozabad, west of
Fars province, Iran. For determination of drinking water quality based on
content of inorganic components, parameters of calcium (Ca), chlorine (Cl),
magnesium (Mg), thorium (TH), sodium (Na), electrical conductivity (EC),
sulfate (SO
Landform characteristics can affect the direction of water movement and water
quality. Hence, in the different landforms, there is different water quality
(Bise, 2013). To this end, studies on the relationship between landform
classes and water quality have received significant attention. For example,
William and Randall (2007) investigated
runoff and water quality from three soil landform units on the Mancos Shale.
A survey of sediment basins in steep, dissected shale uplands indicated that
an average of 1.25 Mg ha yr
Descriptive statistics of the parameters for evaluation of water quality (Fars Regional Water Authority).
Drinking water quality based on content of inorganic component standards (WHO; Shobha et al., 2013).
In addition, different algorithms have been employed for the determination of water quality. Demissie et al. (2012) developed a complementary modelling framework to handle systematic error in physically based groundwater flow model applications that used data-driven models of the errors during the calibration phase. The effectiveness of four error-correcting data-driven models, namely, artificial neural networks (ANN), support vector machines (SVM), decision trees (DT), and instance-based weighting (IBW), was examined for forecasting head prediction errors and subsequently updating the head predictions at existing and proposed observation wells. Rule-based modelling (Manoucher et al., 2010) was used for spatial prediction of groundwater quality in Beaufort West, in the Karoo region of South Africa. The groundwater quality data from about 100 bore wells with a 30-year span collected between 1970 and 2007 were used. The variables used in the analyses included chemicals such as chloride, sulfate, magnesium, sodium, phosphates, and calcium. These were used as predictors for groundwater quality and electrical conductivity. Aliabadi and Soltanifard (2014) used fuzzy inference for determination of the impact of water and soil electrical conductivity and calcium carbonate on wheat crop use. The inference system estimated the performance using soil EC, water EC, and calcium carbonate in the soil as input parameters, and also analysed them.
The aim of this study is the determination of the relationship between
landform classes and drinking water quality based on content of inorganic
components in southern Firozabad, Iran. In this study, drinking water quality
based on content of inorganic components is evaluated using parameters of
calcium (Ca), chlorine (Cl), magnesium (Mg), thorium (TH), sodium (Na),
electrical conductivity (EC), sulfate (SO
The methodology employed in this study is summarized in Fig. 1.
Flowchart for the methodology used in this study to determine the relationship between drinking water quality based on the content of inorganic components and landform classes.
Location of the study area (DEM with a spatial resolution of 30 m;
source:
This study was carried out south of Firozabad, west of Fars province, Iran.
It has an area of 722.91 km
Semivariogram graphs:
Fuzzy AHP procedure for drinking water quality based on content of inorganic components.
Landform classification based on TPI (source: Weiss, 2006).
One of these important factors is drinking water quality based on content of
inorganic components in the study area. In order to predict the variability
of drinking water quality based on content of inorganic components, calcium
(Ca), chlorine (Cl), magnesium (Mg), thorium (TH), sodium (Na), electrical
conductivity (EC), sulfate (SO
Sampling nugget, partial sill, and RMSE of the different interpolated methods for predicted drinking water quality based on content of inorganic components using multiple linear regression (MLR).
Pair-wise comparison matrix for drinking water quality based on content of inorganic components.
The input parameters for determination of drinking water quality based on
content of inorganic components are Ca, Cl, Mg, TH, Na, EC, SO
Interpolated maps of the drinking water quality based on the content of inorganic component parameters generated by using OK.
Fuzzy maps of the study area for the drinking water quality based on the content of inorganic component parameters.
In order to compare the different interpolation techniques, we examined the
difference between known and predicted data using the root mean squared error
(RMSE; Eq. 2):
Drinking water quality based on the content of inorganic components map generated using fuzzy AHP.
Fuzzy logic was initially developed by Zadeh (1965) as a generalization of
classic logic. He defined a fuzzy set by membership functions from properties
of objects. A membership function assigns to each object a grade ranging
between 0 and 1. The value 0 means that
The development of GIS has contributed to facilitating the mapping of
drinking water quality based on content of inorganic components using both
Boolean and fuzzy methods. For each of the parameters, the following function
was used (Shobha et al., 2013):
TPI
maps generated using
Landform classification using the TPI method.
AHP is a structured technique for organizing and analysing complex decisions.
This method is based on a pair-wise comparison matrix. The matrix is called
consistent if the transitivity (Eq. 5) and reciprocity (Eq. 7) rules are
respected:
In a consistent matrix (Eq. 7), all the comparisons
Finally, in order to prepare the drinking water quality based on content of
the inorganic components map, it is necessary to calculate the convex
combination of the raster values containing the different fuzzy parameters
(Bijanzadeh and Mokarram, 2013; Mahdavi et al., 2015).
All the model parameter maps are constructed by interpolation between 50 sampling points using the kriging method. Next, fuzzy logic is applied to create a fuzzy parameter map for each parameter. To arrive at an integrated evaluation using drinking water quality based on content of inorganic component classes, the fuzzy parameter maps were aggregated into a drinking water quality based on content of an inorganic component map following a weighted summation using AHP.
Areas of the drinking water classes.
TPI (Weiss, 2006) compares the elevation of each cell in a digital elevation model (DEM) to the mean elevation of a specified neighbourhood around that cell. Positive and negative TPI values represent locations that are higher and lower than the average of their surroundings respectively. TPI values near zero are either flat areas (where the slope is near zero) or areas of constant slope (where the slope of the point is significantly greater than zero; Weiss, 2006).
TPI (Eq. 8) compares the elevation of each cell in a DEM to the mean
elevation of a specified neighbourhood around that cell. Mean elevation is
subtracted from the elevation value at the centre (Weiss, 2006):
Combining TPI at small and large scales allows a variety of nested landforms to be distinguished (Table 3).
OK was used for the prediction of the drinking water quality based on content
of inorganic component parameters (TH, Ca, Mg, Cl, Na, EC, SO
Each of the water parameter maps that were predicted by OK are shown in
Fig. 5. The lowest SO
The fuzzy maps prepared for the drinking water quality based on content of inorganic component parameters are shown in Fig. 6, where MF is closer to 0 with decreasing drinking water quality based on content of inorganic components, while MF is closer to 1 with increasing drinking water quality based on content of inorganic components (Soroush et al., 2013). Next, the AHP method was applied to the fuzzy parameter maps. The pair-wise comparison matrices used for preparation of the weights for each parameter in AHP are given in Table 5. The drinking water quality based on content of the inorganic components map generated using fuzzy AHP is shown in Fig. 7.
Areas of the landform classes.
The drinking water quality based on the content of inorganic components map
is classified into four classes (Mokarram et al., 2010; Shobha et al., 2013):
Low (not suitable for drinking): < 0.25; Moderate: 0.25–0.50; High: 0.50–0.75; Very high (suitable for drinking): > 0.75.
The results of the classification are shown in Table 6. It is found that
areas with suitable drinking water quality based on content of inorganic
components are located in the south-eastern and south-western parts of the
study area (Fig. 7).
In order to determine the relationship between landform classification and
drinking water quality based on content of inorganic components, a landform
classification map for the study area was prepared using TPI. The TPI maps
generated using small (3 cells) and large (45 cells) neighbourhoods are shown
in Fig. 8. TPI is between
The relationship between drinking water quality based on content of inorganic components and landform classes was determined (Fig. 11). It is found that drinking water quality based on content of inorganic components is high for streams, valleys, upland drainages, and local ridge classes, while the lowest drinking water quality based on content of inorganic components is in the plain small and midslope classes. The characteristics of landform classes, such as slope and geology, determine the drinking water quality based on content of inorganic components. For example, in the plain small class, due to the low slope, there are ample opportunities for high drinking water quality based on content of inorganic components (Christiansen, 2004). Therefore, landform maps can be used to predict drinking water quality based on content of inorganic components without water sampling and analysis in the laboratory.
Relationship between drinking water quality based on content of inorganic components and landform classes.
In this study, fuzzy AHP was used to study the relationship between drinking water quality based on content of inorganic components and landform classes in the south of Firozabad. It was found that 8.29 % of the study area had low water quality; 64.01, moderate; 23.33, high; and 4.38 %, very high. The lands suitable for drinking water are located in the south-eastern and south-western parts of the study area. The relationship between landform class and drinking water quality based on content of inorganic components shows that drinking water quality based on content of inorganic components is high in the stream, valleys, upland drainages, and local ridge classes, while the lowest drinking water quality based on content of inorganic components is in the plain small and midslope classes, so that the study determined that without measurement of water sample characteristics using a DEM and extraction landform classes by the TPI method, we can determine water quality by landform classes. For more accuracy, we suggest using a DEM with a high resolution such as radar and lidar imaging for extraction of landform classes and prediction of water quality by it.
The authors are grateful to the referees and Editor for valuable suggestions and comments which have led to the profound discussion of the results. The authors would like to thanks to all personnel of Agricultural Jihad of Fars province for their kind help. Edited by: R. Shang Reviewed by: S. K. Roy and one anonymous referee