Comparison Between Groundwater Quality at North East and South West of Erbil Governorate for Irrigation Using Some Global Systems and Principal Component Analysis

Dania Mohammed Danish Aladdin Sajadi¹*, Akram Othman Esmail²

^1,2 (Soil and Water Dept., College of Agricultural Engineering Sciences,  Salahaddin University-Erbil, Iraq)

¹Email: dania.sajadi@yahoo.com , ²Email : akram.esmail@su.edu.krd

Abstract: This investigation was conducted during June ,2019 to November ,2019  for  classification of water of  (54) wells in Erbil province 26 of them in north and north east of Erbil and  the other half at south west of Erbil. The studied waters were classified for irrigation purpose depending on some global systems or classification and Principal component analysis was done for the calculated parameters used in classification.. Depending on USDA classification(1954)  the water  of the water of (20 and 6) locations or wells having C2S1 and C3S1 Class respectively .While for the south east Erbil wells the water classes were C1S1, C2S1 , C3S1 and C4S1 for (1,5,17, and 3) wells it means the water quality in  south west of Erbil west  is bad in comparing with the water for wells in north and north east of Erbil. Depending  Ayres and Westcot  (1985)  and Doneen classification the water wells in north and north east of Erbil had better water classes in comparing with  the groundwater in south west of Erbil.,. A scree plot for eigenvalues recorded in this investigation showed the pronounced change of slope after the third eigenvalue. Eigenvalue close or greater than one (unity) were  three factors for north and north east of Erbil and two factors for water of south west of Erbil due to its higher EC value.

Keywords: Groundwater; water classification; Principal component analysis; Erbil.  

* Part of MSc. thesis for the 1^st author.

I. Introduction

Groundwater regards as a source of water resources in Iraq especially in Iraqi Kurdistan region since the irrigation projects are not existing in this part of Iraq except very small irrigation projects in some villages [1]. Groundwater has a significant role in agricultural uses and drinking purpose in rural areas [2]. The groundwater quality is simply the result of the geology and hydrology of the area, types  and chemical composition of rocks, weathering in the source area, and final mineral composition of the sediments are the main factors controlling the  quality and chemical composition of the studied water [2]. Groundwater quality is varied from location to other depending on geological formation of catchment area, chemical composition of aquifer, environmental condition …etc. [3,4]. The large basin of ground water is existing in Erbil governorate in comparing with the area of groundwater basin in other governorates which equal to more than (5000 km²) and the number of drilled wells is  about (10000) wells [1]. Research’s [5] and [6] indicated that the advancement in technology caused  an increase in drilling wells and use of groundwater reached too deep of aquifers to obtain large amount of water, the quality was classified depending on the main chemical properties for irrigation as follow:

1. 1.  Classification of water for irrigation uses:

The  most important global systems for  irrigation water classification was summarized as follow:

1.1.  Richards or USDA classification (1954):

This classification depending on electrical conductivity and sodium adsorption ratio (EC and SAR), separately classified into four classes as shown in Table 1,  and classified into 16 classes when it depending on  both EC and SAR Table 2.

Table 1.Richards classification for irrigation water (1954)

SAR =  ……………  When: SAR = Sodium adsorption ratio

Na = Sodium concentration, Ca = Calcium concentration, Mg = Magnesium concentration (mmol L^-1)

1.2.  Doneen classification (1954):

 Depending on Doneen [7] the irrigation water  were classified as shown in Table 2.

Table 2. Doneen classification of water.

1.3.  Wilcox classification (1955):

This classification of water depending on residual sodium carbonate (RSC), which classified into three classes Table 3.

    Table 3. Water classification depending on RSC.

1.4.   Ayers and Westcot classification (1985).

Ayers and Westcot[8] depended on EC (dS.m^-1), SAR of the soil, in classification of the irrigation water into three classes Table 4.

Table 4. Ayers and Westcot irrigation water classification (1985).

1.5. 
1.5. Don classification (1995):

This classification based on EC, SAR, Na % and pH, the water classification classified into five classes which explained in Table 5.

Table 5. water classification based on TDS, EC, SAR, Na% and pH.

       The groundwater quality for a limited area was conducted by numerous workers such as [3] and [9-12]. Since none of the studies included wide area of Erbil governorate and none of them depended on principal component analysis for comparing between water qualities for this reason this study was done to:

1-       Classification the groundwater in north and south Erbil governorate for irrigation.

2-       Using Principal component analysis for water evaluation.

II. Materials and Methods:

2.1.   Study area:

The study was conducted during the dry season of 2019; the samples were taken from 22 of June to 2 of July 2019 from 52 locations in agricultural lands of Erbil governorate Iraqi Kurdistan region which including (North and north east of Erbil (26) and south west of Erbil(26) samples.  The GPS reading for the studied locations were recorded in table (2-1) and 2-2).

2.2.   Water sampling:

Water samples were collected from (52) wells as mentioned above the depth of the wells was between (150 – 300) m, and the water were taken by using plastic bottle of 330 ml. The water samples were kept in refrigerator at (4 ^oC) and send it to the laboratory for analysis.

2.3.   Water chemical analysis:

Various chemical analyses were analyzed as follow:

1. Electrical conductivity (EC), Potential Hydrogen (pH) and temperature of the water samples were recorded at the site by using EC, pH and temperature meter Model (HI 9814) (APHA,2009)..
2. Cations and trace element of water sample were analyzed in the laboratory by using inductive coupled plasma optical emission spectrometer 2100DV (ICPOES 2100 DV), the cations were (Ca²⁺, K⁺, Na⁺ and Mg²⁺ ).
3. system (ICS 1500), (Dionex ICS 1500), which were (.SO₄^2-, Cl^-, SO₄^2- and HCO₃^-).

III. Statistical Analysis:

 The statistical analysis was conducted using SPSS program version 22 and XLSTAT,2016.

Table 6. GPS reading of the south west of Erbil.

Table 7. GPS reading of North and north east of Erbil.

The chemical variables of properties of the studied water samples and scientific parameters were recorded in table (8,9,10 and 11) and the parameters were calculated depending on the following equations:

Soluble sodium percentage (SSP).

SSP =  X 100. ………… (1)

Sodium adsorption ratio (SAR).

SAR =    ……………….(2)

Ion concentrations must be expressed in mmol_c L^-1(meq L^-1) to calculate SAR using above equation.

Adjusted sodium adsorption ratio (Adj. SAR).

Adj.SAR =  {1+ (8.4 - pHc)}   …………….. (3)

Where:

Na⁺, Ca²⁺, Mg²⁺ = Sodium, calcium, magnesium concentration in meq L^-1 pHc = (pK₂ - pKc) + p (Ca²⁺ + Mg²⁺) + p(ALK) 

(pK₂ - pKc) = The tabulated value for sum of concentrations of

                      (Ca²⁺ + Mg²⁺ + Na⁺) meq L^-1.                                         

pKc = The solubility product of CaCO₃ .

pK₂= Second dissociation constant of calcium bicarbonate.

p(Ca²⁺ + Mg²⁺) = the tabulated value for sum of the concentration of 

                           (Ca²⁺ + Mg²⁺) meq L^-1                                             

p(ALK) = the tabulated value for sum of the concentration of

                (CO₃^2- + HCO₃^-) meq L^-1                                         

Adjusted sodium adsorption ratio (Adj.RNa).

Adj. RNa =    ……………….. (4)

Where;

        Na⁺   = Sodium concentration in meq L^-1.

        Mg²⁺ = Magnesium concentration in meq L^-1.

        Ca_x²⁺= A modified calcium value taken from table adapted from   

                   which depends on EC and   ratio.

Salinity potential:

Salinity potential (SP) =Cl^- +1/2 SO₄^2-(mmolc L^-1) or meq L^-1

Table 8. Some chemical properties of the studied groundwater in North and East of Erbil province.

Table 9. Shows some studied groundwater parameter in North and East of Erbil province.

Table 10. Some chemical properties of the studied groundwater in south east of Erbil province.

Table 11. Shows some studied groundwater parameter in south west Erbil province.

1. IV.  Results and Discussion:

     The minimum, maximum, mean and standard deviation  for the studied variables of the studied water samples at both locations in Erbil governorate were recorded in Table 12 which prepared from Tables 8,9,10 and 11) which were depended in  both classification of  the studied water for irrigation and principal component analysis(PCA) .

Table 12. Shows the values of the studied variables at two locations .

        The statistical analysis indicated to significant difference at level of significant 0.01 between the chemical properties of the studied water samples for the studied wells of north and south Erbil (Table 8 and 10). Similar results were obtained by [11].

3.1.  Water classification for irrigation:

       The studied water was classified according to some global classification depending on Table (8 and 10) for north and north east of Erbil and depending on Table (9 and 11) for south west of Erbil as follow:

3.1.1.   Recharded classification (1954):

The studied water according to this classification were classified as shown from table below. It is appearing that the water of (20 and 6) locations or wells having C₂S₁ and C₃S₁ Class respectively. It means the water class are differing depending on EC value, while they were similar in their class depending on SAR value, since the SAR value of the water in Kurdistan region is low due to high concentration of calcium and magnesium and low concentration of sodium as shown from (3.2) . This result agrees with those recorded by [3].While for the south east Erbil wells the water classes were C1S1, C2S1 , C3S1 and C4S1 for (1,5,17, and 3) wells it means the water quality in  south west of Erbil west  is bad in comparing with the water for wells in north and north east of Erbil .this may be due to the differing in geological formation of the studied two locations and the existing of upper Fars formation in south west of Erbil and Bakhtyari formation in north and north east of Erbil which caused increase in electrical conductivity  and soluble salts in south west of Erbil in comparing with north and north east of Erbil (table,2.3 and 2.5).Similar results were obtained by [3].

3.1.2.   Classification of Wilcox (1954):

Depending on Wilcox (1954) the water of all the studied wells  for both (north and north east and South west) of Erbil having safe class since  the residual sodium carbonate (RSC) below 1.25 mmolc L^-1 or having negative values Table 13. This may be due to  high concentration of calcium and magnesium  in the studied water sample in comparing with the concentration of carbonate and bicarbonate .It is appear from  Table 13  that  the negative values for RSC were  higher  in south west of Erbil in comparing with its value at north and north east of Erbil (Table 13) due to the higher  concentration of calcium ,magnesium , in south west of Erbil. These results agree with [11].

3.1.3.   Ayers and Westcot (1985) classification:  

Depending on some parameters the studied waters were classified according to Ayers and Westcot (1985) as shown from Table 13 In general, the quality of water in north and north east Erbil is better than the studied waters in south west of Erbil. This may be due to differing in the geological formation of the studied locations.

Table 13. Explains water classes according to Ayers and Westcot (1984).

3.2.  Factor analysis:

   In principal component analysis (PCA) analysis the number of components is equal to the number of variables. The eigenvalues was depended to determine the number of (PC) that can be retained for further study. A scree plot for eigenvalues recorded in this investigation showed the pronounced change of slope after the third and second eigenvalues ( Table 14 ,Figure,1and 2) for north and south west Erbil respectively. Eigenvalue close or greater than one (unity) which explained (59.58,26.47 and 7.43) and (71.50 and 15.2) % of the total variation respectively for both locations respectively (Table 14). it means in the north and north east of Erbil three factors had significant loading or influence(Table14) while in south east two factors (F1 and F2) had significant effect on water quality due to higher EC values in south west of Erbil. The comparison between these two factors (F1 and F2) is necessary. It is appear   from eigen values and variability% that the F1 is caused 59.58 and 71.50 %  in water properties in (north and north east) and (south west) of Erbil respectively due to higher EC and ionic concentration in south east of Erbil in comparing with  north and north east of Erbil.  information contained in the original database and Table 14).The cumulative variation of these two factors was( 80.05 and 85.6%) for both locations respectively ,while the third factor (F3) also significant in north and  north east location which caused increase in cumulative 5 variability to93.48%. the variation effect of other factors can be neglect (F3 to F16 ) and (F4 to F16) for both locations respectively . Since  the number of samples were26 samples per location with the degrees of freedom (df) =24 for this reason the correlation coefficient values equal or more than 0.47 is significant and contribute in factors or had significant effect on factors (table,3.3) The locations of the studied parameters  or vectors in Figure (3 and 4) depends on positive or negative correlation between parameters and the factors in Table 14. The parameters which were located in quarter one of the circle it means the correlation of them with both F1 and F2 are positive significantly as shown in Table 15 and their locations depends on the  values of correlation coefficient , also if the angle value between two vectors  equal or more than 90 degrees it means there are significant correlation between them, and the angle greater than 90 degrees indicates to non-significant difference between them. In general the location of vectors in Figure (3 and 4) depends on the sign of factor(F I and F2).

Fig.1.  Scree plot for water parameters in north and north east of Erbil.

Fig.2. Scree plot for water parameters in south west of Erbil.

Table 14. The eigenvalues, variability and cumulative variability for water of wells at both locations in Erbil.

Fig.3. Plot vectors for water of north and north east of Erbil.

Fig.4. Plot vectors for water of south west of Erbil.

*Values  means the significant relation between variables and loading factors.

*Values  means the significant relation between variables and loading factors

For Table 15, F1 =EC, pH, pHc, Ca²⁺ ,K⁺, Na⁺, Mg²⁺ ,SO₄^2-, HCO₃-, Cl-, , RSC, SAR ,Adj.SAR , Adj.RNa and SP.

F2 = Mg and SAR.

Other factors can be neglect since there is no significant correlation or loadings among variables and (F3 to F16).

 For  table (3.4) F1 = EC, pHc, Ca²⁺ ,K₊, Na₊, Mg²⁺ ,SO₄^2-, HCO₃^-, Cl^-, RSC, SAR,Adj.SAR , Adj.RNa and SP.

F2=pH, Na, RSC, SAR, Adj.SAR, Adj.RNa and SSP.

F3 =pH, and K.

 Other factors can be neglect since there is no significant correlation or loadings among variables and (F4 to F16).

Conclusion

        The results indicated that the studied groundwater are differing in their quality among wells or locations and depending on the applied global classification for irrigation water. In general, the water  quality in north and north east of Erbil is better than those in south west of Erbil. The principal component analysis (PCA)  regards as a best method for studying and limiting the factors depending on their significant effect and importance. Recommendations:

Depending on the obtained results the following recommendations were created:

1. Classification the groundwater for different uses.
2. Comparison between cluster analysis and Principal component  analysis in future studies.