Spatial distribution and water quality index of Lahore Canal, Pakistan

Surface water in developing countries like Pakistan is constantly added with pollutants through various sources like domestic and industrial wastewater and animal excretions (Asghar 2018). Such pollutant addition is deteriorating the quality of surface water. In addition, the water ecosystem is also disturbed and somehow destroyed. Therefore, water management is necessary. Precautionary planning, prohibition, and restriction for wastewater discharge, levies, charges, and awareness campaigns, are steps that need to be taken (Jalees et al. 2021a). A recent study by the Pakistan National Council of Research showed that more than 70% of Pakistan's drinking water is unsuitable for drinking based on chemical, physical, and biological analysis (Azizullah et al. 2011). Surface water (canals and rivers) is a significant source of groundwater recharging through seepage (Jalees et al. 2021a). Polluted canals and rivers are also responsible for the deterioration of groundwater. The Lahore branch canal is one of the sources of recharging of groundwater in Lahore City, Pakistan (Mumtaz et al. 2010, 2015).

Lahore Canal (LC) is present in Lahore city, the second largest city in Pakistan. It is used as irrigation water for 4,00,000 acres of agricultural land. The LC water is saline, having a 55 ft water table. The canal passes through Lahore city's centre, a congested residential and commercial area. The people from these commercial and residential areas continuously discharge the wastewater into LC without any treatment. Although the local Government issued the regulation (Law and Parliamentary Affairs Department, G. o. P. (2016)), the Government failed to implement it properly, which resulted in the deterioration of the quality of LC water (Mumtaz et al. 2010, 2015). The LC is approximately 60 km long, east of Lahore city. The detail of LC is given in Table 1. This study deals with the pollutants present in LC. The concentration of pollutants (17 parameters) was used for statistical modelling and to determine Water Quality Index (WQI). In addition, the spatial distribution of selected pollutants was also studied to understand the impact of pollutants better.

The samples were stored in the ice chest and transported to the laboratory for analysis. The chemicals and solutions used in the analysis are of analytical grade (purchased from Merck, Pakistan). The analysis details can be found in the reference, whereas the method number and parameter details are given in Table 3. Control and blank samples and spiked samples analysis were also performed for quality assurance. The samples were run in triplicate, and the average values were used for further analysis. The concentrations of pollutants were used for the source and correlation purposes. Descriptive statistics, Pearson correlation, and analysis of variance (ANOVA) are the well-established models used in this study for correlation, whereas principal component analysis (PCA) and cluster analysis (CA) were used for the source origin. WQI was used to categorize LC water for any usage, and the spatial distribution of pollutants was used to understand the impact of urbanization. All mathematical models and calculations were performed using Minitab^® Software.

The samples collected from LC were analysed in triplicate. The results obtained were then subject to statistical analysis for correlation. In addition, spatial distribution and WQI were also studied.

The total suspended solids increased along the canal from 95 to 295 mg/L (Table 4, Figure 2) due to throwing floating particles or natural storm dust particles (Nishanthi et al. 2021). Total dissolved solids (TDS) are the main contributor to pollution in the water body. The values of TDS were between 79 and 535 mg/L (Table 4), which is within guideline values, i.e., 1,000 mg/L (Panhwar et al. 2022). The total settled solids (TSS) ranged from 0.1 to 1.0 g/L. Values of TSS increase along the canal, the same as TDS. The values of total solids are the sum of all solids, and obviously, it follows the same trend as TDS and TSS, with values ranging from 400 to 992 mg/L. turbidity of the samples is directly associated with suspended and settleable solids. The turbidity of LC is in the range of 8.8–195 NTU. The high value of turbidity is because of silt, organic debris, and soil. Although the values were high, it is within NEQ (National Effluent Standards), i.e., 25–50 NTU (Azizullah et al. 2011).

The LC samples were analyzed for heavy metals, i.e., Pb, Cr, and Cu. The values of metals were beyond the NEQ (Asghar 2018). The concentration of Pb was 0–5.71 mg/L, which was way above the guideline value of 0.5 mg/L. When LC water is used for irrigation, such high Pb values will deposit in plants and pose a cancer risk (Jalees et al. 2021b). The values of Cr in LC samples were 0.02–1.04 mg/L, slightly above the guideline values of 1 mg/L (Asghar 2018). Such high Cr contents can pose a cancer risk if LC is used for irrigation. The values of Cu were in the range of 0–3.5 mg/L, which is above the guideline values of 1 mg/L. Regarding irrigation water criteria, the heavy metals are above the guidelines values, making LC water unsuitable for irrigation (WWF-Pak 2007). The heavy metal trend and source of pollution will be discussed in the coming section.

To access the information about the source and correlation of various pollutants, statistical tools were used, i.e., Pearson correlation, ANOVA, PCA, and CA. Average values of all tested parameters were used for this.

The Pearson correlation was used to get information about the probable source of pollutants. It is a regression analysis used to understand the correlation of experimental data (Jalees & Asim 2016). The results of the Pearson correlation are given in Table 5. The coefficient (r) > 0.7 indicates a strong relationship; r = 0.3–0.7 indicates a moderate relationship, while r < 0.3 indicates a poor one. The positive and negative signs indicate a direct and inverse relationship, respectively (Jalees & Asim 2016). A strong positive relationship indicated that the pollutant may have originated from the same source. In Table 5, a positive moderate to poor correlation was observed in SS with pH & TDS and vice versa; in turbidity with pH, TDS, SS, and TS; in hardness with Pb, TDS, and SS. The negative moderate to poor relationship is observed in TDS with pH; in DO with TDS, TS, and turbidity; in alkalinity with Pb and TDS. These findings suggested a hypothesis that the parameters (pollutants) originated from the same source. This hypothesis was further studied using ANOVA.

To evaluate correlations between the pollutants, a null hypothesis was formulated as per the literature (Jalees & Asim 2016). The null hypothesis statement was ‘no correlation among the parameters studied in LC, i.e., r = 0’. Contrary to it, an alternative hypothesis was also formulated, which stated that ‘there is some correlation between the pollutants in LC, i.e., r ≠ 0″. The results of ANOVA (single tail) are given in Table 6. It generated two parameters, i.e., ‘F_crit’ and ‘p.’ If p>F_crit then the values are not significant and if p<F_crit then the values are significant. Significant values further suggested that pollutants/parameters correlate with them, i.e., r ≠ 0 (Jalees & Asim 2016). The values in Table 6 showed that the p<F_crit, which means correlation among the pollutants and hence the results of Pearson analysis were confirmed.

The PCA is an imaginary eigenvalues-based system. In this system, the eigenvalues < 1 are ignored, and greater values > 1 are grouped based on the same source. The data are divided into groups known as components. The component which explains most of the results is termed as 1st component of PC-1, the 2nd component in line is termed PC-2, and so on (Nguyen et al. 2020). Parameters having the same source or origin are grouped in the same component. The results of PCA are given in Table 7. PCA obtained three components, i.e., PC-1, PC-2, and PC-3. Collectively these three components explain 55.05% of the variance. PC-1 explains 29.88%, PC-2 explains 14.14%, and PC-3 explains 10.91% of the total variance. The PC-1 showed strong loading moderate to strong loading for pH (0.5), TS (0.5), TDS (0.71), Turbidity (0.86), DO (-0.83), and Fe (0.72). It suggested that the source of these parameters may be the same (Jalees & Asim 2016). PC-2 showed moderate to strong loading for SS (0.62) and Pb (0.76), while PC-3 showed moderate loading for TS (0.63) and Cu (−0.62). The parameters with moderate to strong loading and the same PC are assumed to originate from the same source (Jalees & Asim 2016). To verify this, CA was performed.

The values of WQI of LC were compared with the literature data. Many authors from India, China, Iran, and Iraq have studied the rivers, canals and lakes for WQI. The data showed that the water quality of LC is severely deteriorated as compared to other literature-available data (see Table 9). It suggested that the Government should immediately consider the remedies for LC.

The physical and chemical parameters of LC were analyzed. The results showed that the physical parameters were within the NEQS of Pakistan. The low DO and high BOD values indicate that low oxygen conditions are dominant in LC. The metal concentrations were beyond Pakistan's NEQ and irrigation water standards. The highest concentration was Ca, i.e., 75 mg/L, and the lowest was for Cu, i.e., 1 mg/L. The metals concentration shows the following trend Ca > Fe > Pb > Cr > Cu. The Pearson correlation coefficient showed a moderate to strong correlation as values were within 0.3–0.7, suggesting that the sources of pollutants were the same. The ANOVA gives F_crit > P, supporting the hypothesis of a similar pollutant origin. The loadings of PCA were three components with a total variance of 55%, whereas the CA also gave three groups based on Euclidean distance, confirming the hypothesis that the origin of pollutant is from the same source, i.e., small household industries in the study area. The WQI of LC showed values above 300 at all locations, which indicated that the LC water is unsuitable for any usage. The spatial distribution of pollutants shows variation along LC, although the variation was associated with population density and urbanization. Worse conditions were present at L-4, which is highly populated and has an abundance of small industries. The Government may focus on implementing rules and regulations for water quality; otherwise, the environmental conditions will become more severe.

The authors are thankful to the University of Engineering and Technology, Lahore, for providing the necessary facilities for study.

No ethical approval is required for this manuscript.

All authors have given their consent about the content of this manuscript.

All authors have been permitted to submit and publication of the manuscript.

Muhammad Irfan Jalees: Original Idea and manuscript writing; Iffat Irfan: Experimental and result compilation; Asif Ali: initial draft; Madeeha Batool: Manuscript review

There was no funding for this research

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.