This study aimed to assess spatiotemporal water quality variation and its suitability for irrigation and domestic purposes in Lah River using the irrigation water quality index (IWQI) and the weighted arithmetic water quality index (WAWQI). The IWQI analysis result showed that the sodium absorption ratio, residual sodium carbonate, potential salinity, Kelly index, magnesium ratio, sodium percentage, and permeability index were found to be 1.07 mEq/L, −0.43 mEq/L, 0.8 mEq/L, 0.78 mEq/L, 43.01%, 42.95%, and 63.46%, respectively. The IWQIs revealed that the water quality of the river was appropriate for agricultural use during the dry season. Furthermore, the calculated WAWQI of the river water ranged from 123.13 to 394.72 during the wet season, indicating the high pollution levels in the Lah River and incompatibility for drinking purposes. On the other hand, the principal component analysis identified two pollution sources during the wet season and three during the dry season. In addition, the positive matrix prioritization model predicted the pollution source's contribution quite well with a signal-to-noise ratio of >2 and a residual error between −3 and 3 for both seasons. This study suggests that water quality of Lah River is degrading periodically necessitating proper pollution management.

  • In this study, water quality and its suitability were assessed using multivariate statistical techniques and receptor models conjunctively.

  • The middle segment of the Lah River is more polluted than the upstream and downstream sections necessitating coordinated intervention measures.

  • During the dry season, the Lah River is suitable for irrigation.

  • Integrated water quality index and multivariate statistical techniques can be used for preliminary pollution management in Lah River.

These days, the disposal of urban wastes and untreated effluents from various industries, agriculture, and houses are continually increasing the pollution load in surface water resources and deteriorating water quality (Ali et al. 2016; Angello et al. 2021). In addition, urban rivers in developing countries also carry organic and inorganic nutrients, which are a threat to the local communities (Hamid et al. 2020) and have implications for aquatic ecology as well as human health (Khouni et al. 2021). In general, poor water quality directly impacts human health, treatment costs, and availability of safe drinking water. Therefore, ensuring drinking water safety is a basis for the prevention of waterborne diseases, which signifies the importance of assessing the spatiotemporal variability of water quality (Shishaye & Asfaw 2020).

Despite the high socioeconomic importance of rivers in urban areas, their increased exposure to pollution sources has led to environmental threats (Chen 2017). However, many countries have set environmental guidelines that limit the number of pollutants that are released into the water resources based on the self-purification capacity of the receiving water body (Chinyama et al. 2016). The legal enforcement in developing countries is, however, very limited partly due to complicated pollution sources (Xu et al. 2019). Although a number of factors trigger the pollution of these rivers, high urbanization, poor sanitation infrastructures, poor public understanding on waste handling, and poor monitoring and inspection are some to mention.

Previous studies on water quality assessment in developing countries involve complex processes where the management plans are based on personal judgement and experiences. Despite the availability of some of the water quality data, its handling, management, and interpretation have remained bottlenecks. According to Özdemir (2016), accurate characterization of water quality having large monitored data is often difficult for two main reasons. First, the collected data are large leading to difficulty in management and data distortion. Second, the operational cost for the effective handling and processing of these large data is high. However, the introduction of statistical tools through the use of multivariate statistical techniques (MSTs) has become very successful for such cases in multiple dimensions (Barakat et al. 2016). Principal component analysis (PCA) and factor analysis (FA) are two common MSTs and are multidimensional data analysis approaches widely used to determine potential pollution sources and statistically determine independent source tracers (Li et al. 2015). In recent days, the application of MSTs has become very common in water quality management and pollution control programs such as the determination of spatial and temporal variability of pollutants (Wang et al. 2012), identification of pollution sources (Huang et al. 2010) and identification of pollution hotspots (Bushero et al. 2022). PCA has the potential to provide valuable qualitative information about pollution sources. On the other hand, multivariate receptor models (MRMs) such as positive matrix factorization (PMF) (Gholizadeh et al. 2016) and the UNMIX model (Angello et al. 2021) have become useful in providing quantitative information on the contribution and composition of pollution sources (Hossain et al. 2015). In addition, the introduction of techniques such as the use of the water quality index (WQI) for the determination of pollution statuses of surface water sources has become very effective without much information loss. WQI is a single dimensionless value that aggregates the measurements of various parameters to express water quality in a much simpler form. They are very useful instruments for assessing and managing river water quality and are effective tools for communicating water quality information to the concerned policymakers (Tokatli 2019).

In Ethiopia, river pollution specifically in urban areas is associated with the uncontrolled expansion of cities, agricultural practices in the catchments, poor urban runoff management, and improper waste release into the streams (Bushero et al. 2022). In addition, poor wastewater treatment and the absence of treatment plants are impacting the river's water quality (Yohannes & Elias 2017). In addition, industrial wastewater is also identified as the most polluting source of surface water resources in the country (Yilma et al. 2018). This study is conducted with the aim of assessing the pollution status in the study area based on integrated MSTs, MRM, and WQI.

Description of the study area

The study area (Finote Selam) is located in the West Gojjam Zone of the Amhara Regional State, Ethiopia, and is bounded between 10,039′00″–10,043′00″ N and 37,013′20″–37,018′00″ E having an elevation of 1,917 m above mean sea level. Lah River is one of two rivers crossing the town, draining from north to south. On the other hand, the temperature of the study area typically varies from 11 to 30 °C and is rarely below 8 or above 33 °C. The area is characterized by both dry and rainy seasons. The minor rainy season occurs during April and May and the major rainy season from June to September. Based on the Ethiopian agro-ecological classification, the Lah River catchment is categorized under the Woina Dega zone having an annual rainfall of 1,450.3 mm (Figure 1).
Figure 1

The study area.

Site selection and water sampling

For this study, the water samples were collected during both dry (December and January) and rainy seasons (July and August) from 16 sampling sites using grab sampling (Figure 2). Four sampling sites from undisturbed points (S1, S2, S3, and S4) and nine sampling sites within the urban boundary (S5, S6, S7, S8, S9, S10, S11, S12, and S13) with extensive human activities are characterized by excessive liquid waste effluent discharge, car-washing, solid waste disposal, open defecation, open bathing, and cloth washing. The sampling sites (S14, S15, and S16) were taken from the downstream end of the river relative to the city, which has less human settlement than midstream and is characterized by high agricultural activities. The sampling sites were selected after preliminary site investigation and are based on factors such as the level of anthropogenic influence, accessibility, water use, and budget. The samples were collected using polyethylene bottles. Before sample collection, the bottles were washed carefully and rinsed three times with sample water. The grab sampling technique was used for collecting the samples and collected at approximately 20–30 cm below the water surface. After collection, the water sample was preserved below 4 °C and transported to the laboratory for subsequent examination according to APHA (2012).
Figure 2

Monitoring stations of the Lah River.

Figure 2

Monitoring stations of the Lah River.

Close modal

Water quality analysis

A total of 14 water quality variables were analyzed on all monitoring stations during the dry and rainy seasons as shown in Table 1. The water quality parameters were selected based on the objective of the study in consideration of its representativeness for the physical and chemical characteristics of the river under study. pH, total dissolved solids (TDS), and electrical conductivity (EC) were measured in situ with a portable multimeter probe (HQ40D, USA), while the remaining were analyzed in the laboratory according to APHA (2012). Accordingly, biochemical oxygen demand (BOD5) is analyzed by the manometric method, whereas phosphate, nitrate, and sulfate are analyzed by the spectrophotometric method. The titrimetric method was used to analyze calcium, magnesium, carbonate, and bicarbonate. All other analytical methods are shown in Table 1.

Table 1

Water quality parameters and their analytical techniques

ParametersInstrument usedMethod
BOD5 (mg/L) BOD incubator and titration assembly Manometric method 
(mg/L) UV-spectrophotometer Spectrophotometric 
(mg/L) UV-spectrophotometer Spectrophotometric 
Cl (mg/L) Titrimetric Argentometric method 
K+ and Na+ (mg/L) Flame photometer Flame photometric 
Ca2+ and Mg2+ (mg/L) Titrimetric Titrimetric 
and (mg/L) Titrimetric Titrimetric 
(mg/L) UV-spectrophotometer Spectrophotometric 
ParametersInstrument usedMethod
BOD5 (mg/L) BOD incubator and titration assembly Manometric method 
(mg/L) UV-spectrophotometer Spectrophotometric 
(mg/L) UV-spectrophotometer Spectrophotometric 
Cl (mg/L) Titrimetric Argentometric method 
K+ and Na+ (mg/L) Flame photometer Flame photometric 
Ca2+ and Mg2+ (mg/L) Titrimetric Titrimetric 
and (mg/L) Titrimetric Titrimetric 
(mg/L) UV-spectrophotometer Spectrophotometric 

Suitability assessment of water for drinking

In this study, the suitability of Lah River's water for drinking purposes was calculated using the weighted arithmetic WQI (WAWQI). This method was chosen because it has been widely used by different scholars to assess the suitability of river water quality for drinking purposes (Lkr et al. 2020). The water quality status standards are presented in Table 2 and it is calculated using Equation (1).
formula
(1)
Table 2

Water quality rating based on WAWQI

WAWQI valueRating of water qualityUsage possibilities
0–25 Excellent water quality Drinking, irrigation 
25–50 Good water quality Drinking, irrigation 
50–75 Poor water quality Irrigation 
75–100 Very poor water quality Irrigation 
>100 Unsuitable for drinking purposes Proper treatment is required before use 
WAWQI valueRating of water qualityUsage possibilities
0–25 Excellent water quality Drinking, irrigation 
25–50 Good water quality Drinking, irrigation 
50–75 Poor water quality Irrigation 
75–100 Very poor water quality Irrigation 
>100 Unsuitable for drinking purposes Proper treatment is required before use 
The quality rating scale (Qi) for each parameter is calculated using Equation (2):
formula
(2)
where Vi is the estimated concentration of ith parameter in the analyzed water; Vo is the ideal value of this parameter in pure water and Vo = 0 (except pH = 7.0) (Goher et al. 2014); and Si is the recommended standard value of ith parameter.
The unit weight (Wi) for each water quality parameter is calculated using Equation (3):
formula
(3)
where K is the proportionality constant and can also be calculated using Equation (4):
formula
(4)

Irrigation water quality index

The suitability of the Lah River's water for agricultural purposes was assessed using the irrigation water quality index (IWQI) where the indices used to determine the suitability were sodium adsorption ratio (SAR), residual sodium carbonate (RSC), potential salinity (PS), Kelly index (KI), permeability index (PI), magnesium ratio (MR), and sodium percentage (Na %) (Chukwuma et al. 2016). A detailed description of the determination of IWQI is presented in Table 3, and the suitability classes are shown in Table 4.

Table 3

Irrigation water quality indices and the corresponding mathematical equation

IndexEquation
SAR  
RSC  
PS  
MR  
KI  
PI  
Na%  
IndexEquation
SAR  
RSC  
PS  
MR  
KI  
PI  
Na%  
Table 4

Classifications of irrigation water quality indices

IndexValueClassIndexValueClass
KI <1 Suitable SAR <10 Excellent 
>1 Unsuitable 10–18 Good 
PS <3 Suitable 18–26 Permissible 
>3 Unsuitable >26 Unsuitable 
MR <50 Suitable PI <25 Unsuitable 
>50 Unsuitable 25–75 Good 
Na % < 60 Safe >75 Excellent 
>60 Unsafe RSC <1.25 Safe 
  1.25–2.5 Marginally unsuitable 
  >2.5 Unsuitable 
IndexValueClassIndexValueClass
KI <1 Suitable SAR <10 Excellent 
>1 Unsuitable 10–18 Good 
PS <3 Suitable 18–26 Permissible 
>3 Unsuitable >26 Unsuitable 
MR <50 Suitable PI <25 Unsuitable 
>50 Unsuitable 25–75 Good 
Na % < 60 Safe >75 Excellent 
>60 Unsafe RSC <1.25 Safe 
  1.25–2.5 Marginally unsuitable 
  >2.5 Unsuitable 

Multivariate statistical analysis

An MST such as PCA/FA is used to analyze large datasets with multidimensional content. They are also used to provide a more detailed and meaningful interpretation of large datasets in water quality without significant information loss contained in original data. In this study, PCA was used to determine the pollution source type contributing to the Lah River where a similar approach has been adopted (Dutta et al. 2018; Fathi et al. 2018). On the other hand, the suitability of the data for PCA was determined based on the suitability test determined by Kaiser–Meyer–Olkin (KMO). KMO values greater than 0.5 are often recommended for the PCA/FA. On the other hand, Bartlett's test having a significance level of under 0.05 implies significant relationships within variables (Yilma et al. 2018). The components extraction was done by using principal components method and varimax rotation with eigenvalue greater than one. For effective interpretation, PCA factor loadings need to be strong enough so that all the contributing pollution sources are well explained. According to Kilonzo et al. (2014), PCA factor loadings >0.75 are classified as strong, whereas 0.5–0.75 and 0.3–0.5 are classified as moderate and weak loading, respectively. The interpretation behind each factor loading depends on the magnitude of the loads after varimax rotation.

PMF receptor model

MRMs such as PMF have been used in many studies for the determination of the contribution of a pollution source based on the identification of source type (Angello et al. 2021). The model has been found to be efficient in assessing the water quality data of interest (Xiao et al. 2020), and it is effectively applied by various researchers in many watersheds (Jiang et al. 2019; Xiao et al. 2020; Wang et al. 2022). The PMF model is used to decompose a matrix X, into factor contributions (G), factor profiles (F), and the residual error (E) of the pollution sources of the known profile (Jiang et al. 2019). It cannot allow zero and negative uncertainty for the species and give the non-negative constraint on the factor contribution rate (Li et al. 2021). In the PMF, the reliability of variables considered is determined using the signal-to-noise (S/N) ratio value, which is the ratio between the concentration of the variable and its uncertainty. According to Celen et al. (2022), data with a low S/N (between 0.2 and 2) are taken as weak, data with an S/N lower than 0.2 are considered as bad, and data with an S/N greater than 2 are considered strong species. In PMF, two input files were considered: the concentrations of the examined water quality parameters and their uncertainty value are calculated using Equation (5) (Xiao et al. 2020) and Equation (6) (Hsieh et al. 2022):
formula
(5)
when C ≤ MDL.
formula
(6)
when C ≥ MDL.
C is the concentration of the parameters and MDL is the method detection limit of each chemical species. The concentration of each parameter (X) is estimated using Equation (7) as:
formula
(7)
where G is factor contributions, F is the source profile, and E is the residual error (Jiang et al. 2019).
The prime aim of the receptor model such as PMF to apply the chemical mass balance between the species in concentration versus the identified pollution source profiles is expressed by Equation (8):
formula
(8)
where p is the number of factors, f is the species profile of each source, g is the amount of mass contributed by each factor to each individual sample, and eij is the residual for each sample/species.
In PMF, factor contributions and profiles are determined by minimizing the objective function (Q) in the PMF model and are given by Equation (9) as:
formula
(9)
where Q is a critical parameter for PMF expressed in two terms: Q(true) is the goodness-of-fit parameter calculated including all points and Q(robust) is the goodness-of-fit parameter calculated excluding points not fit by the model, defined as samples for which the uncertainty-scaled residual is greater than 4.

Hydrochemistry of the Lah River

In the study area, the pH varied slightly between sampling points during the dry (7.45–8.52) and wet seasons (6.6–7.6). A minimum pH was recorded at sampling point S12 partly due to the entry of acidic wastes from local garages, car wash centers, and solid waste disposal, which is also reported in the work of Barakat et al. (2016). On the other hand, the maximum pH was measured at S3 during the wet season. The higher pH could be attributed to the accumulation of a high amount of bicarbonate from its non-perennial streams (El Morhit & Mouhir 2014). The maximum pH during the dry season was recorded at sampling point S14, potentially attributed to high photosynthetic activity, absorption of dissolved carbon dioxide in the water, and low water levels increasing the concentration, which is also stated in the study of Yusuf (2020). Similarly, the TDS varied from 42.81 (S3) to 81.37 mg/L (S12) during the wet season and from 107.15 (S2) to 188.15 mg/L (S13) during the dry season. The measured TDS and EC were higher in the dry season than in the wet season and in the middle section of the river than at the downstream and upstream sections of the river water. Conversely, the concentrations of TDS and EC showed a decreasing trend downstream of the river, especially at sampling point S16 partly due to the increased self-purification process of the river water and the low pollution load at the sampling point. The maximum TDS and EC concentration were recorded at sample point S12 during the wet season, which could be due to the improper disposal of solid waste near the sampling point, and the entry of waste from urban areas, garages, and agricultural lands (Elsayed et al. 2020). Furthermore, the maximum concentration of TDS and EC was detected at sampling point S13 during the dry period attributed to high evaporation rates and the absence of a dilution in the dry seasons. The lowest TDS and EC were measured at sampling point S2 during the dry season. In the present study, the measured TDS and EC were within the permissible limits of FAO (2,000 mg/L and 3,000 μs/cm) and WHO (1,000 mg/L and 2,500 μs/cm) guidelines at all sample points for irrigation and drinking purposes.

Calcium is the most common constituent present in natural water, and its salts are important contributors to the hardness of water. In this study, the mean calcium concentration of the Lah River ranged from 5.06 (S1) to 14.2 mg/L (S12) during the dry season and 6.1 (S1) to 32.1 mg/L (S12) during the rainy season. The maximum calcium concentration was measured during the rainy season rather than during the dry season. Comparatively, the highest calcium concentration was observed in the middle section of the river rather than in the upstream and downstream segments of the river. Studies showed that the concentration of calcium is elevated because of an increase in pollution load by anthropogenic impacts such as domestic sewage, nutrients from an agricultural area, and the presence of organic matter in the water (Bhutiani et al. 2016). On the other hand, the magnesium concentration varied between 5.5–17.54 mg/L and 1.3–7 mg/L during the wet and dry seasons, respectively. The highest Mg2+ ion was observed during the wet season rather than in the dry season. Higher Mg2+ concentration was observed at sampling point S13 during the dry season. This could be due to the presence of clay minerals in the river water that contain high magnesium ions. On the contrary, a lower magnesium concentration was measured at the upstream and downstream sections of the river, especially at sampling points S1 and S16 during both seasons. According to Dubey & Ujjania (2016), this may be due to the less anthropogenic activity, its accumulation in the bottom deposits, and the high dilution effect of rainwater. On the other hand, the chloride concentration in the study area was observed at sampling points S12 (20.1 mg/L) and S15 (40.13 mg/L) during the wet and dry seasons, correspondingly. Higher chloride concentration was measured in the middle section of the river rather than the upstream and downstream sections of the river. In addition, the highest values were recorded during the dry season than the rainy season. Higher chloride concentration was measured at sampling point S15 during the dry season mainly due to the entry of human waste and livestock manure into the river.

Phosphate in rivers is often associated with runoff from agricultural lands applied as fertilizers and domestic waste discharged households (Khan & Wen 2021). The mean phosphate in the Lah River ranged from 0.24 to 0.82 mg/L and 0.102 to 0.36 mg/L during the rainy and dry seasons, respectively. Relatively, the concentrations were higher in the middle section of the river than the upstream and downstream sections of the river. The phosphate concentration was decreased along the downstream side of the river, especially at sampling point S16 due to the self-purification process of the river water. In the study area, the highest phosphate concentration was observed at sampling point S13 during the rainy season potentially due to the washouts of waste from the nearby abattoir and agricultural land. Similarly, Angello et al. (2021) described that the highest proportion of in the rainy season was observed at Little Akaki River in Ethiopia as a result of agricultural waste from agricultural land.

The organic pollution load is one of the determining factors for the pollution load in water resources. They are mainly attributed to anthropogenic influences. In the study area, the concentration of BOD5 showed a decreasing trend along the downstream end of the river due to the improved self-purification capacity of the river. The measured BOD was found above the acceptable limits of WHO guideline standard (5 mg/L) for drinking purposes during the wet season, while all sampling points were above the permissible limit except at sampling points S1, S4, and S16 during the dry season. On the other hand, the maximum BOD concentration was recorded at sampling points S12 (32 mg/L) and S13 (15.57 mg/L), whereas the minimum lower concentration was observed at sampling points S1 (10.4 mg/L) and S4 (3.15 mg/L) during the wet season and dry season, respectively. Studies revealed that the high organic pollution load is predominantly contributed by factors such as improper disposal of solid waste, open defecation, leaves, woody debris, dead plants, and animal manure.

The nitrate concentration in the Lah River ranged from 1.35 mg/L (S1) to 6.61 mg/L (S14) and from 10.1 mg/L (S3) to 30.2 mg/L (S13) during the dry season and wet season, respectively. However, the concentration of was within the permissible limit at all monitoring stations for drinking purposes as set by WHO guidelines (50 mg/L). Generally, the determined values of nitrate were higher in the middle section of the river during the rainy season. The elevated concentration of nitrate was observed at sampling point S12 during the rainy period and could be associated with the agricultural runoff that carries nitrogen-containing fertilizers from nearby farmland, and the entry of domestic sewage from urban areas, which was also explained in the study of Wondim et al. (2016).

In the study area, Na+ concentration varied from 5.8 mg/L (S3) to 16.4 mg/L (S13) and from 8.3 mg/L (S1) to 27.42 mg/L (S13) during the wet and dry seasons, respectively. The highest sodium concentration was observed during the dry season rather than in the wet season. These values were within the permissible limit of WHO (200 mg/L) and FAO (919 mg/L) guidelines set for drinking and irrigation purposes, correspondingly. The elevated value of sodium observed at sampling point S13 during the wet season was probably due to the entry of agricultural wastes from the agricultural land with runoff into the river water. In addition, the highest sodium concentration was observed in the middle part of the studied river section, especially at S13, and potentially related to the entry of sewage from urban areas, which was also proposed in the works of Xiao et al. (2020). On the other hand, the sulfate of the study river ranged from 12.08 to 38.97 mg/L during the wet season and from 4.2 to 13.95 mg/L during the dry season. The highest value was measured in the middle section of the river rather than in the downstream and upstream sections of the river especially at sampling point S12 during the rainy season and at sampling point S11 during the dry season. The maximum sulfate concentration was observed at sampling point S12 during the rainy season partly related to the discharge of sulfate-containing municipal sewages from the city and organic fertilizers from agricultural activities undertaken on the riverside. A similar phenomenon was reported in the study of Gupta et al. (2021).

The potassium concentration of the Lah River was found to range from 1.6 to 5.4 mg/L during the rainy season and from 1.4 to 4.9 mg/L during the dry season. When comparing the wet period analysis with the dry period, higher concentrations were measured in the wet period at all sampling points. Relatively the highest value of potassium was recorded at the middle section of the river than at the downstream and upstream sides of the river. The highest value observed during the rainy period at sampling point S13 may be due to runoff from the nearby agricultural lands and weathering of rocks from the surroundings, which was also reported in another study (Elango 2018).

Suitability of Lah River for domestic (drinking) purposes

As shown in Table 5, during the wet season, the WAWQI in the Lah River ranged from 123.1 to 394.8 indicating that the river water quality is unsuitable for drinking purposes at all sampling points. On the other hand, during the dry season, WAWQI in the Lah River water quality status has varied from poor (51.9) to unsuitable (173.5) classes. The spatial variances of both WAWQIs were shown by an inverse distance weightage (IDW) map (Figure 3). According to the geospatial interpolation maps for WAWQI, the river water throughout the whole watershed was not fit for human consumption during the wet season.
Table 5

Classification of the Lah River water quality status according to WAWQI

Sampling pointsWet seasonStatusDry seasonStatus
S1 145.4 Unsuitable 51.9 Poor 
S2 193.0 Unsuitable 57.3 Poor 
S3 123.1 Unsuitable 63.0 Poor 
S4 140.9 Unsuitable 68.7 Poor 
S5 212.2 Unsuitable 57.0 Poor 
S6 182.0 Unsuitable 77.0 Very poor 
S7 245.5 Unsuitable 85.3 Very poor 
S8 172.4 Unsuitable 98.2 Very poor 
S9 234.6 Unsuitable 135.7 Unsuitable 
S10 257.8 Unsuitable 173.5 Unsuitable 
S11 292.8 Unsuitable 120.6 Unsuitable 
S12 371.6 Unsuitable 147.4 Unsuitable 
S13 394.8 Unsuitable 86.5 Very poor 
S14 316.2 Unsuitable 78.6 Very poor 
S15 274.1 Unsuitable 62.9 Poor 
S16 213.4 Unsuitable 53.3 Poor 
Sampling pointsWet seasonStatusDry seasonStatus
S1 145.4 Unsuitable 51.9 Poor 
S2 193.0 Unsuitable 57.3 Poor 
S3 123.1 Unsuitable 63.0 Poor 
S4 140.9 Unsuitable 68.7 Poor 
S5 212.2 Unsuitable 57.0 Poor 
S6 182.0 Unsuitable 77.0 Very poor 
S7 245.5 Unsuitable 85.3 Very poor 
S8 172.4 Unsuitable 98.2 Very poor 
S9 234.6 Unsuitable 135.7 Unsuitable 
S10 257.8 Unsuitable 173.5 Unsuitable 
S11 292.8 Unsuitable 120.6 Unsuitable 
S12 371.6 Unsuitable 147.4 Unsuitable 
S13 394.8 Unsuitable 86.5 Very poor 
S14 316.2 Unsuitable 78.6 Very poor 
S15 274.1 Unsuitable 62.9 Poor 
S16 213.4 Unsuitable 53.3 Poor 
Figure 3

IDW map for WAWQI during dry and wet season.

Figure 3

IDW map for WAWQI during dry and wet season.

Close modal

Suitability of Lah River for irrigation purposes

Sodium adsorption ratio

In this study, the SAR varied from 0.66 to 1.50 mEq/L during the dry season (Table 6) where the highest and lowest SAR were observed at monitoring stations S13 and S9, respectively. Accordingly, the calculated SAR at all locations revealed that the irrigation water was under the excellent class.

Table 6

Water quality classes of Lah River for irrigation use during the dry season

Sampling pointsSARRSCPSKIMRNa%PI
S1 0.85 −0.16 0.48 1.00 29.96 47.06 77.52 
S2 1.01 −0.25 0.59 0.97 39.89 46.83 76.23 
S3 1.04 −0.25 0.75 0.96 49.70 47.60 78.17 
S4 1.29 −0.49 0.82 0.98 49.71 46.87 71.09 
S5 1.19 −0.64 0.65 0.85 36.27 43.91 63.79 
S6 1.04 −0.26 0.54 0.86 49.88 43.94 62.15 
S7 1.05 −0.32 0.82 0.80 49.62 43.10 61.89 
S8 0.98 −0.31 0.57 0.82 44.03 43.18 62.17 
S9 0.66 −0.87 0.69 0.38 45.25 29.48 45.31 
S10 1.04 −0.40 0.88 0.60 34.91 41.64 59.70 
S11 1.30 −0.46 1.09 0.77 33.03 48.54 69.95 
S12 1.21 −0.99 1.00 0.71 44.39 41.59 55.68 
S13 1.50 −0.96 1.11 0.87 46.36 46.28 60.58 
S14 1.09 −0.63 1.09 0.65 41.79 39.99 57.80 
S15 0.84 −0.91 1.27 0.54 47.75 34.10 50.08 
S16 0.98 −0.39 0.99 0.79 45.75 43.09 63.37 
Sampling pointsSARRSCPSKIMRNa%PI
S1 0.85 −0.16 0.48 1.00 29.96 47.06 77.52 
S2 1.01 −0.25 0.59 0.97 39.89 46.83 76.23 
S3 1.04 −0.25 0.75 0.96 49.70 47.60 78.17 
S4 1.29 −0.49 0.82 0.98 49.71 46.87 71.09 
S5 1.19 −0.64 0.65 0.85 36.27 43.91 63.79 
S6 1.04 −0.26 0.54 0.86 49.88 43.94 62.15 
S7 1.05 −0.32 0.82 0.80 49.62 43.10 61.89 
S8 0.98 −0.31 0.57 0.82 44.03 43.18 62.17 
S9 0.66 −0.87 0.69 0.38 45.25 29.48 45.31 
S10 1.04 −0.40 0.88 0.60 34.91 41.64 59.70 
S11 1.30 −0.46 1.09 0.77 33.03 48.54 69.95 
S12 1.21 −0.99 1.00 0.71 44.39 41.59 55.68 
S13 1.50 −0.96 1.11 0.87 46.36 46.28 60.58 
S14 1.09 −0.63 1.09 0.65 41.79 39.99 57.80 
S15 0.84 −0.91 1.27 0.54 47.75 34.10 50.08 
S16 0.98 −0.39 0.99 0.79 45.75 43.09 63.37 

Sodium percentage

The Na% in this study was found between 29.48 and 48.54% during the dry season (Figure 4(b)) where the highest and lowest Na% were observed at sampling points S11 and S9, respectively, revealing that the Lah River is suitable for irrigation. Moreover, the spatial distribution map of Na% (Figure 4(b)) shows that the water was classified as an excellent class for irrigation use during the dry season.
Figure 4

The spatial distribution of SAR (a), Na% (b), KI (c), RSC (d), MR (e), and PS (f) during the dry season.

Figure 4

The spatial distribution of SAR (a), Na% (b), KI (c), RSC (d), MR (e), and PS (f) during the dry season.

Close modal

Kelly index

The KI in Lah River during the dry season ranged between 0.38 to 0.99 mEq/L (<1) at all stations and is considered appropriate for irrigation use. The spatial distribution of KI indicates the water was appropriate for irrigation application during the dry season as shown in Figure 4(c). The highest and the lowest KI were observed at sampling points S1 and S9, respectively.

Residual sodium carbonate

RSC is taken as one of the efficient tools used for evaluating water quality for irrigation purposes (Aydin et al. 2020). The RSC in this study varied from −0.99 to −0.16 mEq/L during the dry season revealing the Lah River as a good water quality class for irrigation purposes. During the dry season, the highest and the lowest values of RSC were observed at sampling points S1 and S12, respectively (Figure 4(d)).

Magnesium ratio

As shown in Table 6, the MR varied between 29.96 and 49.88% during the dry period. In this study, the MR was also found less than 50% at all sample points. In addition, the IDW map also shows the MR values of the water sample were less than 50% during the dry season. Based on this value, the sampled water from each sampling point was suitable for irrigation purposes. The spatial distribution of MR values of the tested irrigation water samples is shown in Figure 4(e).

Potential salinity

The PS values greater than 3 mEq/L indicate that the water is unsuitable for irrigation purposes, whereas a PS less than 3 mEq/L indicates that the water is suitable for irrigation purposes (Arslan 2017). The PS in the study area varied between 0.48 and 1.27 mEq/L. The highest and the lowest values of PS were observed at sampling points S15 and S1, respectively. The spatial distribution of PS values of the tested irrigation water samples is shown in Figure 4(f).

Permeability index

PI values of all water samples in the current study ranged from 45.31 to 78.17%, which were between 25 and 75% except at sampling points S1, S2, and S3 where the water was categorized as good for irrigation purposes, and the remaining three sampling points were categorized as excellent for irrigation purposes. Similarly, the spatial distribution map describes the PI value as 25–75% except at S1, S2, and S3 during the dry seasons, whereas when the PI of the water is between 25 and 75%, the water is good for irrigation purposes.

Seasonal pollution source apportionment in Lah River

During PCA and FA in this study, the KMO was found to be 0.693 and 0.613 for the wet season and dry season, respectively (p < 0.001) indicating the suitability of the data for PCA to interpret Lah River's water quality extracting three and two principal components and explaining the cumulative variance of 73.473 and 79.875% for the dry and wet seasons, respectively.

During the wet season, the first principal component (PC1) explained 70.92% of the total variance (Table 7) with strong positive loadings for , Cl , , and and moderate loadings for Na+ and EC having a component loading of 0.76, 0.89, 0.81, 0.78, 0.62, and 0.71, respectively, while pH (−0.887) has a strong negative loading for PC1. The strong positive loading could show that the source of pollution is potentially the non-point source of pollution generated by various sources during the rainy season. The study conducted by Hong et al. (2022) in the Tien Giang province showed that the waste released from the domestic area has shown an increasing trend for the parameters.

Table 7

Loadings of water quality variables on three significant PCs in the varimax rotated component matrix

ParametersWet season
Dry season
PC1PC2PC1PC2PC3
pH −0.887   −0.783  
TDS 0.495 0.770 0.492 0.740  
EC 0.715 0.608 0.464 0.763  
Ca2+ 0.449 0.708 0.808   
Mg2+ 0.467 0.781 0.878   
  0.693   0.864 
     0.779 
Cl 0.895  0.772   
 0.815  0.486 0.521  
K+  0.803 0.501 0.617  
 0.782 0.529   −0.584 
NO3 0.759 0.514 0.800   
BOD5  0.882  0.797  
Na+ 0.618 0.637 0.634 0.609  
Eigenvalue 9.219 1.164 7.070 1.849 1.368 
Variability (%) 70.919 8.957 50.497 13.208 9.769 
Cumulative % 70.919 79.857 50.497 63.704 73.473 
ParametersWet season
Dry season
PC1PC2PC1PC2PC3
pH −0.887   −0.783  
TDS 0.495 0.770 0.492 0.740  
EC 0.715 0.608 0.464 0.763  
Ca2+ 0.449 0.708 0.808   
Mg2+ 0.467 0.781 0.878   
  0.693   0.864 
     0.779 
Cl 0.895  0.772   
 0.815  0.486 0.521  
K+  0.803 0.501 0.617  
 0.782 0.529   −0.584 
NO3 0.759 0.514 0.800   
BOD5  0.882  0.797  
Na+ 0.618 0.637 0.634 0.609  
Eigenvalue 9.219 1.164 7.070 1.849 1.368 
Variability (%) 70.919 8.957 50.497 13.208 9.769 
Cumulative % 70.919 79.857 50.497 63.704 73.473 

The second principal component (PC2) explaining 8.96% of the variation was strongly contributed by BOD5, K+, Mg2+, and TDS and moderate loading for EC, Ca2+, Na+, and having a component loading of 0.88, 0.80, 0.78, 0.77, 0.61, 0.71, 0.64, and 0.69, respectively. The strong loading of potassium, BOD5, TDS, and magnesium in PC2 could indicate that the possible source of pollution is associated with anthropogenic factors such as the release of urban wastewater, which was also reported in the work of Dutta et al. (2018). The study done by Yilma et al. (2018) on Little Akaki River indicated that the presence of organic pollutant constituents from food waste, detergent, livestock operations, and solid waste dumping along the river has increased the concentration of BOD5, K+, and TDS. In addition, the significant contribution of Ca2+, , and Mg2+ may indicate that the river is affected by natural sources of pollution such as atmospheric deposition and weathering of soils, rocks, and minerals; this reason is also mentioned by the study Jaiswal et al. (2019).

In the dry season, Factor 1 explained 50.50% of the total water quality parameters and was associated with relatively high concentrations of Ca2+, Mg2+, Cl, and and moderate loading on Na+ with a component loading of 0.81, 0.88, 0.77, 0.80, and 0.63, respectively. According to Zhang & Wang (2019), the level of nitrate in surface water is mainly derived from domestic sewage in the dry season due to a lack of rainfall runoff from an agricultural area that carries nitrogen-containing chemical fertilizers.

During the dry season, the second principal component (PC2), explaining 13.21% total variance, had strong loading on EC (0.76) and BOD5 (0.80) and moderate loading on TDS, K+, and Na+ with a component loading of 0.74, 0.62, and 0.61, respectively. This component also has a negative loading on pH (−0.78). According to Dutta et al. (2018), the negative contribution of pH is interpreted as the organic pollution in the river, originating from the regular discharge of domestic wastewater into the river.

PC3 had strong positive loadings on bicarbonate (0.864) and carbonate (0.779) explaining 9.769% of the total variance. The higher concentrations of and in surface water may be the result of enhanced water–rock interactions and accelerated rock dissolution. Thus, PC3 was consistent with water–rock interactions (non-point source of pollution).

Pollution source contribution in Lah River

The PMF model analysis in the study area revealed that all the parameters except phosphate used in the PMF model were classified as strong species as the S/N ratio was greater than 2 during the wet season. However, during the dry season, all the parameters were classified as strong species and carbonate was deleted from the datasets because of having zero uncertainty.

The variable with S/N ratio of less than 2 is considered as weak whereas S/N ratio greater than 2 is considered as strong (Celen et al. 2022). The scale residual values between measured and predicted parameters were between −3 and 3 in the dry and wet seasons. The species with scaled residuals are between −3 and +3, indicating that the data were suitable for the PMF (Xiao et al. 2020). High R2 values (0.66–0.9 and 0.5–0.84) were found for all parameters except pH and Ca2+ during the wet season and except pH, K+ , , and during the dry season. This indicates that the water quality parameters estimated by the source factors selected in the PMF model are well explained (Zhang et al. 2012). On the other hand, after 20 and 17 iterations, two and three factors were determined by achieving the minimum Q value and its convergence during the wet and dry seasons, respectively. Q(robust) was equal to Q(true) during both seasons, which indicated that outliers did not influence the model disproportionately. Several things could cause the non-convergence, including uncertainties that are too low or specified incorrectly, or inappropriate input parameters. Table 8 shows the contribution and composition of each source to each parameter in the Lah River during the wet and dry seasons estimated by the PMF model.

Table 8

Source composition and contribution (% in parentheses) of Lah River constituents for the dry and wet seasons

ParametersSeasonFactor 1 (F1)Factor 2 (F2)Factor 3 (F3)
pH Wet 3.77 (54.0) 3.215 (46.1) 
Dry 0.56 (7.0) 2.66 (33.0) 4.84 (60.1) 
TDS Wet 27.425 (46.4) 31.63 (53.6) 
Dry 16.7 (12.7) 55.8 (42.4) 59.2 (45.0) 
EC Wet 53.57 (44.5) 66.74 (55.5) 
Dry 21.7 (9.1) 98.4 (41.3) 118.3 (49.6) 
Ca2+ Wet 9.62 (67.2) 4.7(32.8) 
Dry 3.57 (39.0) 3.57 (38.9) 2.03 (22.1) 
Mg2+ Wet 4.76 (45.2) 5.767 (54.8) 
Dry 2.98 (66.1) 1.3 (28.8) 0.23 (5.1) 
 Wet 23.2 (53.6) 20.08 (46.4) 
Dry 1.46 (9.0) 5.96 (36.8) 8.79 (54.2) 
Cl Wet 4.82 (36.0) 8.58 (64.0) 
Dry 4.08 (15.6) 13.17 (50.3) 8.96 (34.2) 
 Wet 8.57 (36.9) 14.7 (63.1) 
Dry 0.51 (5.5) 5.15 (54.9) 3.72 (39.6) 
K+ Wet 1.65 (48.8) 1.73 (51.2) 
Dry 1.83 (72.1) 0.003 (0.1) 0.7 (27.8) 
 Wet 0.185 (44.5) 0.23 (55.5) 
Dry 0.02 (12.5) 0.04 (34.3) 0.06 (53.3) 
 Wet 6.03 (32.9) 12.2 (67.1) 
Dry 0.56 (16.7) 2.35 (69.5) 0.47 (13.8) 
BOD5 Wet 12.34 (55.9) 9.72 (44.1) 
Dry 1.09 (12.6) 5.89 (68.4) 1.64 (19.0) 
Na+ Wet 4.13 (39.8) 6.24 (60.2) 
Dry 6.47 (39.6) 5.59 (34.2) 4.28 (26.2) 
ParametersSeasonFactor 1 (F1)Factor 2 (F2)Factor 3 (F3)
pH Wet 3.77 (54.0) 3.215 (46.1) 
Dry 0.56 (7.0) 2.66 (33.0) 4.84 (60.1) 
TDS Wet 27.425 (46.4) 31.63 (53.6) 
Dry 16.7 (12.7) 55.8 (42.4) 59.2 (45.0) 
EC Wet 53.57 (44.5) 66.74 (55.5) 
Dry 21.7 (9.1) 98.4 (41.3) 118.3 (49.6) 
Ca2+ Wet 9.62 (67.2) 4.7(32.8) 
Dry 3.57 (39.0) 3.57 (38.9) 2.03 (22.1) 
Mg2+ Wet 4.76 (45.2) 5.767 (54.8) 
Dry 2.98 (66.1) 1.3 (28.8) 0.23 (5.1) 
 Wet 23.2 (53.6) 20.08 (46.4) 
Dry 1.46 (9.0) 5.96 (36.8) 8.79 (54.2) 
Cl Wet 4.82 (36.0) 8.58 (64.0) 
Dry 4.08 (15.6) 13.17 (50.3) 8.96 (34.2) 
 Wet 8.57 (36.9) 14.7 (63.1) 
Dry 0.51 (5.5) 5.15 (54.9) 3.72 (39.6) 
K+ Wet 1.65 (48.8) 1.73 (51.2) 
Dry 1.83 (72.1) 0.003 (0.1) 0.7 (27.8) 
 Wet 0.185 (44.5) 0.23 (55.5) 
Dry 0.02 (12.5) 0.04 (34.3) 0.06 (53.3) 
 Wet 6.03 (32.9) 12.2 (67.1) 
Dry 0.56 (16.7) 2.35 (69.5) 0.47 (13.8) 
BOD5 Wet 12.34 (55.9) 9.72 (44.1) 
Dry 1.09 (12.6) 5.89 (68.4) 1.64 (19.0) 
Na+ Wet 4.13 (39.8) 6.24 (60.2) 
Dry 6.47 (39.6) 5.59 (34.2) 4.28 (26.2) 

During the wet season, the first source of pollution (F1) was identified as domestic and natural sources of pollution, due to the significant contribution of constituents such as pH, BOD5, Ca2+, and (Table 8). The study conducted by Xiao et al. (2020) at the Beichuan River in China showed that the highest loading of BOD5 indicates the contamination of the river by organic pollutants that come from domestic sewage and animal wastes. On the other hand, the highest loading of Ca2+ and on the source may indicate the dominance of natural sources of pollution. The second source of pollution (F2) entering into the Lah River has a significant loading on parameters such as , , , Cl , Na+, K+, TDS, EC, and Mg2+. The highest loading on and probably shows the contamination of the river by agricultural waste that comes from farmlands and garden areas, which are also informed in the study of Fathi et al. (2018). In addition, the highest loading of K+ , EC, TDS, Mg2+, Cl , Na+, and on this component may indicate the dominance of domestic sources of pollution. Based on the above reason, Factor 2 was preliminarily identified as the domestic and agricultural source of pollution.

In the dry season, Factor 1 was strongly associated with K+ (72.1%) and Mg2+ (66.1). Accordingly, Factor 1 can be explained by natural sources of pollution. The second factor (F2) was mainly characterized by , BOD5, Cl , and . High concentrations of in the surface water may be from domestic sewage due to much lower rainfall in the dry season, which is also explained in the study of Xiao et al. (2020). In addition, the strong contribution of BOD5, Cl, and in the river water probably indicated that the water is affected by organic pollutants from domestic sewage and different non-point sources of pollution, similar findings were also reported in the works of Khan & Wen (2021) and Celen et al. (2022). Based on this analysis, domestic source pollution was determined as the second factor. The third factor was weighted heavily on (53.3%), (54.2%), and pH (60.1%). The significant contribution of is related to the rock weathering and dissolution of carbonate in the water (Hui et al. 2020). In view of the above, the third factor was interpreted as the natural and domestic source of pollution.

The study evaluated the physicochemical water quality for drinking and irrigation water uses based on conjunctive application of WQI, multivariate statistical analysis, and MRMs. This integrated approach not only gives clear information for policymakers but also pinpoints the areas of focus specifically in data-scarce areas. Accordingly, the water quality analysis results examined from different sites of the river showed that most of the parameters were higher during the rainy season than the dry season and at the middle sample point than the upper and downstream river sections. According to the assessment of river water at different sampling points in both dry and wet seasons, most of the physicochemical water quality constituents were within the guideline standard, whereas some of the parameters exceeded the limit. According to the WAWQI, the quality of the Lah River's water was categorized as an unsuitable class (WAWQI > 100) at all sampling points during the rainy season. In addition, during the dry season, the water quality status ranged from poor to unsuitable categories (51.9–173.5) for drinking purposes. It was found that relatively the quality of the river water was highly polluted during the rainy season than the dry season. This was due to the entry of agricultural waste and other non-point sources of pollution into the river with runoff during the wet season. Irrigational suitability of the river water in terms of calculated values of SAR, Na%, RSC, PI, PS, KI, and MR was assessed at 16 sampling points, which restricted the use of the river water for agricultural purposes in the dry season. The values of all IWQIs were found to be within the limits of suitability for irrigating the crops during the dry season except PI. However, the lower PI values make the river water categorized under good class for agricultural purposes at all sampling points except at S1, S2, and S3. On the other hand, FA in the study area identified two significant sources of pollution for the Lah River during the wet season, namely, domestic and agricultural sources of pollution (PC1), and domestic and natural sources of pollution (PC2), which explained the 79.87% total variance. Similarly, during the dry season, three components were extracted, such as domestic and natural sources of pollution as PC1, a domestic source of pollution as PC2, and a natural source of pollution as PC3, which explains the 73.47% total variance. The PMF model also quantifies three sources of pollution with S/N > 2 and the residual error between +3 and −3 during both dry and wet seasons. Therefore, domestic waste, agricultural waste, and natural sources of pollution are the main sources of pollution in the Lah River. In conclusion, the study pinpointed the potential pollution zones and hotspot areas of the most socioeconomically important river. More focus is better given to the middle segment of the river by applying river ecological restoration and mitigation options such as buffer zoning. In addition, more targeted works on the reduction of anthropogenic influences need to be devised. Furthermore, future studies focusing more on the quantification of non-point sources of pollutants and their impact on the receiving river based on continuous and long-term river water quality monitoring could improve the water resources management of the study river.

The authors are very grateful to the Water Supply and Environmental Engineering faculty of Arba Minch University for allowing the Water Quality laboratory free of charge.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.

Ali
M. M.
,
Ali
M. L.
,
Islam
M. S.
&
Rahman
M. Z.
2016
Preliminary assessment of heavy metals in water and sediment of Karnaphuli River, Bangladesh
.
Environ. Nanotechnol.
5
,
27
35
.
APHA
.
2012
Standard Methods for the Examination of Water and Wastewater
,
Vol. 10
.
American Public Health Association
,
Washington, DC
.
Aydin
H.
,
Ustaoğlu
F.
,
Tepe
Y.
&
Soylu
E. N.
2020
Assessment of water quality of streams in northeast Turkey by water quality index and multiple statistical methods
.
Environ. Forensics
22,
1
18
.
Barakat
A.
,
El Baghdadi
M.
,
Rais
J.
,
Aghezzaf
B.
&
Slassi
M.
2016
Assessment of spatial and seasonal water quality variation of Oum Er Rbia River (Morocco) using multivariate statistical techniques
.
Int. J. Soil Water Conserv.
4
(
4
),
284
292
.
Bhutiani
R.
,
Khanna
D. R.
,
Kulkarni
D. B.
&
Ruhela
M.
2016
Assessment of Ganga river ecosystem at Haridwar, Uttarakhand, India with reference to water quality indices
.
Appl. Water Sci.
1947
,
107
113
.
Celen
M.
,
Oruc
H. N.
,
Adiller
A.
,
Töre
G. Y.
&
Engin
G. O.
2022
Contribution for pollution sources and their assessment in urban and industrial sites of Ergene River Basin, Turkey
.
Int. J. Environ. Sci. Technol.
19
(
12
),
11789
11808
.
Chinyama
A.
,
Ncube
R.
&
Ela
W.
2016
Critical pollution levels in Umguza River, Zimbabwe
.
Phys. Chem. Earth
93
,
76
83
.
Chukwuma
C. E.
,
Chukwuma
C. G.
,
Uba
I. J.
,
Orakwe
C. L.
&
Ogbu
K. N.
2016
Irrigation water quality index assessment of Ele River in parts of Anambra State of Nigeria
.
Arch. Curr. Res. Int
. 4, 1–6.
Dubey
M.
&
Ujjania
N. C.
2016
Seasonal variation in water quality of weir cum-causeway, Tapi River (India)
.
Pollut. Res.
35
(
2
),
429
433
.
Dutta
S.
,
Dwivedi
A.
&
Kumar
M. S.
2018
Use of water quality index and multivariate statistical techniques for the assessment of spatial variations in water quality of a small river
.
Environ. Eng. Res
. 190, 718.
Fathi
E.
,
Zamani
R.
,
Rafat
A.
&
Bidaki
Z.
2018
Water quality evaluation using water quality index and multivariate
.
Appl. Water Sci.
8
(
7
),
1
6
.
Goher
M. E.
,
Hassan
A. M.
,
Abdel-Moniem
I. A.
,
Fahmy
A. H.
&
El-Sayed
S. M.
2014
Evaluation of surface water quality and heavy metal indices of Ismailia Canal, Nile River, Egypt
.
Egypt. J. Aquat. Res.
40
(
3
),
225
233
.
Hamid
A.
,
Bhat
S. U.
&
Jehangir
A.
2020
Local determinants influencing stream water quality
.
Appl. Water Sci.
10
(
1
),
1
16
.
Hossain
M. A.
,
Ali
N. M.
,
Islam
M. S.
&
Hossain
H. M. Z.
2015
Spatial distribution and source apportionment of heavy metals in soils of Gebeng industrial city, Malaysia
.
Environ. Earth Sci.
73
(
1
),
115
126
.
Hsieh
P. Y.
,
Lin
H. C.
,
Wang
G. S.
,
Hsu
Y. J.
,
Chen
Y. J.
,
Wang
T. H.
,
Wang
R. D.
,
Kuo
C. Y.
,
Wang
D. W.
,
Liao
H. T.
&
Wu
C. F.
2022
Hazard ranking of wastewater sources in a highly polluted river in northern Taiwan by using positive matrix factorization with metal elements
.
Sustain. Environ. Res.
32
,
33
.
Huang, F., Wang, X., Lou, L., Zhou, Z. & Wu, J. 2010 Spatial variation and source apportionment of water pollution in Qiantang River (China) using statistical techniques. Water Res. 44 (5), 1562–1572.
Jaiswal
M.
,
Hussain
J.
,
Gupta
S. K.
,
Nasr
M.
&
Nema
A. K.
2019
Comprehensive evaluation of water quality status for entire stretch of Yamuna River, India
.
Environ. Monit. Assess.
191
,
208
.
Kilonzo, F., Masese, F. O., Van Griensven, A., Bauwens, W., Obando, J. & Lens, P. N. 2014 Spatial-temporal variability in water quality and macro-invertebrate assemblages in the Upper Mara River basin, Kenya. Phys. Chem. Earth. 67, 93–104.
Li
H.
,
Hopke
P. K.
,
Liu
X.
,
Du
X.
&
Li
F.
2015
Application of positive matrix factorization to source apportionment of surface water quality of the Daliao River basin, northeast China
.
Environ. Monit. Assess.
187
(
3
),
1
12
.
Özdemir
Ö
.
2016
Application of multivariate statistical methods for water quality assessment of Karasu-Sarmisakli creeks and Kizilirmak river in Kayseri, Turkey
.
Pol. J. Environ.
25
(
3
),
1149
1160
.
Wondim
Y. K.
,
Mosa
H. M.
&
Alehegn
M. A.
2016
Physico-chemical water quality assessment of Gilgel Abay River in the Lake Tana Basin, Ethiopia
.
Civ. Environ. Res.
4,
56
64
.
Xu
Z.
,
Xu
J.
,
Yin
H.
,
Jin
W.
,
Li
H.
&
He
Z.
2019
Urban river pollution control in developing countries
.
Nat. Sustain.
2
(
3
),
158
160
.
Yilma
M.
,
Kiflie
Z.
,
Windsperger
A.
&
Gessese
N.
2018
Assessment and interpretation of river water quality in Little Akaki River using multivariate statistical techniques
.
Int. J. Environ. Sci. Technol.
16
,
3707
3720
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).