Hydrochemical characterisation of a landslide prone area, South Western Ghats, India

ABSTRACT

This study focuses on the Manimala River Basin (MMRB) situated in the Western Ghats region of Kerala, which has been significantly affected by landslides. The water quality in the MMRB is likely to be impacted by heavy rainfall and the resulting debris flows. A comprehensive analysis was conducted for 157 water samples collected from the MMRB, with an objective to evaluate the hydrogeological conditions and estimation of 23 physico-chemical parameters to determine the water quality. Pearson’s correlation matrix for geochemical data was performed to investigate the relationships among physical and chemical parameters, and various ions in the MMRB. The present study also aimed to identify the dominant geochemical facies present in the water samples and determine the geochemical processes responsible for modifying the hydrochemistry. The results revealed that elevated concentrations of Ca²⁺ and SO₄^2- ions, which exert influence on the hydrochemistry of the region. Moreover, we utilized a fuzzy logic tool to categorize the water samples based on a water quality index, which indicated that approximately 92% of the MMRB samples are suitable for human consumption. This study offers valuable insights into the interplay of geochemical factors and water quality assessment, contributing to effective strategies for water resource management and conservation.

Graphical Abstract

KEYWORDS:

• Chemical weathering
• solute transport
• hydrochemistry
• water quality index
• fuzzy logic

Introduction

Access to safe drinking water is a universal right, yet UN statistics for 2020 reveal staggering disparities: 3.6 billion people lack proper sanitation, and 2 billion lack clean drinking water. By 2050, nearly half of the world’s urban population might face severe water scarcity (He et al., 2021). Factors like pollution, climate change, and inadequate infrastructure strain global water quantity and quality. Crucially, of 286 transboundary river and lake basins across 153 countries, only 58% have functional water cooperation arrangements (unece.org., UN World Water Development Report, 2023).

The Ministry of Jalsakthi in India conducted a comprehensive water body census in 2018–19 to evaluate, repair, and enhance water resources nationwide. The report identified 24,24,540 water bodies, with 97.1% (23,55,055) in rural areas and only 2.91% (69,845) in urban areas. Ponds make up the largest portion (59.5%), followed by tanks (15.7%), reservoirs (12.1%), lakes (0.9%), percolation tanks (9.3%), and check dams (2.5%). Around 83% are utilized, while 16.3% remain unused due to issues like salinity, siltation, damage, or drying. Private entities own 55.2%, while 44.8% are publicly owned (Water bodies, first census report, Vol. 1, 2023).

Kerala, situated in southwest India across 38,863 Km² (geographic coordinates: 8° 18'–12° 48' N & 74° 52'–77° 22' E), encounters both water abundance and scarcity in its diverse landscape. Its population of 33,406,061, with a density of 859/Km², is evenly split between rural (48%) and urban (52%) areas (ecostat.kerala.gov.in, 2014–15 report). The predominance of crystalline rocks limits groundwater potential in the region (Shaji, Nayagam, Kunhambu, & Thampi, 2009). Kerala harbors 55,734 water bodies, primarily in rural areas (89.21%), predominantly featuring ponds (91.51%), check dams (6%), tanks (1.5%), reservoirs (0.11%), and lakes (0.007%). Around 83.52% of these water bodies are utilized, primarily for irrigation (72.99%), with 70.67% privately owned and 29.32% publicly owned (State wise report of the first census of water bodies, Vol. 2, 2023). These statistics underscore the critical need for continuous surface and groundwater flow, necessitating effective water resource development for sustainable progress.

The crystalline rocks of Kerala are overlain by Tertiary sedimentary layers and recent alluvium, hosting various groundwater conditions (Shaji, Nayagam, Kunhambu, & Thampi, 2009). Factors like lithology, topography, and overburden thickness influence water table depth, varying from near-surface to 25 meters below ground level (mbgl), with highlands at around 10 mbgl and coastal areas at about 6 mbgl (CGWB, 2021).

Kerala’s groundwater potential is limited compared to other states, estimated at 5590 Mm³. Extraction mainly occurs through dug wells and borewells. In crystalline aquifers, static water levels range from 3 to 16 mbgl, while borewells go as deep as 240 mbgl, with productivity between 60 and 170 mbgl. Laterite and coastal aquifers have depths from 1 to 25 mbgl and 0.5 to 6 mbgl, respectively (CGWB, 2021).

Neglecting groundwater resources and harmful practices like chemical misuse can have severe consequences, necessitating long-term and costly recovery efforts (Adimalla, Li, & Venkatayogi, 2018). India faces groundwater contamination from various sources, including agrochemicals, industrial waste, and leaking storage systems (Jaiswal, Mukherjee, Krishnamurthy, & Saxena, 2003; Li, Karunanidhi, Subramani, & Srinivasamoorthy, 2021; Mir, Ahad, Inayatullah, Ali, & Ahmed, 2023). Approximately, 279 rivers in India were reported as polluted based on the Biochemical Oxygen Demand (BOD) parameter by the Central Pollution Control Board (pib.gov.in). Addressing these sources is crucial for preventing further contamination and protecting communities’ well-being.

Several studies emphasize rising groundwater’s substantial impact on slope stability. Elevated water table heights raise pore pressures in litho units, diminishing cohesion and leading to slope failures (Bogaard et al., 2007; Lindenmaier, Zehe, Dittfurth, & Ihringer, 2005). Seasonal groundwater variations, coupled with clay mineral changes from element leaching, crucially influence mineral dissolution and precipitation along fractures, worsening slope instability (Anson & Hawkins, 2002).

Considering all these important points into an account, the objective is frame to assess the hydrochemical quality and hydrological characteristics of the Manimala River Basin (MMRB) in an area prone to landslides situated in the central part of Kerala. The data obtained on the hydrogeology and hydrochemistry of MMRB can serve as a valuable baseline for future research focused on landslides in the region.

Study area-regional geology and extent

The study area, within Southern Granulite Terrain, Peninsular India, features an array of Archaean rocks: schistose (3200 Ma), layered basic-ultrabasic (3100–3000 Ma), Peninsular granitic gneiss (3000 Ma), khondalite (3070–2830 Ma), and charnockite (2600 Ma). These are overlaid by Proterozoic migmatites (2500–2200 Ma), basic intrusive rocks (2100–1600 Ma), ultra-basic rocks (700–600 Ma), younger charnockites (550 Ma), and granites (550–390 Ma). Mesozoic intrusions appear as NNW-SSE trending dyke swarms of basic-ultrabasic composition (Ambili & Narayana, 2022; Geology and mineral resources of the states of India, 2005; Yellappa & Rao, 2018). Additionally, Tertiary sedimentary rocks, with fossiliferous limestone, calcareous marl, sandstone, clay, and lignite seams, rest atop the Proterozoic formations. Overlying these are Quaternary formations consisting of recent to sub-recent sediments. Notably, the Vaikom and Warkalli formations are potential tertiary aquifers.

The state’s physiographic divisions comprise mountain peaks (>1800 m, 0.64%), highlands (600–1800 m, 20.35%), midlands (300-600 m, 8.44%), lowlands (300-10 m, 54.17%), and coastal plains/lagoons (0–10 m, 16.40%) (Soman, 2002). The Western Ghats (WG), a significant escarpment rising over 2500 m and stretching 1600 km along the west coast of the Indian peninsula, exhibits steep slopes, thick soil cover, and weathered rocks, heightening its susceptibility to landslides (Martha, Roy, Khanna, Mrinalni, & Vinod Kumar, 2019). Among the 41 rivers originating from the WG, some flow westward, merging with the Arabian Sea, while the rest flow eastward.

In recent years, Kerala has witnessed a shift in its atmospheric patterns, leading to heavy rainfall during monsoon seasons. Notably, from August 15–17, 2018, the region experienced intense rainfall, recorded 588 mm in the drainage basin (CWC, 2018). The state received 3505 mm of rainfall from April 2021 to March 2022, marking a 12% increase from the previous year. Monthly data for 2021–22 highlights significant deviations from normal rainfall: 157% above normal rain (abnr) in May 2021 due to the “TAUKTAE” cyclonic storm, 93% abnr in October, and 163% abnr in November 2021 (CGWB, 2021). When analyzing temporal annual rainfall data from 1901 to 2018, Kerala has encountered both heavy rainfall events (HRE) and droughts. Notably, heavy rainfall in 1924 and 1961 was more than that in 2018 (Mishra & Shah, 2018).

MMRB (drainage basin area: 1075.95 Km²) encompasses a 7^th order river. It is the 10^th longest drainage system in Kerala, originating from the southern WG, located between the geographic coordinates 9°21' and 9°40 ' N and 76° 33 ' and 76° 58 ' E (Figure 1). The mean annual rainfall of the area is more than 3000 mm, and the mean monthly temperature varies from 30°C to 35°C (Achu, Thomas, & Reghunath, 2020). Major rock types in the basin include massive charnockite, charnockite gneiss, quartzo-feldsphathic gneiss, cordierite gneiss, gabbro, quartzite, granite, laterite, and dolerite intrusive. Low-land regions of the basin are characterized by the dominance of Tertiary and Quaternary sediments, which unconformably overlay Precambrian crystalline rocks (Soman, 2002).

Figure 1. Location map of the study area.

Groundwater mainly occurs in phreatic aquifers of laterite, weathered crystalline rocks, and alluvium in MMRB. People in this region mainly depend on dugwells and borewells for groundwater extraction. Water from river and springs are also utilized to meet their domestic water demand. Crystalline rocks, charnockite and feldspar rich gneiss with secondary porosity and laterite are the main aquifers of MMRB. High land and parts of the midland physiographic division of MMRB are characterized by rubber, coffee, pepper and cardamom cultivation.

MMRB is prone to landslides (debris flow) under heavy rainfall. Many disastrous landslide events have been reported from the basin in September 2020 and October 2021. Kokkayar, Kuttickal, and Kavali areas of MMRB have been severely affected by the landslide events occurred under the influence of heavy rain (Ajin et al., 2022). The heavy rainfall events, subsequent debris flow, and denudation might have affected the hydrogeochemistry of the region.

Methodology

During June 2022, field investigations were conducted in the sub-basins of the Manimala River, to study the hydrogeological characteristics and collected 157 water samples for assessing the water quality. The sampling locations were carefully selected, and data on static water level, well depth, aquifer type, and geographic coordinates were recorded. A comprehensive well inventory was compiled (see Appendix 1) to facilitate future analysis. The collected water samples consisted of 134 samples from dug wells, 9 samples from bore wells, 12 samples from the river, and 2 samples from springs. These samples were collected to conduct hydrochemical analysis and to evaluate the quality of drinking water as well as its suitability for irrigation purposes. New polyethylene bottles were utilized for sample collection, and standard procedures were followed to prevent any potential contamination and to ensure the accuracy. Parameters such as temperature, color, and odor of the samples were recorded in the field, pH and electrical conductivity (EC) measurements were performed using a portable water quality meter, while the total dissolved solids (TDS) were measured using a digital TDS meter.

The hydrochemical analysis of water samples were carried out as per the protocols of the American Public Health Association (APHA, 2005) by estimating major cations – (Ca²⁺, Mg²⁺, Na⁺ and K⁺), and anions – (Cl⁻, CO₃^2-, HCO₃⁻. K⁺ and Mg²⁺). Fe concentration in the samples was determined using the Atomic Absorption Spectrum (AAS). The flame photometric method was applied for the analysis of Na⁺. Total Hardness (TH) and concentration of Ca²⁺, Cl⁻, HCO₃⁻ were determined using volumetric analysis.

The physico-chemical parameters of the collected water samples were evaluated to assess their suitability for drinking and irrigation purposes. Hydrogeochemical facies and potable water quality of MMRB were interpreted from the Hill-Piper (1953) trilinear diagram by plotting the concentration of anions and cations in milli-equivalent percentages. Chadha diagram was used to interpret the geochemical classification of water samples (Chadha, 1999; Thakur, Rishi, Naik, & Sharma, 2016). Parameters such as electrical conductivity (EC), sodium absorption ratio (SAR), residual sodium carbonate (RSC), and permeability index (PI) were evaluated for irrigation suitability. Additionally, the United States Salinity Laboratory Staff diagram (USSL Salinity Lab, 1954) is used to evaluate the water quality for irrigation.

The mechanism controlling the hydrochemistry is interpreted from the Gibbs diagram. The ionic ratio and ion exchange index estimated from the Chloro- Alkaline Indices (CAI) were used for interpreting hydrochemical sources and mechanisms.

Basic statistical analysis, including Pearson’s correlation matrix and standard deviation, provided insights into the relationship between variables in the samples. The positive and negative linear relationships between physical and chemical parameters, and various ions are interpreted from the Pearson’s correlation analysis. Standard deviation (SD) of the data were determined to measure the dispersion of the analytical data set from the mean value.

To establish a benchmark for comparison, the hydrochemical data of the MMRB was compared against the quality standards recommended by the World Health Organization (WHO) and the Bureau of Indian Standards (BIS). This comparison provided insights into the compliance of the water samples with the established quality standards, further aiding in the determination of their suitability for both drinking and irrigation purposes.

The Water Quality Index (WQI) of MMRB was evaluated with the aid of fuzzy logic software, offering a comprehensive assessment of water quality, considering the complex patterns and relationships in the hydro-physico-chemical variables, that are difficult to express by the traditional statistical models.

Results and discussion

The estimation of physical parameters and the concentration of ions in the water is vital for determining its suitability for domestic and irrigation purposes (Tatawat & Chandel, 2008). The mean, standard deviation, sum, minimum and maximum values of the physico-chemical parameters of 157 water samples from the MMRB and the water quality standards of WHO (2004) and BIS (2012) are summarized in Table 1. Charnockite and laterite are the major aquifers in MMRB. The static level ranges from 1.1 m to 8.2 m, with an average value of 3.72 m (Appendix 1).

Table 1. Descriptive statistics of physio-chemical parameters of MMRB samples.

Parameter	Min	Max	Sum	Mean	SD	Water quality standards
						WHO (2004)	BIS (2012)
pH	4.34	9.20	1208.59	7.70	0.95	6.5–8.5	6.5–8.5
TDS (ppm)	10	480.00	8560	54.52	62.57	1500	500
EC (µs/cm)	7.3	601.80	11052.22	70.40	76.09	1500	-
Temperature^0 C	25	31.00	4546.5	28.96	1.46	-	-
Na^+ (ppm)	0.76	44.53	873.73	5.57	6.85	200	200
K^+ (ppm)	0.01	2.74	51.01	0.32	0.42	12	12
Ca^2+ (ppm)	0.6	34.48	989.85	6.30	5.40	200	200
Mg^+ (ppm)	0	2.39	41.25	0.26	0.35	150	100
Cl^− (ppm)	7.09	42.55	3006.57	19.15	7.02	600	250
SO4^2- (ppm)	40	261.40	9064.9	57.74	27.87	250	400
HCO3^−(ppm)	0.282	317.20	2589.36	16.49	34.35	500	600
CO3 (ppm)	0	0.00	0	0	0.00	-	-
Fe (ppm)	0.0223	0.37	30.44	0.19	0.04	2.0	1.0
TA	14.62	27.89	2296	1	260	-	200
TH	19.75	15.40	3100	4	116	300	300

pH is defined as the negative logarithm of the hydrogen ion concentration, with the value ranging from 0 to 14, representing the neutral (pH = 7) or acidic (pH < 7) or alkaline nature (pH > 7) of the water (Covington, Bates, & Durst, 1985). The pH value of the water samples in the study area ranges from 4.34 to 9.2. About 23.56% of water samples are acidic, and the remaining 76.43% are alkaline in nature. 72.62% of the water samples fall within the permissible range of pH for drinking water (6.5–8.5) according to the water quality standards of both WHO and BIS. The remaining 27.38% of water samples, including two surface water samples (sample nos. 58 and 62) are unhealthy for human consumption as per the guidelines of WHO and BIS.

The TDS values of the water samples range from 10 ppm to 480 ppm, and all the samples fall within the permissible range of water quality standards suggested by WHO and BIS. The electrical conductivity (EC) values of the samples range from 7.3 µs/cm to 601 µs/cm, which are within the desirable limit of the EC value suggested by WHO and BIS.

The concentration of Na⁺ present in the water samples ranges from 0.76 ppm to 44.53 ppm, and the concentration of K⁺ in the water samples ranges from 0.01 ppm to 3.48 ppm. The estimated values of Na⁺ and K⁺ fall within the limits of WHO and BIS standards. The concentration of Ca²⁺ ranges from 0.12 ppm to 34.48 ppm, and that of Mg²⁺ ranges from 0.0004 ppm to 2.39 ppm. The concentration of Cl⁻ present in the water samples ranges from 7.09 ppm to 42.55 ppm. The concentration of bicarbonates ranges from 0.28 ppm to 317.2 ppm. The concentration of Ca²⁺, Mg²⁺, Cl⁻ and HCO₃⁻ falls within the desirable limits of WHO (2004) and BIS (2012) standards.

The concentration of sulfates in the water samples ranges from 40 ppm to 261.4 ppm. The permissible limit of sulfate is 250 ppm according to WHO (2004) and 400 ppm according to BIS (2012) standards. 10.79% of samples collected from MMRB are not suitable for drinking according to the WHO (2004) standards for sulfate. The cation and anion concentration of water samples are represented in the box chart (Figure 2). The anionic concentrations of water samples of MMRB is in the order SO₄^2->HCO₃⁻> Cl⁻> CO₃⁻ and cations in the order Ca²⁺>Na⁺>K⁺>Mg²⁺

Figure 2. Box diagram depicting the concentration of various ions in the water samples of MMRB.

The variation in the physical and chemical parameters of water samples collected from 134 dug wells, 9 bore wells, 12 samples from the river and 2 samples from springs are represented in Table 2. Figure 3 depicts the variation in ionic concentrations in water samples collected from dug wells, bore wells, river and springs. It is observed that HCO₃-, SO₄^2-, Na⁺ and Ca²⁺ are relatively high in samples collected from bore wells. However, the concentrations of Mg²⁺, K⁺ and Cl⁻ are slightly high in samples collected from dug wells. The ionic concentrations are low in samples collected from springs and river.

Figure 3. Variation in ionic concentrations (mean value) in water samples.

Table 2. Physico-chemical characteristics of water samples.

Spring sample	River sample	Open Well	Bore well	Well type
8.4	7.01	4.34	5.9	pH(min)
8.4	8.7	9.2	8.5	pH(max)
8.40	8.22	7.62	7.73	pH (mean)
10	10	20	50	TDS (min)
10	80	180	480	TDS (max)
10.00	34.29	45.81	240.91	TDS(mean)
19.54	20.17	20.27	7.3	EC (min)
21.12	49.15	255.5	601.8	EC (max)
20.33	33.61	62.73	269.40	EC (mean)
1.53	1.37	0.76	2.05	Na^+ (min)
1.72	4.98	20.25	44.53	Na^+ (max)
1.63	2.78	4.96	20.98	Na^+ (mean)
0.13	0.05	0.01	0.06	K^+ (min)
0.23	0.22	2.74	0.50	K^+ (max)
0.18	0.12	0.36	0.28	K ^+(mean)
3.4	2.5	0.6	1.7	Ca^2+ (min)
10.90	8.41	34.48	31.96	Ca^2+ (max)
7.15	5.76	6.16	12.46	Ca^2+(mean)
0.05	0.04	0.0004	0.00	Mg^2+ (min)
0.06	0.15	2.39	0.66	Mg^2+(max)
0.06	0.09	0.30	0.25	Mg^2+ (mean)
7.09	14.18	7.09	7.09	Cl^− (min)
14.18	28.37	42.55	35.45	Cl^− (max)
10.64	17.73	19.55	19.34	Cl^− (mean)
46.50	46.00	40.00	46.80	SO4 ^2- (min)
48.90	90.40	261.40	113.20	SO4 ^2- (max)
47.70	61.71	58.52	67.72	S04 ^2- (mean)
0.35	0.28	0.28	0.40	HCO3^− (min)
0.35	14.64	58.56	317.2	HCO3^− (max)
0.35	6.60	13.03	101.38	HCO3^− (mean)
0.17	0.13	0.02	0.16	Fe (min)
0.17	0.25	0.37	0.22	Fe (max)
0.17	0.21	0.19	0.18	Fe (mean)
8.00	6.00	6.00	4.00	TH (min)
10.00	18.00	88.00	116.00	TH (max)
9.00	12.00	20.06	38.55	TH (mean)
6.00	4.00	1.00	6.00	TA (min)
6.00	12.00	48.00	260.00	TA (max)
6.00	8.71	10.96	91.09	TA(mean)
−10.83	−4.19	−11.72	−8.54	RSC(min)
−3.34	3.12	4.73	2.32	RSC (max)
−7.09	−1.34	−0.71	−1.94	RSC(mean)
0.74	1.01	0.36	1.29	SAR (min)
1.17	2.54	14.68	20.29	SAR(max)
0.95	1.60	2.70	9.07	SAR (mean)
0.46	0.80	0.004	0	MR (min)
1.73	2.86	58.70	4.56	MR (max)
1.10	1.69	5.00	2.00	MR (mean)
13.58	19.70	8.55	13.78	SSP (min)
30.66	40.25	84.9	82.42	SSP(max)
22.12	31.12	39.25	52.98	SSP(mean)
14.45	20.75	12.17	14.50	%Na(min)
33.72	41.74	85.05	82.48	%Na(max)
24.08	32.08	40.77	53.82	%Na(mean)
0.16	0.25	0.09	0.16	KR(min)
0.44	0.67	5.62	4.69	KR(max)
0.30	0.47	0.80	2.06	KR(mean)
16.31	35.53	25.26	20.93	PI(min)
37.60	76.78	248.00	139.98	PI(max)
26.96	53.38	78.35	79.86	PI(mean)

Total alkalinity (TA) is the measurement of the concentration of alkaline substances like bicarbonate, carbonate, and hydroxide compounds of calcium, sodium, and potassium in water (Patil & Patil, 2010). The TA of all the water samples in the study area ranges between 1 and 260 ppm, and it falls within the permissible limit of TA for drinking water as per BIS standards. It is observed that TA is higher in bore well water samples than in dug well water samples. Low TA is reported in spring water samples.

Total hardness (TH) is the amount of dissolved calcium and magnesium in the water. The desirable limit of TH of potable water is 200 ppm, and the permissible limit is 600 ppm (BIS 10500, 2012). Since the TH of all the water samples ranges between 4 and 116 ppm, they are fit for human consumption. However, the TH of bore well samples are slightly high and a low value is reported from spring water samples.

Standard deviation (SD) is a measure of the variability of the data set. If SD is higher than the mean value, the data value spreads over a wide range, and vice versa. The SD values of pH, TDS, temperature, Cl⁻, SO₄^2- HCO₃⁻, TA, and TH are high, which indicates the variability of these parameters in the water samples of MMRB.

The relative variation in the physical and chemical parameters of the water samples collected from 137 dug wells, 9 bore wells, 12 samples from the river, and 2 samples from springs is summarized and represented in Table 2.

Hill-piper trilinear diagram

From the hydrochemical facies interpretation by plotting the cations and anions in the Hill-Piper trilinear diagram, it is observed that Ca²⁺ is the dominant cation in the water samples, and 66.24% of water samples fall within the calcium domain. It is followed by the alkali domain (28.02%), the mixed domain (5.09%), and the magnesium domain (0.6%) (Figure 4).

Figure 4. Hill - Piper trilinear diagram representing the quality of potable water.

The groundwater samples in the study area exhibit domination of SO₄^2-anion. Approximately 82.80% of the water samples are categorized within the sulfate domain, followed by 15.9% mixed domain and 1.2% bicarbonate domain. Analyzing the ionic concentrations of the water samples from the MMRB using the Hill-Piper trilinear diagram reveals that approximately 51.59% of the samples fall within the area 6, indicating a groundwater facies characterized by calcium-magnesium-chloride-sulfate composition. The area 7, accounting for 24.20% of the samples, represents sodium-potassium-chloride-sulfate facies, while area 9 (23.56%) displays a mixed geochemical facies where no single cation–anion combination exceeds 50%. Finally, area 8 represents sodium-potassium-bicarbonate facies, constituting only 0.6% of the samples.

Chadha diagram

Water samples were also geochemically classified with the aid of Chadha diagram (Chadha, 1999; Thakur, Rishi, Naik, & Sharma, 2016) by plotting (Ca²⁺+Mg²⁺) – (Na⁺+K⁺) on the X-axis and (CO₃^2-+HCO₃⁻) – (SO₄^2-+ Cl⁻) on the Y-axis in milli-equivalent percentages (Figure-5). Ionic concentration can be interpreted from this diagram, which makes it distinct from the Hill-piper trilinear diagram. The Chadha diagram is characterized by four major rectangular fields, with eight sub-fields. Water samples that fall in the eight sub-fields of the diagram are characterized by (1) alkaline earths exceeds alkali metals (2) alkali metals exceed alkaline earths (3) weak acidic anion exceed strong acid anions (4) strong acid anions exceed weak acid anions (5) alkaline earths and weak acidic anion exceed alkali metals and strong anions, indicating temporary hardness. (6) alkaline earth exceeds alkali metals and strong acidic anions exceed weak acidic anions. Such water has permanent hardness and does not deposit residual sodium carbonate. (7) alkali metals exceed alkaline earths and strong acidic anions exceed weak acidic anions, creating salinity issues (8) alkali metals exceed alkaline earths and weak acidic anions exceed strong acidic anions, causing the deposition of residual sodium that leads to foaming issues in the water.

From the Chadha diagram (Figure 5), it is evident that the majority of water samples from MMRB fall in sub-field 1, in which alkaline earths exceed alkali metals. The calcium-magnesium- bicarbonate hydrochemical facies underpins the role of recharge and weathering in determining hydrochemistry (Raza, Farooqi, Niazi, & Ahmad, 2016).

Figure 5. Chaddha diagram representing the quality of water in MMRB.

Assessment of the quality of water for irrigation

Inhabitants in the high and midlands of MMRB rely on agriculture for their livelihood, therefore it is essential to estimate the quality of water for irrigation. Classification of water samples suitable for irrigation based on RSC, SAR, MR, %Na, KR and PI (Doneen, 1964; Kelly, 1940; McGeorge, 1954; Todd, 2001; Wilcox, 1955) are summarized in Table 3.

Table 3. Classification of water samples for its suitability based on for irrigation quality standards.

Parameter	Range	Classification	No. of samples within the range
RSC	<1.25	Good	154
	1.25–2.5	Fair	1
	>2.5	Unsuitable	2
SAR	<10	Excellent	151
	10–18	Good	4
	18–26	Fair	2
	>26	Unsuitable
MR	<50	Suitable	155
	>50	Unsuitable	2
%Na	<20	Excellent	13
	20–40	Good	64
	40–60	Permissible	66
	60–80	Doubtful	10
	>80	Unsuitable	4
KR	<1	Suitable	126
	>1	Unsuitable	31
PI	>75	Excellent	60
	25–75	Good	95
	<25	Unsuitable	2

RSC value is an indicator of soil degradation due to Na⁺ retention in the soil, which leads to soil dispersion, reduced permeability which eventually leads to stunted plant growth (Sharma, Rishi, & Keesari, 2017). The measured RSC values of water samples range from −11.72 to 4.732 meq/l (Table 2). RSC water quality classification scheme (Table 3) suggests that 98.08% of samples meet excellent quality standards for irrigation, whereas the remaining percentage (sample no. 62 and 44) have poor quality for irrigation.

SAR value indicates the degree to which irrigation water tends to enter into cation-exchange reactions in soil. Sodium replacing adsorbed calcium and magnesium is a hazard as it causes damage to the soil structure and makes the soil compact and impervious (Raju, 2007). The measured SAR of water samples in the study area ranges from 0.355 ppm to 20.286 ppm (Table 3). SAR water quality criteria (Table 3) suggest that 96.17% of samples fall in the excellent category and two samples are unsatisfactory for irrigation purposes.

Magnesium is an essential element for the growth of plants, but an excess concentration of magnesium in irrigation water retards plant growth and becomes a hazard (Szabolcs & Darab, 1964). Magnesium retention (MR) value of water samples collected from the study area varies from 0% - 58%, and the MR water quality scheme (Table 3) suggests that 98.72% of the samples are suitable for irrigation.

The amount of sodium in water, denoted as percent sodium or sodium percentage (%Na) is a water quality parameter considered to evaluate its suitability for irrigation purposes (Wilcox, 1955). In all natural water, sodium combining with carbonate forms alkaline soils, while sodium combining with chloride forms saline soils (Wilcox, 1948). Either type of sodium-enriched soil will support little or no plant growth (Todd, 1980). The value of % Na in the study area varies between 12.17 and 85.05 ppm (Table 3). About 91.71% of samples fall within the permissible limit for %Na, and the remaining 8.29% are unsuitable for irrigation.

Potassium retention (KR) is the amount of sodium measured against calcium and magnesium, which indicates the excess level of sodium in water, describing the suitability of water for irrigation (Kelly, 1963). The value of KR in the study area varies between 0.093 and 5.62 (Table 3), and found about 80.25% of samples are good for irrigation.

Permeability index (PI) denotes the movement of water through the soil. Soil permeability is affected by long-term use of irrigation water with high salt content viz. sodium, calcium, magnesium, chloride, and bicarbonate (Janardhana Raju, Shukla, & Ram, 2011). PI of the study area ranges between 16.3 and 248 ppm (Table 1). About 98.72% samples meet the permissible limit, for water suitability for irrigation purposes whereas the remaining percentage (sample no 109 and 156) of samples are unsuitable for irrigation.

US salinity chart is prepared to interpret the suitability of water samples for irrigation purposes (Figure 6). The water samples have low to medium salinity hazard and sodium hazard and are recommended as good to excellent for irrigation.

Figure 6. US salinity chart representing the irrigation water quality of the water samples of MMRB.

Statistical analysis

Pearson’s correlation analysis enables the exploration of relationships between various chemical constituents and physical parameters of water, facilitating the identification of patterns and trend in the dataset (Kumari & Rai, 2020). The correlation matrix analysis (refer to Table 4), reveals a highly positive correlation between total dissolved solids (TDS) and electrical conductivity (EC) (correlation coefficient, r-0.899), as well as between bicarbonate (HCO₃⁻) (r-0.603), sodium (Na⁺) (r-0.816), and total alkalinity (TA) (r-0.636). Moreover, EC exhibits a strong relationship with Na⁺(r-0.871), HCO₃⁻(r-0.665) TA (r-0.667), and total hardness (TH) (r-0.527). Potassium (K⁺), chloride (Cl⁻), and salinity also exhibit similar high correlation trend with r values 0.571 and 0.996 respectively. Additionally, TH demonstrates a strong correlation with calcium (Ca²⁺) (r-0.834).

Table 4. Pearson’s correlation matrix of the physical and chemical parameters of MMRB.

pH

1

TDS

−0.024

1.000

pH

1.000

EC

−0.040

0.899**

1.000

Temp (^0C)

−0.530

0.113

0.123

1.000

Na^+

0.044

0.816**

0.871**

0.081

1

K^+

−0.022

0.360*

0.423*

0.119

0.286*

1

Ca^2+

0.062

0.416*

0.492*

−0.016

0.361

0.417

1

Mg^+

−0.076

0.140

0.137

0.240*

0.042

0.179

0.106

1

Cl^−

0.022

0.252*

0.335*

0.177*

0.269*

0.579*

0.288*

0.202*

1

SO4^2-

−0.164

0.110

0.075

0.088

0.048

−0.018

−1E–04

−0.011

−0.070

1

HCO3-

−0.005

0.603**

0.665**

0.166*

0.731**

0.060

0.249*

−0.002

0.078

0.021

1

CO3^2-

–

–

–

–

–

–

–

–

–

–

–

–

Fe

0.237*

−0.075

−0.087

−0.283

−0.060

−0.0124

−0.068

−0.126

0.0072

2.8E–04

−0.055

–

1

Salinity

0.025

0.249*

0.334*

0.174*

0.267*

0.571*

0.290*

0.202*

0.996**

−0.068

0.078

–

0.009

1

TA

0.018

0.636**

0.667**

0.110

0.721*

0.041

0.240*

−0.023

0.056

0.045

0.976**

–

−0.079

0.056

1

TH

−0.132

0.433*

0.527*

0.187*

0.351*

0.428*

0.834**

0.198*

0.338*

0.031

0.226*

–

−0.059

0.335*

0.194*

1

Hydrochemistry and mechanism: Interpretation from the Gibbs diagram

Gibbs diagram is the plot of TDS versus ratio of Cl⁻ and Cl⁻ + HCO₃⁻ in ppm, and TDS versus ratio of Na⁺+ K⁺ and Na⁺+K⁺+ Ca²⁺ (Gibbs, 1970). Gibbs proposed that the chemistry of groundwater and surface water is controlled mainly by three mechanisms viz. rock – water interaction (in which the geochemistry of rock controls the hydrochemistry), rate of evaporation, and atmospheric precipitation (Ribinu, Prakash, Khan, Bhaskar, & Arunkumar, 2023).

The water samples have an average TDS concentration of 54.52 ppm, which underscores the role of chemical weathering in controlling the hydrochemistry. The majority of the water samples from MMRB fall in the precipitation dominance and rock dominance fields of the Gibb’s diagram, almost equally, which indicates that the recent heavy rainfall events in MMRB and ionic exchange through rock–water interaction play significant roles in governing the hydrochemistry of the study area (Figure 7). Gibbs proposed that Na⁺ ions may be entered to the surface and groundwater through atmospheric precipitation, even in regions far away from the coast. MMRB is located approximately 100 km from the coast, and the precipitation imprints in the hydrochemistry of the region interpreted from the Gibbs diagram is valid.

Figure 7. A, b. Gibb’s diagram depicting the relation between TDS, cations and anions and interpretation of the mechanism for hydrochemistry of MMRB.

Hydrochemistry sources and mechanisms: Interpretations from ionic ratio and ion exchange index

Ionic exchange, oxidation, reduction, and dissolution of minerals in conjunction with the weathering of rocks contributing ions and alter the hydrochemistry (Ribinu, Prakash, Khan, Bhaskar, & Arunkumar, 2023). Cation exchange reactions can be inferred from the high concentration of Na⁺ in relation to Cl⁻ or vice versa. In the normal ion exchange reaction, Ca²⁺ ions are retained in the aquifer, whereas Na⁺ ions are liberated into the water. Cl⁻ or SO₄^2-ions, did not balance the excess Na⁺ ions generated. In the reverse ion exchange reactions, Na⁺ ions are retained in the aquifer and Cl⁻ ions are released to water (Senthilkumar & Elango, 2013).

The plots of Ca²⁺ + Mg²⁺ and Na⁺ + K⁺ versus total cations (TZ), Cl⁻ versus Na⁺, Ca²⁺ + Mg²⁺ versus HCO₃⁻ + SO₄^2- (meq/L) and HCO₃⁻ versus Ca²⁺ + Mg²⁺ (mmol/L) of samples were prepared and are depicted in Figure 8 a, b, c, d respectively.

Figure 8. (a) Distribution of Ca²⁺ + Mg²⁺ in relation to total cations (TZ) (meq/L), 8 b. Correlation between Na⁺, and K⁺ concentrations (meq/L), 8 c. Relationships between Na⁺ and Cl⁻ in water samples (meq/L) 8 d. SO₄^2- + HCO₃⁻ concentration (meq/L) in relation to Ca²⁺ + Mg²⁺.

The TZ plot (refer to Figure 8(a,b)) illustrates that the alkaline earths surpass the alkalis in the studied area. The combined concentration of Ca²⁺ and Mg²⁺ ions exceeds that of Na⁺ and K⁺ ions, primarily originating from the weathering of silicate minerals present in the charnockite and gneiss formations within the MMRB. The Na⁺ versus Cl⁻ plot (Figure 8c) indicates that the majority of the samples fall above the 1:1 line, suggesting the possibility of Na⁺ ions being replaced by Ca²⁺ ions. However, the Na⁺/Cl⁻ ratio remains below 1, indicating that these ions did not originate from silicate weathering processes (Meybeck, 1987). This finding supports the conclusion that a certain proportion of Na⁺ and Cl⁻ ions in the samples are derived from rainfall.

The graph depicting Ca²⁺ + Mg²⁺ versus HCO₃⁻ + SO₄^2- (refer to Figure 8d) shows that the samples are positioned on the right side of the plot, below the 1:1 line. This observation suggests the occurrence of ion exchange reactions, which contribute these ions to the water samples. The trend observed in the water samples of the MMRB consistently falls below the 1:1 line, indicating that the ions are likely derived from the weathering of silicate and carbonate rocks through ion exchange mechanisms.

Based on the graph illustrating Ca²⁺ + Mg²⁺ versus HCO₃⁻ (mmol/L) (refer to Figure 9), it can be interpreted that the majority of the samples are positioned above the 0.5 slope. This observation suggests that ion exchange plays a significant role in shaping the hydrochemistry of the MMRB. Furthermore, the majority of the samples fall above the equiline 1:1, indicating that silicate weathering is the predominant process influencing the hydrochemical composition.

Figure 9. Ca²⁺ + Mg²⁺ concentration (mmol/L) and their relation to HCO₃⁻.

In the bivariate plot of Mg²⁺/Na⁺ versus Ca²⁺/Na⁺ (refer to Figure 10a) representing the sodium- normalized alkaline earth metal concentrations, three zones are demarcated – carbonate dissolution, silicate weathering and evaporate dissolution zones. The MMRB sample points spread in the area between silicate weathering and evaporate dissolution, suggesting the mixing of ions from the silicate weathering process. In the Ca²⁺/Na⁺ versus HCO₃⁻/Na⁺, bivariate plot (refer to Figure 10b) the samples fall in the area between silicate weathering and carbonate dissolution.

Figure 10. (a) Bivariate analysis of Mg⁺/Na⁺ in relation to Ca²⁺/Na⁺, and the correlation between Ca²⁺/Na⁺ and HCO₃-/Na⁺ (Figure 10b) in the water samples, reveals the mixing of ions derived from silicate weathering.

A Ca²⁺/Mg²⁺ ratio greater than 1 suggests the dissolution of dolomite, while a ratio greater than 2 indicates silicate weathering (Vasu et al., 2017). The Ca²⁺/Mg²⁺ ratio in water samples from the MMRB is higher than 2, which means that silicate weathering is the main process that affects the ion composition. The plot of Ca²⁺ versus HCO₃⁻ (refer to Figure 11a) for MMRB water samples reveals that the majority of the samples fall above the equiline, indicating higher Ca²⁺ levels compared to HCO₃⁻. Furthermore, the plot of SO₄^2- versus Ca²⁺ (Figure 11b) demonstrates elevated concentrations of SO₄^2-.

Figure 11. (a) Plot illustrating the relationship between Ca²⁺ and HCO₃- suggests that a majority of the samples exhibit elevated concentration of Ca²⁺ compared to HCO₃-. Figure 11b, depicting the correlation between SO₄^2- and Ca²⁺ reveals a notable concentration of SO₄^2-in the water samples.

The weathering of pyroxene in charnockite contributes to the enrichment of Ca²⁺ and Mg²⁺ ions, as well as HCO₃⁻ in the water (Adithya et al., 2016; Senthilkumar & Elango, 2013)).

1 $\mathrm{Ca}\mathrm{Mg}\left(\mathrm{S}{\mathrm{i}}_2{\mathrm{O}}_6\right)+4\mathrm{C}{\mathrm{O}}_2+6{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}{\mathrm{a}}^{2+}+\mathrm{M}{\mathrm{g}}^{2+}+4\mathrm{HC}{\mathrm{O}}_3-+2\mathrm{Si}{\left(\mathrm{OH}\right)}_4.$1

Additionally, water often contains a significant amount of CO₂, which leads to an increase in bicarbonate levels (Egbueri, Mgbenu, & Chukwu, 2019). The relatively high concentration of SO₄^2- indicates that CO₂ dissolution is minimal, revealing the oxygen-rich environment of the study area.

The weathering of plagioclase feldspar (anorthite) in gneiss has also contributed to the Ca²⁺ content in the water (Vinnarasi et al., 2020).

2 $\mathrm{CaA}{\mathrm{l}}_2\mathrm{S}{\mathrm{i}}_2{\mathrm{O}}_8+2\mathrm{C}{\mathrm{O}}_2+3{\mathrm{H}}_2\mathrm{O}\to \mathrm{A}{\mathrm{l}}_2\mathrm{S}{\mathrm{i}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4+\mathrm{C}{\mathrm{a}}^{2+}+2\mathrm{HC}{\mathrm{O}}_3-$2

The ratio of Ca²⁺ + Mg²⁺/Tz is higher (0.58) compared to the ratio of Na⁺ +K⁺/Tz (0.42) in charnockite, suggesting that the weathering of pyroxenes in charnockite is more significant than in gneiss (Li, Wu, & Qian, 2013). The presence of Ca²⁺ ions may replace Na⁺ ions, resulting in relatively high levels of SO₄^2- in the water samples. The oxidation of pyrite in charnockite contributes to the presence of Fe and SO₄^2- (Raj et al., 2023).

3 $2\mathrm{Fe}{\mathrm{S}}_2+2{\mathrm{H}}_2\mathrm{O}+7{\mathrm{O}}_2\to 2\mathrm{F}{\mathrm{e}}_2{++4\mathrm{S}\mathrm{O}}_4^{2-}+4{\mathrm{H}}^{+}$3

Additionally, the K⁺ ions derived from weathering form compounds with Cl⁻, which is supported by the correlation coefficient analysis.

Chloro- alkaline indices (CAI) serve as indicator of ion exchange (Li, Wu, & Qian, 2013; Schoeller, 1967). The indices were derived from the following equations

4 $\mathrm{CAI}-1=\mathit{C}{\mathrm{l}}^{-}-\left(\mathrm{N}{\mathrm{a}}^{+}+{\mathrm{K}}^{+}\right)/\mathit{C}{\mathrm{l}}^{-}$4

5 $\mathrm{CAI}-2=\mathit{C}{\mathrm{l}}^{-}-\left(\mathrm{N}{\mathrm{a}}^{+}+{\mathrm{K}}^{+}\right)/\mathit{HC}{\mathrm{O}}_3-+\mathrm{S}{{\mathrm{O}}_4}^{2-}\text{break}+\mathit{C}{{\mathrm{O}}_3}^{2-}+\mathrm{N}{{\mathrm{O}}_3}^{2-}$5

Positive CAI indices indicate the ion exchange of Na⁺ and K⁺ ions from groundwater to Ca²⁺ and Mg²⁺ ions in the aquifer, known as direct base exchange. Conversely, negative CAI values indicate the reverse trend, termed indirect base exchange, which implies a chloro-alkaline imbalance. The negative Schoeller index values presented in Table 5 for the water samples of MMRB suggest the likelihood of ionic exchange between Ca²⁺ and Mg²⁺ ions in water with Na⁺ and K⁺ ions in the aquifer, leading to a chloro-alkaline imbalance. The Na⁺/Na⁺ + Cl⁻ ratio was calculated, and the resulting values (0.2) are less than 0.5, indicating the replacement of Na⁺ ions by Ca²⁺.

Table 5. Schoeller index of water samples of MMRB.

Location number	CAI – 1	CAI – 2
1	−4.03	−0.06
2	−1.56	−0.02
3	−4.77	−0.07
4	−8.02	−0.12
5	−4.12	−0.07
6	−2.84	−0.05
7	−9.02	−0.15
8	−5.63	−0.1
9	−3.8	−0.06
10	−9.63	−0.14
11	−4.04	−0.05
12	−2.28	−0.03
13	−2.18	−0.04
14	−4.42	−0.04
15	−4.64	−0.07
16	−2.71	−0.04
17	−4.56	−0.04
18	−2.55	−0.03
19	−2.23	−0.04
20	−10.41	−0.13
21	−0.77	−0.01
22	−4.84	−0.09
23	−4.28	−0.07
24	−7.53	−0.09
25	−8.96	−0.11
26	−8.28	−0.08
27	−1.74	−0.03
28	−2.87	−0.06
29	−5.56	−0.07
30	−1.93	−0.03
31	−2.88	−0.04
32	−1.39	−0.03
33	−2.22	−0.03
34	−4.57	−0.08
35	−3.64	−0.06
36	−4.88	−0.08
37	−4.24	−0.05
38	−4.68	−0.07
39	−3.48	−0.06
40	−2.18	−0.03
41	−2.17	−0.04
42	−11.98	−0.23
43	−2.12	−0.04
44	−4.01	−0.06
45	−2.71	−0.05
46	−9.07	−0.16
47	−5.11	−0.09
48	−3.52	−0.06
49	−3.71	−0.07
50	−19.03	−0.23
51	−2.44	−0.04
52	−3.93	−0.08
53	−4.13	−0.07
54	−4.1	−0.06
55	−3.15	−0.06
56	−3.94	−0.08
57	−3.14	−0.05
58	−3.01	−0.05
59	−44.08	−0.12
60	−42.06	−0.14
61	−5.13	−0.09
62	−4.51	−0.07
63	−4.37	−0.07
64	−13.78	−0.1
65	−13.32	−0.21
66	−12.7	−0.15
67	−21.92	−0.2
68	−19.4	−0.43
69	−20.03	−0.33
70	−4.7	−0.05
71	−3.14	−0.06
72	−3.75	−0.05
73	−5.16	−0.1
74	−13.22	−0.24
75	−3.77	−0.07
76	−6.71	−0.13
77	−3.63	−0.06
78	−1.59	−0.02
79	−5.68	−0.1
80	−2.57	−0.04
81	−2.88	−0.05
82	−2.5	−0.03
83	−7.72	−0.14
84	−2	−0.03
85	−36.06	−0.5
86	−3.72	−0.07
87	−1.87	−0.03
88	−5.05	−0.06
89	−2.32	−0.02
90	−2.17	−0.02
91	−2.94	−0.03
92	−2.23	−0.03
93	−1.61	−0.03
94	−2.62	−0.06
95	−3.12	−0.06
96	−7.99	−0.19
97	−3.97	−0.08
98	−3.29	−0.07
99	−2.78	−0.05
100	−2.43	−0.05
101	−3.69	−0.04
102	−4.22	−0.09
103	−3.61	−0.06
104	−1.72	−0.02
105	−1.66	−0.02
106	−1.42	−0.02
107	−2.21	−0.05
108	−2.22	−0.05
109	−1.85	−0.04
110	−1.76	−0.04
111	−3.85	−0.05
112	−1.9	−0.04
113	−7.34	−0.15
114	−7.88	−0.14
115	−3.27	−0.07
116	−12.65	−0.24
117	−2.62	−0.05
118	−3.66	−0.05
119	−4.55	−0.06
120	−5.04	−0.09
121	−2.22	−0.04
122	−4.79	−0.09
123	−4.33	−0.07
124	−4.12	−0.07
125	−1.95	−0.03
126	−2.57	−0.04
127	−3	−0.06
128	−3.65	−0.06
129	−2.45	−0.04
130	−1.82	−0.03
131	−2.19	−0.04
132	−3.56	−0.06
133	−4.71	−0.08
134	−6.3	−0.1
135	−4.42	−0.07
136	−1.97	−0.03
137	−4.26	−0.08
138	−5.21	−0.1
139	−4.29	−0.04
140	−3.62	−0.03
141	−2.33	−0.01
142	−4.45	−0.02
143	−6.37	−0.07
144	−7.19	−0.11
145	−13.75	−0.25
146	−8.28	−0.14
147	−14.05	−0.21
148	−5.44	−0.1
149	−7.52	−0.1
150	−1.97	−0.02
151	−7.88	−0.14
152	−16.42	−0.21
153	−3.71	−0.03
154	−3.4	−0.04
155	−44.94	−0.16
156	−5.53	−0.08
157	−2.97	−0.03

The groundwater in MMRB undergoes a base exchange reaction where Ca²⁺ and Mg²⁺ ions replace Na⁺ ions, resulting in the formation of HCO₃⁻ and softened water (Handa, 1969; Thakur, Rishi, Naik, & Sharma, 2016). When the concentration of Ca²⁺ + Mg²⁺ is greater than HCO₃⁻, it is classified as base exchanged hardened water. The water samples from MMRB fall into the category of softened water. These ion exchange mechanisms play a significant role in shaping the hydrochemistry of the region.

Water quality index

Water quality index (WQI) is a collective numerical rating of water quality parameters that highlights the influences of various parameters on the overall quality of water for specific use (Alobaidy, Abid, & Maulood, 2010). WQI was calculated to assess the suitability of potable water based on WHO standards (WHO, 2011). WQI of the study area was evaluated with 13 parameters viz. pH, TDS, EC, Na⁺, K⁺, Ca²⁺, Mg²⁺, Cl⁻, SO₄^2-, HCO₃ ⁻, Fe⁺, TA and TH.

WQI is calculated by the following equations:

6 $\mathit{Wr}=\mathit{Wi}/\sum \mathit{Wi}$6

where W_r is the relative weightage for each parameter and W_i is the weight of each parameter. Water quality rating scale for each parameter is calculated by

7 $\mathit{Qi}=\left(\mathit{Ci}/\mathit{Si}\right)\ast 100$7

where Ci represents the concentration of parameters and Si represents the drinking water quality standards for each parameter according to WHO (2011). Finally, the sub-indices (SI_i) and aggregate WQI are evaluated using the following equations:

8 $\mathrm{S}{\mathrm{I}}_{\mathrm{i}}=\mathrm{Wr}\times {\mathrm{Q}}_{\mathrm{i}}$8

9 $\mathit{WQI}=\sum \mathit{SIi}$9

The water quality assessment based on standard values was conducted using fuzzy membership functions. To simplify the model, linear membership functions were adopted, as suggested by Zhang et al. (2012). The membership function is expressed as:

10 $\begin{array}{rl}& \mathrm{rij}=0,\mathrm{ifCi}<\mathrm{Sij}-1\mathrm{orCi}>\mathrm{Sij}+1\\ & \left(\mathrm{Ci}-\mathrm{Sij}\right)/\left(\mathrm{Sij}-\mathrm{Sij}-1\right),\mathrm{ifSij}-1\le \mathrm{Ci}\le \mathrm{Sij}\\ & \left(\mathrm{Sij}+1-\mathrm{Ci}\right)/\left(\mathrm{Sij}+1-\mathrm{Sij}\right),\mathrm{ifSij}\le \mathrm{Ci}\le \mathrm{Sij}+1\\ & 1,\mathrm{ifCi}=\mathrm{Sij}\end{array}$10

where rij represents the fuzzy membership of indicator i to class j, Cj is the analytical value of water quality indicator i, and Sij is the allowable water quality indicator.

The weight (Wi) of each water quality indicator is calculated using the equation:

11 $\mathrm{Wi}=\mathrm{Ci}/\mathrm{Si}$11

where Wi represents the weight of water quality index i, Ci is the analytical value of water quality indicator i, and Si is the arithmetic mean of the allowable value of each class.

The normalized weight of each indicator is determined by

12 $\mathit{ai}=\mathit{Wi}/\left(\sum \mathit{Wi}\right)$12

Where ai is the normalized weight of indicator i, Wi is the weight of the water quality indicator i, and ∑Wi is the sum of the weights of all water quality parameters (Table 6).

Table 6. Weight and relative weight of the water samples of MMRB.

Parameters	Weight (wi)	Relative weight Wi$=\frac{wi}{{\sum }_i^{n}1\mathit{wi}}$
pH	4	0.089
TDS(ppm)	5	0.1111
EC (µs/cm)	4	0.089
Na^+ (ppm)	2	0.044
K^+ (ppm)	2	0.044
Ca^2+ (ppm)	2	0.044
Mg^2+ (ppm)	1	0.022
Cl- (ppm)	3	0.067
SO4^2- (ppm)	4	0.089
HCO3- (ppm)	4	0.089
Fe (ppm)	5	0.11
Salinity (ppm)	4	0.089
TA (ppm)	2	0.044
TH (ppm)	3	0.067

The fuzzy A matrix represents the weights of each water quality indicator. The water quality assessment using fuzzy logic membership is performed based on the matrix

13 $\mathrm{B}=\mathrm{A}\ast \mathrm{R}$13

The fuzzy B matrix represents the membership of each water quality class. The water sample is classified into the class with the maximum membership.

Water quality index (WQI) classification for the samples of MMRB is calculated and summarized in Table 7.

Table 7. Water quality index (WQI) classification.

Sl. No	WQI values	Classification type
1	388.27	Very Good
2	399.62	Very Good
3	404.96	Good
4	453.11	Good
5	367.25	Very Good
6	466.08	Good
7	489.59	Good
8	352.88	Very Good
9	332.24	Excellent
10	427.67	Good
11	462.24	Good
12	377.80	Very Good
13	326.77	Excellent
14	554.26	Poor
15	397.88	Very Good
16	362.68	Very Good
17	590.71	Poor
18	463.23	Good
19	393.16	Very Good
20	388.79	Very Good
21	340.97	Excellent
22	329.39	Excellent
23	398.74	Very Good
24	435.25	Good
25	424.68	Good
26	518.50	Poor
27	350.78	Very Good
28	324.08	Excellent
29	414.50	Good
30	368.15	Very Good
31	352.37	Very Good
32	360.31	Very Good
33	367.03	Very Good
34	405.72	Good
35	381.44	Very Good
36	388.30	Very Good
37	453.70	Good
38	398.99	Very Good
39	341.62	Excellent
40	381.70	Very Good
41	359.94	Very Good
42	433.72	Good
43	372.95	Very Good
44	378.10	Very Good
45	354.81	Very Good
46	374.58	Very Good
47	391.60	Very Good
48	378.87	Very Good
49	381.06	Very Good
50	547.33	Poor
51	388.91	Very Good
52	377.73	Very Good
53	385.12	Very Good
54	404.14	Good
55	463.40	Good
56	399.16	Very Good
57	443.98	Good
58	390.04	Very Good
59	807.15	Very Poor
60	789.99	Very Poor
61	376.57	Very Good
62	415.86	Good
63	427.81	Good
64	577.00	Poor
65	444.58	Good
66	512.65	Poor
67	574.43	Poor
68	381.22	Very Good
69	478.01	Good
70	426.52	Good
71	388.38	Very Good
72	386.36	Very Good
73	228.43	Excellent
74	454.88	Good
75	448.02	Good
76	447.64	Good
77	441.44	Good
78	390.70	Very Good
79	469.18	Good
80	464.44	Good
81	321.26	Excellent
82	360.18	Very Good
83	456.71	Good
84	471.79	Good
85	569.01	Poor
86	397.64	Very Good
87	309.88	Excellent
88	404.59	Good
89	424.63	Good
90	433.21	Good
91	424.39	Good
92	421.43	Good
93	413.40	Good
94	394.07	Very Good
95	413.45	Good
96	423.73	Good
97	479.10	Good
98	415.48	Good
99	344.43	Excellent
100	355.02	Very Good
101	456.23	Good
102	345.67	Excellent
103	456.96	Good
104	399.93	Very Good
105	420.39	Good
106	411.59	Good
107	422.92	Good
108	426.18	Good
109	343.84	Excellent
110	335.51	Excellent
111	398.71	Very Good
112	342.21	Excellent
113	373.34	Very Good
114	426.31	Good
115	340.83	Excellent
116	424.53	Good
117	392.93	Very Good
118	357.22	Very Good
119	373.58	Very Good
120	359.05	Very Good
121	360.77	Very Good
122	373.24	Very Good
123	367.68	Very Good
124	366.91	Very Good
125	323.07	Excellent
126	357.58	Very Good
127	332.33	Excellent
128	329.17	Excellent
129	333.55	Excellent
130	335.61	Excellent
131	328.39	Excellent
132	317.10	Excellent
133	310.98	Excellent
134	327.68	Excellent
135	316.00	Excellent
136	304.90	Excellent
137	323.45	Excellent
138	318.48	Excellent
139	359.00	Very Good
140	344.26	Excellent
141	421.05	Good
142	429.59	Good
143	360.70	Very Good
144	365.87	Very Good
145	401.87	Good
146	360.84	Very Good
147	439.34	Good
148	365.95	Very Good
149	426.94	Good
150	396.60	Very Good
151	412.35	Good
152	483.87	Good
153	382.78	Very Good
154	379.40	Very Good
155	743.67	Very Poor
156	571.80	Poor
157	373.32	Very Good

The status of the drinking water quality was assessed and classified based on the aggregate WQI, and are evaluated as (Class 1) excellent (WQI < 350), (Class 2) very good (350.1–400), (Class 3) good (400.1–500), (Class 4) poor (500.1–600), (Class 5) very poor (>600.1) (Boateng, Opoku, Acquaah, & Akoto, 2016).

Fuzzy logic is a soft computing tool used in multiple domains for logical decision-making by manipulating multi-data sets (Kambalimath & Deka, 2020). This tool is widely used in hydrochemical studies to manage the uncertainties or vagueness of complex physico-chemical data of water samples to arrive at logical interpretations based on symbolic (linguistic) and numeric (mathematics). It considers the non-linear relationship between variables and can model it, even though the data is vague. Water quality class assessment of 157 samples of MMRB by fuzzy membership function is given in Table 8.

Table 8. Water quality class assessment by fuzzy membership function.

Water sample class	Number of samples within each class					Total
	Class 1	Class 2	Class 3	Class 4	Class 5
Total Number of sample within class	29	62	54	9	3	157

Conclusion

Hydrochemistry of MMRB reveals that major ions are present in the following order: SO₄^2- > Cl⁻ > HCO₃⁻ > Ca²⁺ > Na⁺ > K⁺ > Mg²⁺ > Fe > CO₃^2-, and the pH values indicate the alkaline nature of water. Comparison of WHO and BIS standards with the hydrochemical data suggests that all the physical parameters and ion concentrations fall within the desirable range for human consumption as well as irrigation. The Hill-Piper trilinear diagram categorizes 51.59% of the water samples as Ca²⁺-Mg²⁺-Cl⁻-SO₄^2- geochemical type, 24.2% as Na⁺-K⁺-Cl⁻-SO₄^2- type, and 23.56% as mixed type. Chadha diagram classifies the water samples into calcium-magnesium- bicarbonate hydrochemical facies, providing insights into the role of recharge and weathering in determining hydrochemistry. The water quality index, interpreted using fuzzy logic tools, indicates that 18%, 39% and 34% of the samples are of excellent quality, very good quality and good quality, respectively. The US salinity chart classifies 99% of the samples as excellent and good for irrigation purposes. The parameters such as RSC, SAR, MR, %Na, KR, and PI also substantiate the suitability of water for irrigation. The concentrations of HCO₃⁻, SO₄^2-, Na⁺, and Ca²⁺ were relatively higher in the water samples from bore wells in MMRB. Pearson’s correlation matrix shows strong positive correlation between TDS and EC, HCO₃⁻ Na⁺, and TA. EC exhibits a strong relationship with Na⁺, HCO₃⁻, TA, TH, K⁺, and Cl⁻. Strong correlations were also observed between TH and Ca²⁺

The modification of hydrogeochemistry was interpreted through mechanisms identified from the Gibbs diagram, which indicated that the ions originated from the interaction between infiltrating precipitation and the aquifer. Silicate weathering is the primary mechanism responsible for the hydrochemical modifications in the region. The weathering of pyroxenes and calcic plagioclase may have played a role in regulating the water chemistry. The relatively high concentration of SO₄^2- in the samples could be attributed to the oxidation of pyrite in charnockite and/or anthropogenic inputs, such as fertilizers. Furthermore, the negative value of the Schoeller index indicates ionic exchange between Ca²⁺ and Mg²⁺ ions in the water with Na⁺ and K⁺ ions in the aquifer.

The study reveals that the primary mechanism responsible for the hydrochemical transformations in the humid tropical region is chemical weathering, which triggers ionic exchange reactions. This mechanism may also have implications for the cohesiveness of litho units and soil profiles, potentially impacting the slope stability of the region. Future studies that focus on landslides can utilize the hydrochemical data from MMRB as a baseline for further investigations.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data are included as electronic supplementary material.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/23570008.2024.2331392.

Additional information

Funding

No fund is received from any agency to conduct this research.