Hydrogeochemical assessment of spring water resources around Melamchi, Central Nepal

Groundwater in the hills and mountains appears on the surface as springs, the natural discharge features of aquifers/groundwater systems. Springs are major lifelines for people in Nepal's mountainous regions, and are precious resources that play important roles in securing human requirements, maintaining balance in the ecosystem and providing riverine base flows.

According to Kresic & Stevanovic (2010), a spring is a location where groundwater discharges from the aquifer create visible flow on the land surface. The discharge is caused by the difference in elevation of the aquifer's hydraulic head and the land surface at that point. Geology is important in aquifer formation and, consequently, on groundwater accumulation and movement. It is also important in relation to local groundwater quality because of rock-water interactions. Reactions between groundwater and aquifer materials change water quality, which is also important in understanding groundwater genesis (Cederstrom 1946), because groundwater quality depends on the total amounts and kinds of minerals dissolved from the surrounding rocks. The concentration of dissolved ions depends on groundwater flow paths, and the interactions between rock and water (Hem 1985). The main cations in groundwater are derived mainly from the solution of minerals during chemical weathering, while the anions may be derived from non-lithologic sources.

Water quality has been an important issue in groundwater studies. The hydrogeochemical study of groundwater includes measurement of physicochemical parameters such as temperature, pH, electrical conductivity (EC), total dissolved solids (TDS), and dissolved oxygen (DO), and the concentration of major cations: calcium (Ca²⁺), magnesium (Mg²⁺), sodium (Na⁺) and potassium (K⁺), and the major anions: bicarbonate (HCO₃⁻), sulfate (SO₄²⁻) and chlorine (Cl⁻). These are important tools and decisive factors for water quality assessment (Ali & Ali 2014; Bharti et al. 2017).

Groundwater chemistry arises from geological and anthropogenic activities (Subramani et al. 2005). Different geochemical processes within the groundwater and aquifer materials affect the water's chemical quality. Study of the hydrogeological regime and groundwater chemistry shows the possible geological and anthropogenic influences on water quality (Kumar et al. 2006; Bhat & Jeelani 2016). It also aids water resource management by identifying the water's quality and its suitability for different purposes. For spring conservation, the term ‘springshed’ is introduced and the importance of springshed management in the Nepal Himalaya is highlighted (Rijal 2016). The aim of this study was to investigate seasonal variations in the in-situ physicochemical parameters of the groundwater, evaluate its hydrogeochemical characteristics with respect to water type, and identify the groundwater chemistry control mechanism. The water's suitability for drinking was also evaluated.

The study area is in the Sindhupalchok district, northeast of the Kathmandu Valley, in Central Nepal. It includes Jyamire and Dubachaur, two major villages within Melamchi municipality. Geographically, the study area extends from 26°49′30″ N to 27°53′00″ N latitude and 85°32′30″ E to 85°38′00″ E longitude. The study area and relevant spring locations are shown in Figure 1.

An inventory of springs was an important part of this work and began with a field survey of springs using GPS (Global Positioning System). The physicochemical parameters: pH, EC, TDS, DO and temperature, were measured in-situ along with notes of the spring type, nature of discharge, surrounding land use and type of deposit. In-situ parameters were measured with a Mettlor Toledo field kit and discharge was determined using a bucket and stopwatch. The in-situ parameters were measured in May (pre-monsoon) and October 2017 (post-monsoon). The samples for laboratory analysis were collected post-monsoon only. Samples were collected in standard 1,000 ml bottles. Geological parameters, including rock type, structure, inclinations of rock strata with horizontal plane, dip and strike were also noted at the spring location where rock was exposed.

A total of 18 springs were sampled post-monsoon (October 2017) for laboratory determination of the major cations and anions, and the data obtained were used for hydrogeochemical characterization using Piper diagrams and Gibbs plots. Piper (1944) uses a trilinear diagram for water classification with two triangles, one each for cations and anions. The diamond-shaped field between them is used to represent the water's composition with respect to both ionic groups. The Gibbs plot (1970) is used to determine the groundwater hydrogeochemistry control mechanisms and explain the relationship between groundwater chemistry and aquifer materials. The plot has three zones representing evaporation, rock, and precipitation dominance from top to bottom.

The study area lies in the Higher Himalaya region of Nepal Himalaya, where the major rock types observed were schist and gneiss. Dhital et al. (2002) note that the area is covered by the Talamarang Formation, with light to dark grey kyanite schist, garnetiferous schist, banded gneiss and quartzite (Figure 2). The sequence dips generally to the NW. The part of the study area east of Indrawati Nadi exhibits mostly medium to thick bedded, light to dark grey, slightly weathered schist. In some locations, garnetiferous schist and quartzite were also observed. Kyanite schist was observed mainly north of Siteni Khola. West of Indrawati Nadi, the geology mainly comprises highly weathered schist and medium to thick bedded gneiss exposed at a few locations.

The type and nature of spring observed is characteristic of the local geology and structures. The springs mostly emerged from unconsolidated sediments in topographic depressions, mostly as seepages or concentrated discharges. Some springs emerged, however, from the contact between sediments and underlying strata. A third type of spring emerges from planes of weakness – e.g., rock fractures. Using Bryan's classification system (Bryan 1919) of the geological occurrence of springs, the majority were found to be depression and fracture springs. The presence of thick unconsolidated sediments on the slopes make depression springs prevalent in the region, while fracture springs were also observed. Fractured gneiss and schist account for groundwater storage and transmission. Eighteen springs were selected for study and are described in Table 1.

The water's physicochemical parameters were analyzed statistically. The results are presented in box and whisker plots (Figure 3). The pH values fell within the range 5.5–7.0 pre-monsoon and 5.7–7.2 post-monsoon; that is, local spring waters are slightly acidic to marginally alkaline. The minimal alkalinity may be attributed to the presence of bicarbonate ions produced by the formation of carbonic acid. The spring water EC ranges from 29.3 to 234 μS/cm pre-monsoon and 20.9 to 189.8 μS/cm post-monsoon. Pre-monsoon values are higher than post-monsoon because there is less groundwater recharge before the monsoon and groundwater residence time is longer. After the monsoon, when there has been significant groundwater recharge and shorter residence time, less rock-water interaction has led to lower ion concentrations. Springs showed changes in EC levels between seasons. In the majority of springs, the change was less than 30 μS/cm but in a few (5) it was between 30 and 70 μS/cm. The TDS ranged from 14.6 to 116.9 mg/l pre-monsoon and 10.5 to 94.9 mg/l post-monsoon, the high pre-monsoon TDS values and their significant decrease post-monsoon indicate rocks as the source of dissolved solids in the water. Similarly, DO ranged from 4 to 7 mg/l pre-monsoon and 4.6 to 7.3 mg/l post-monsoon. The concentration of DO decreased during summer as temperatures rose, but biochemical processes and water-rock interactions might also have played role in the change. Slight differences between spring water temperatures were recorded in May and October 2017, with a range from 18.7 to 24.3 °C pre-monsoon and 19.2 to 23.7 °C post-monsoon.

The concentration ranges of major ions obtained from the laboratory analyses are presented in Table 2 – see also Figure 4, which shows the related box and whisker plots.

Groundwater is derived from precipitation such as rainfall, which contains very low concentrations of dissolved minerals. When it seeps into the ground, it begins by moving through the inter-connected pores in the soil, unconsolidated sediments, and rock fractures. Mineral dissolution in soil, sediments and rocks is slow and depends on their solubility. Because of this, the water from spring J20 (outlier in Figures 3(d) and 4), which is at a lower elevation (1,217 m) than the others, contains high concentrations of ions. The exception is spring J42 (1,187 m), which has lower concentrations than spring J40, possibly due to the different spring types – J40 is a depression spring and J42 a fracture spring. The fracture networks provide direct pathways for water flow towards discharge points; that is, springs, allowing less time for interactions than the paths to depression springs.

Comparing the spring waters’ physical and chemical parameters, all are within the limits in the National Drinking Water Quality Standards (NDWQS 2005) and the World Health Organization's recommendations (WHO 2011), except that the majority of spring waters report pH below 6.5.

The discharges for 13 springs at the two sampling times are shown in Figure 5. Spring discharges are higher post-monsoon than pre-monsoon, and ranged from 1.7 to 62.5 L/m pre-monsoon and 5.5 to 65.7 L/m post-monsoon. Post-monsoon spring discharges are very high in most springs, indicating enhancement by monsoonal precipitation. However, two fracture springs (J42 and J35) have higher pre-monsoon than post-monsoon discharge because the high discharge rates disturbed the original fracture network, creating new flow-paths during the summer monsoon, all of which could not be captured during discharge measurement. In some places, regolith removal by shallow, rain-induced landslides is also related to such spring discharge behavior patterns. The change in discharge rate in springs seem to be similar except spring J12, which shows much more significant increase after the monsoon than any other. J12 emerges in a topographic depression whose primary storage is the unconsolidated soil and sediments around the spring, which are recharged quickly, increasing the flow.

Identifying areas with different groundwater types helps in the sustainable management of groundwater resources. The hydrogeochemical regime of groundwater in the study area can be determined by plotting the spring waters’ analytical values on a Piper (1944) trilinear diagram. The classification was done using USGS GW Chart (Version 1.29) to plot the cationic and anionic concentrations. All seven ionic species were included in the data clustering. Of the spring waters studied, 16 are of Ca-HCO₃ type, one (D16) appears in field 4, indicating a mixed Ca-Mg-Cl type, and one (D9) in field 3, indicating mixed Ca-Na-HCO₃ type (Figure 6).

The study suggests that, in the waters of the study area, alkaline earth metals (Ca²⁺ and Mg²⁺) predominate over alkali metals (Na⁺ and K⁺) and weak acids (HCO₃⁻) predominate over strong ones (Cl⁻ and SO₄²⁻). Spring D9 reported the lowest calcium, sodium and bicarbonate concentrations, while spring D16 reported the highest chlorine concentration. The majority having Ca-HCO₃ water type indicates that most springs are fed by shallow aquifers.

Gibbs plots (Gibbs 1970) were used to study the relationship between water composition and the aquifers’ lithologic characteristics. As shown in Figure 7, lithology and precipitation are the most significant contributors to spring water chemistry.

The concentration of major cations and anions present in the groundwater depends on the hydrogeochemical processes occurring in the aquifer system (Glover et al. 2012). As the groundwater travels from recharge to discharge, a variety of hydrogeological processes alter its chemical content so that, at the discharge zone, it has higher mineral content that in the recharge area due to its residence time and contact with the aquifer matrix (Freeze & Cherry 1979; Sridhar et al. 2013). The low ionic concentrations in spring D9, which is at 1,700 m elevation, are related to the limited contact time and interactions with the aquifer matrix there. The analytical results for most samples fall into the rock-water interaction zone of the Gibbs plot, indicating the predominance of interaction between the rocks and the percolating water.

Variation in the in-situ physicochemical parameters (pH, EC, TDS, DO, temperature) of spring waters in different seasons were studied. The type of water was characterized with respect to content of major ionic species (Ca²⁺, Mg²⁺, Na⁺, K⁺, HCO₃⁻, SO₄²⁻ and Cl⁻) and the hydrogeochemical processes responsible for the ionic concentrations identified. Geologically, the study area comprises light to dark grey kyanite schist, garnetiferous schist, banded gneiss and quartzite. The local geological conditions indicate a majority of depression and fracture springs. Seasonal variations in the in-situ physicochemical parameters and spring discharge were observed mainly due to the influence of monsoon precipitation.

The spring water is slightly acidic to marginally alkaline. The general order of dominance of major cations is Ca²⁺ > Na⁺ > Mg²⁺ > K⁺ and major anions HCO₃⁻ > Cl⁻ > SO₄²⁻. Piper and Gibbs diagrams were used to study the waters’ hydrogeochemical composition. The major ion concentrations, when plotted in the Piper diagram, indicate calcium and bicarbonate dominating the aquifer, suggesting that springs are fed by shallow water-bearing zones. The Gibbs plot shows that groundwater quality in the area is influenced mainly by rock-water interactions.

This study was funded by World Wildlife Fund (WWF), Nepal. The authors would like to thank Hariyo Ban Program WWF, Nepal for research funding support under a student thesis research grant awarded to the first author. The support of Maria Maharjan, Sima Humagain, Gunanidhi Pokhrel and Rupak Gyawali during field work is also acknowledged.

The views and interpretations in this publication are those of the authors and they are not necessarily attributable to their organizations.

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