Hydro-Geochemical Conditions under Projected Climate Change Scenarios of Marshyangdi River, Nepal

Assessment of hydro-geochemical processes in a Himalayan River fed by snow and glaciers in the context of global climate change is crucial to understanding the changes in water quality due to natural and anthropogenic in�uences. Thus, the hydro-geochemical status of water quality was analyzed in a snow-fed Himalayan Watershed, Marshyangdi located in western Nepal for current and future scenarios under the medium (RCP 4.5) and pessimistic (RCP 8.5) representative concentration pathways (RCPs) for two seasons (pre-and post-monsoon, 2019) based on multiple regional climate models. Flow at each sampling site of a total of twenty-one sites was estimated from a soil and water assessment tool (SWAT) hydrological model and then the concentration of water quality for the future was determined. A descriptive analysis of water quality was carried and a Piper plot diagram for evaluating the spatiotemporal variation as well as the hydro-geochemical status of water for the current and future scenarios. The results reveal alkaline water in the watershed based on pH values that follow the pattern of average ionic dominance Ca 2+ > Mg 2+ > Na + > K + for cations and HCO 3 − > Cl − > NO 3 − for anions indicating the carbonate-dominated lithology in the Marshyangdi Watershed for the current scenarios. However, for future scenarios dominance of cations is different for the respective seasons Ca 2+ > Na + +K + >Mg 2+ and Na + +K + > Mg 2+ > Ca 2 with similar anion composition with respect to current scenarios under both RCPs.


Introduction
Many water resource challenges are rooted in hydrogeochemistry, a eld that combines hydrology and mineral/rock chemistry (Zhu and Schwartz 2011).Therefore, future changes to hydrogeochemical processes that affect water quality must be considered in efforts to solve issues related to global warming.Furthermore, the demand for water is increasing rapidly due to population increase, industrialization, urbanization, and mechanization, which causes water supplies to be depleted thereby causing pollution (Falkenmark, 1994).Though water can be recycled and treated for both industrial and domestic usage, the expense of managing and treating it may be substantial (Eyankware et al., 2014).
Hence, in order to manage water resources sustainably meeting human requirements, hydrogeochemical data on water resources are crucial.The world's surface water chemistry is mainly controlled by three major mechanisms; atmospheric precipitation, rock dominance, and the evaporation-crystallization process (Gibbs et al., 1970).The hydrogeochemical processes reveal the zones and the suitability of the water quality for different uses, including drinking, farming, and industry.It also helps to understand the changes in water quality due to rock-water interaction as well as anthropogenic in uences (Kumar et al., 2006).The hydrochemistry of river water is impacted by a number of anthropogenic activities, such as mining, quarrying, and dumping, as well as natural processes like weathering, precipitation, and ion exchange.As a result, water pollution has emerged as one of the world's most signi cant environmental issues today ( Singh et al., 2008).Hence, to manage water resources sustainably and meet human requirements, hydro-geochemical data on water resources and uxes are crucial (Zhu et al., 2011).Hydrogeochemical studies have been carried out around the globe by several researchers.Kumar et al. (2008) studied the chemical characteristics of the surface, groundwater, and mine water at the upper catchment of the Damodar River Basin, India to evaluate the major ion chemistry and the geochemical processes controlling water composition and suitability of water for domestic, industrial and irrigation uses.The research showed that the water was acceptable for irrigation as well as domestic usage.Similarly, Gupta et al. (2016) focused on the hydrogeochemical investigation of water samples from the Rangit River, Sikkim, India for human consumption and agricultural purposes.The study visualized that the major hydrogeochemical facies in upstream river water as (K + -Ca 2+ -Mg 2+ -HCO 3 − ), whereas (Ca 2+ -K + -HCO 3 -) predominated in downstream sections of the river.The high equivalent ratios of (Ca 2+ -Mg 2 +)/ (Na + -K + ) and the low ratio of (Na + -K + ) in the study evidenced that the chemical such watersheds, this study aims to assess the spatio-temporal composition of major ion chemistry along with the assessment of the water quality based on physicochemical characteristics with respect to cations and anions in current and future contexts in a snow-fed Himalayan River Marshyangdi.In addition to this, other stressors such as the alteration of the ow regime, particularly as a result of the development of hydropower may also impact the hydrogeochemical process in a Marshyangdi Watershed.There are 47 hydropower plants in the various phases of operation in the Marshyangdi Watershed (DOED, 2020), where three major hydropower plants are already in operation.So, there might be a possibility of changing the hydrology of rivers impacting water quality.

Study Area
Marshyangdi Watershed is a snow-fed perennial river that originated from the Tethyan Himalayan Sequence THS located in West-Central Nepal.It covers four districts namely Manang, Gorkha, Tanahu and Lamjung.The river is around 150 km long and it extends from 27°50 '42" to 28°54' 11" N latitudes and from 83°47'24" to 84°48'0 " E longitudes covering an area of 4,748 km 2 (Fig. 1).The elevation of this basin varies from 274 to 8,042 m.This river rises on the northern slopes of the Annapurna Himalaya, ows east through the arid valley around Manang, and then swings south to join the Trisuli River at Mugling.The major source of this river is the glaciers of Annapurna Himalaya range, Manaslu Himalaya range, and Larkya Himalayan sub range.The general climate varies from sub-tropical in the lower belt to arctic frost in the higher altitudes (Karki et al., 2016)

Sampling Sites and Analytic Methods
A total of 21 sampling sites (M1-M21) were identi ed covering the mainstream and its tributaries (Fig. 1).Detail characteristics of the sampling sites are provided in Table 1.
The surface water samples with three replicates were collected, composited, and stored in a clean 500 mL polyethylene bottle.Then, the sample bottles were sealed, labeled properly, and transported to the laboratory in an ice box maintaining a temperature of 4 o C. Before the collection of the water sample, each sample bottle was rinsed twice with the same river water and then washed with nitric acid for cations.For the analysis of chemical oxygen demand (COD), sample was preserved with a few drops of conc sulphuric acid in a separate sample bottle of 200 mL.Water quality parameters like pH, temperature (T), Total dissolved solids (TDS), and electrical conductivity (EC) were measured using the multipurpose meter (HANNA; HI98129).Dissolved oxygen (DO) was directly measured in the eld using the portable digital meter (EcosenseDO200A).These probes were immersed in composite water in a beaker and the parameter was noted after the stabilization of the instrument.The major cations such as Ca 2+ and Mg 2+ , Na +, and K + were analyzed using Flamephotometer (JENWAYPFP7) and anions like bicarbonate (HCO

Climate change impacts on hydrogeochemical conditions
For the assessment of climate change's impact on hydrogeochemical status, the climate was projected at the watershed with the help of dynamic downscaling using three RCMS as described in Singh et al.
Further, the Soil and Water Assessment (SWAT) model was set up then calibrated and validated with acceptable statistics at three hydrological stations to study the impact of climate change on hydrogeochemical conditions within the watershed (Singh et al., 2022).After the development of the model, projected time series data of ow (Q) at each sampling site was calculated by using the drainagearea ratio method and then using Eq. 1.The concentration of water quality was calculated for two future periods and under two RCPs (4.5 and 8.5) scenarios.

Prediction of water quality condition
The water quality component was predicted by using Eq. 1, assuming the types and quantities of pollutants from human settlements remain constant (Zhao et al., 2019).Pollution and runoff concentrations is stated as: where, F is the measured runoff, C is the measured concentration of pollutants, F′ is the predicted future runoff, and C′ is the concentration of pollutants.

Results And Discussion
3.1 Characterization of physico-chemical processes

Current scenarios
The status of water quality based on the statistical summary of physico-chemical parameters among the 21 studied sites within the Marshyangdi Watershed has been explained brie y and presented in (Table 2) by categorizing them into physical, chemical, and nutrient parameters.
River temperature is a key physical parameter that in uences river ecology (Medupi 2016;Webb & Walsh 2004) indirectly in uencing the mobilization as well as the toxicity of pollutants.The temperature in the basin ranges from 4.9 to 30.3 ± 6.5 (Table 2) in pre-monsoon whereas in post-monsoon 3.3 to 25.1 ± 5.6 which might be due to the altitudinal variations within the watershed (Table 1).However, a number of factors, including diurnal daylight, weather, terrain, slope, aspects, different sampling times, altitudinal and seasonal change, ow of water, numerous biotic and abiotic components, and surface radiation (as well) may have an impact on the temperature of the surface water (Trivedi & Goel, 1986; Singh et al., 2017).The overall temperature is crucial for regulating freshwater's physical, chemical, and biological processes.
All the water quality parameters representing physical status (pH, EC, TDS) are within the range of permissible limits for the sustenance of the aquatic life within the watershed (Table A1).In current scenarios, the mean pH values for both seasons show that the water is naturally alkaline.This pH parameter signi cantly impacts the composition of aquatic macroinvertebrates as well as their metabolic and physicochemical characteristics (Tadesse et al., 2018).Similarly, mean values of all the parameters represent water chemical status (DO, Cl − , COD, BOD, NH 3 -N & SO 4 2− ).For the case of TH, water is very hard (> 180 mg/L) in pre-monsoon and hard in post-monsoon.In addition, for alkalinity, it exceeded its permissible limit for the sustenance of aquatic life (> 200mg/L) in the post-monsoon season (Table 2).
Finally, nutrient parameters (NO 3 − & PO 4 3− ) were also within the permissible level for the survival of aquatic organisms (Table A1) except total phosphate whose both maximum and mean values exceeded their limits for both studied seasons (> 0.1mg/L) (Table 2).Such exceedance in total phosphate might be due to non-point sources in the watershed, which might cause the problem of eutrophication in the fuMean values of cations like (Ca 2+ , Mg 2+ , K + and Na + ) are also within the acceptable limits for the survival of aquatic life in current scenarios.Future scenarios (Near future, RCPs 4.5 and 8.5) The descriptive statistics have been calculated and presented for both seasons under both RCPs scenarios for the near future (Table 3a and Table 3b) and for the mid-future (Table A2a and Table A2b).Mean pH values in both seasons indicate a continuation of alkaline water in the near future under both RCPs.All the physical parameters (pH, EC, and TDS) are within acceptable limits for the sustenance of aquatic organisms (Table A1).Likewise, chemical parameters (DO, Cl − , COD, BOD, TH, TA, NH 3 -N & SO 4 2− ) have been also predicted and within the standard limits for the sustenance of aquatic life in both seasons under both RCPs.Furthermore, nutrient parameters have been also predicted and are within the permissible limits (Table A1) for both seasons under both RCPs.Similarly, predicted mean values of cations (Ca 2+, Mg 2+ , K + , & Na + ) have also fallen within acceptable limits for the survival of aquatic life.Mid-Future (RCP 4.5 and 8.5) The mid-future descriptive statistics under both RCPs are presented in Table A2a and Table A2b.For the mid-future also mean values of all the physical parameters, were within the acceptable limits for both seasons under both RCPs.For the case of chemical parameters, their mean values have been predicted to be their acceptable limits (Table A2a) for both RCPs.Similarly, nutrient parameters of water quality have been predicted as within the acceptable limits except for total phosphate whose both maximum as well as mean values exceeded their permissible level (> 0.05mg/L) (Table A2a) for both seasons under both RCPs.For the case of cations, all of them have been predicted to be within the acceptable levels for the survival of aquatic life for both seasons under both RCPs (Table A1)

Current scenarios
The hydro-chemical characteristics of the Marshyangdi Watershed were analyzed for two seasons (preand post-monsoon) from headwater to the mouth along the 21 sampling sites based on a piper trilinear diagram (Piper, 1944).
In pre-monsoon, based on mean values of cations, Ca-HCO 3 mg/L was dominant (Fig. 2a) revealing the dominance of calcium in the river water followed by magnesium (Ca 2+ >Mg 2+ >Na + +K + ).The dominance of calcium and magnesium ions exhibited 44 and 58 percent, respectively more than the global mean (Gaillardet et al., 1999 The overall characteristics of the water chemistry reveal the dominance of the alkaline earth metals (Ca 2+ and Mg 2+ ) over the alkalis (Na + + K + ) and the weak acids (HCO 3 − ) over the strong acids (Cl − and SO 4 2− ) in pre-monsoon (Fig. 2a).In general, the sampling points in the piper diagram can be classi ed into six elds.The types are 1.Ca-HCO 3 , 2. Na-Cl, 3. Mixed: Ca-Mg-Cl, 4. Mixed: Ca-Na-HCO 3 , 5. Ca-Cl and 6.Na- However, in post-monsoon (Fig. 2b), sodium and potassium ions predominate the water chemistry followed by calcium and magnesium (Na + +K + >Ca 2+ >Mg 2+ ), whereas in the case of anions bicarbonate dominates the water chemistry followed by chloride and sulphate (HCO 3 − >Cl − >SO 4 2− ).As the basin is geologically carbonate-dominated, concentrations of bicarbonate are much higher than the other anions which is true for other basins like Gandaki (Pant et al., 2018).The diamond plot reveals the water type falls in class 1(Ca-HCO 3) and 3(Ca-Mg-Cl) for this season too.
Similarly, in post-monsoon, cations composition follows different patterns for RCP 4.5NF (Ca 2+ >Na + +K + >Mg 2+ ) and for the rest of the three scenarios Na + +K + > Mg 2+ > Ca 2+ .The domination of sodium and potassium ions could be due to the dissolution of evaporite and the weathering of silicates (Galy and France-Lanord, 1999).Water type falls in classes 3 and 6 for all scenarios in future periods based on the concentration of post-monsoon water quality data.For RCP8.5MF, sodium and potassium ions have shown many variations concerning current scenarios in both seasons.Also, for both seasons, it is interesting to note that in all scenarios and future periods anions composition of water is dominated by bicarbonate and follows the pattern for current scenarios (HCO 3 − >Cl − >SO 4 2− ).

Conclusions
The water quality of the Marshyangdi Watershed has been analyzed based on the statistical summary of physicochemical and hydrogeochemical status for current and future scenarios for medium and pessimistic representative concentration pathways for two seasons (pre-and post-monsoon).All the physicochemical parameters were within the acceptable limit for the sustenance of aquatic life for current and future scenarios for both RCPs except total phosphate.The exceedance limit of this

Figures Figure 1
Figures

Figure 2 Current
Figure 2 composition of river water is mostly in uenced by carbonate weathering, with partial contribution from silicate weathering.Further, Eyankware et al. (2016) assessed the anthropogenic activities on the hydrogeochemical quality of water resources of Ekaeru Inyimagu and its environs at Ebonyi State of Nigeria.The study reveals that a decline in water quality poses a major health challenge to the inhabitant of the study area.In addition, similar studies on the hydrogeochemical composition of ChuTalas River basin located in Central Asia revealed the dominancy of calcium bicarbonate as the major ion (Ma et al., 2020).Similar composition (CaMg-HCO 3 ) was also observed in the Shiyang River of China mainly from the weathering of silicates and carbonates (Zhang et al., 2021).However, there were very few hydrogeochemical investigations conducted in Nepal, most of which were on rivers and lakes that were situated at high altitudes.For example, in the Khumbu and Imja Khola at elevations ranging from 4530 to 5480 m, Tartari et al. (1998) examined the chemistry of 31 lakes while taking atmospheric loads, the geo-lithological and morphometric properties of the watershed, and surface waters into account.According to the study, the ionic content of the water in those lakes was mostly caused by the weathering of the rocks in the watershed.The study also revealed a direct connection between weathering events and the existence of glaciers, indicating that global atmospheric warming may have decreased the length of snow cover, increasing the contact time between

Table 2
Statistical summary of current physicochemical condition for pre-and post-monsoon Note: All values are expressed in mg/L except Temp (°C), EC (µS/cm), and Turbidity (NTU); SD is the Standard deviation

Table 3a
Characterization of seasonal variations in physicochemical conditions in near-future(RCP 4.5) ; Meybeck 2003,) which re ects the predominance of carbonate weathering in the watershed.The anion plot reveals that most samples fall on the left corner indicating the HCO 3 type, and Boenish et al.,2009).High concentrations of Ca 2+ and Mg 2+ in the water could be related to the weathering of crystalline dolomitic limestones.A similar Ca 2+ dominance also has been observed in a snow-fed Himalayan Gandaki River Basin (Pant et al., 2018).The studied major ions in the Marshyangdi Watershed indicate that the observed chemical species are of natural origin.
Ghezzi et al. (2019)g 2+ ) than silicate under natural conditions.Similarly, based on the ternary plot, Wolff-Boenish et al. (2009) also observed the predominance of calcium and magnesium in TSS and GHS-drained watersheds, which was supported byGhezzi et al. (2019)on the basis of the dominance of Mg 2+ and Ca 2+ and, to a lesser extent, HCO 3 -at super cial water sites of THS and GHS domain except at two sites in THS domain (Mg-Ca-HCO 3 -SO 4 ).