Uncorrected Proof

© 2023 The Authors AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 1 doi: 10.2166/aqua.2023.176

Long-term hydrological simulation for the estimation of snowmelt contribution of

Alaknanda River Basin, Uttarakhand using SWAT

Kuldeep Singh Rautelaa, Dilip Kumara, *, Bandaru Goutham Rajeev Gandhia, Ajay Kumara and Amit Kumar Dubeyb
a
Department of Civil Engineering, G B Pant Institute of Engineering and Technology, Pauri (Garhwal), Uttarakhand, India
b
Space Applications Centre-ISRO, Ahmedabad, Gujarat, India
*Corresponding author. E-mail: jhadilip27@gmail.com

ABSTRACT

A large population depends on water resources generated due to runoff from Himalayan River basins. They provide enough water for drink-
ing, domestic, industrial, and irrigation. Also, these rivers have a high hydropower potential. A lack of in-depth studies has made it difficult to
understand how these rivers respond hydrologically to climate change (CC) and, thus, impact the environment. In this paper, modelling the
Alaknanda River Basin (ARB) using the Soil and Water Assessment Tool (SWAT) has been conducted to understand the hydrological response
and assess its water balance components. The result shows that the basin’s water yield and Evapotranspiration (ET) vary from 58 to 63% and
34 to 39% of precipitation, respectively. The average annual contribution of snowmelt to the total riverine flow will range from 20 to 24%
throughout the simulation period. SFTMP, TLAPS, SMTMP, CN2, SMFMX, and GW_DELAY is found to be most sensitive at the significance
level of less than 0.05, showing the contribution of the snowmelt is significant in streamflow, while delay in the groundwater will affect
the contribution of surface runoff and groundwater in the streamflow. Based on the results, it is highly recommended that the spatial
and temporal hydro-meteorological should be investigated in-depth.

Key words: ARB, hydrological modelling, IHR, snowmelt, sustainable strategies, SWAT

HIGHLIGHTS

• A methodology has been proposed to obtain the streamflow pattern of a high-altitude river.
• The model can also derive the snowmelt contribution to the total streamflow.
• The SWAT model is applied to the snow domination basin of the Himalayan region of Uttarakhand. The evaluation of the results shows that
the model can obtain streamflow fluctuation.

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/).

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 2

GRAPHICAL ABSTRACT

LIST OF ABBREVIATIONS

ARB Alaknanda River Basin

SWAT Soil and Water Assessment Tool
ET Evapotranspiration
SFTMP Snowfall temperature
TLAPS Temperature lapse rate
PLAPS Precipitation lapse rate
CN curve number
SMTMP Snow melt base temperature.
SMFMX maximum melt rate for snow during the year (occurs on summer solstice).
SMFMN minimum melt rate for snow during the year (occurs on winter solstice).
TIMP snow pack temperature lag factor.
SNOCOVMX minimum snow water content that corresponds to 100% snow cover.
SNO50COV snow water equivalent that corresponds to 50% snow cover.
SOL_AWC available water capacity of the soil layer
ESCO soil evaporation compensation factor
ALPHA_BF base-flow recession coefficient
RCHRG_DP deep aquifer percolation fraction
GWMQN Threshold depth of water in the shallow aquifer required for return flow to occur (mm)

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 3

GW_DELAY groundwater delay

REVAPMN Threshold depth of water in the shallow aquifer for ‘revap’ to occur (mm)
GW_REVAP Groundwater ‘revap’ coefficient
CH_K2 Effective hydraulic conductivity in main channel alluvium
CWC Central Water Commission
TRMM Tropical Rainfall Measurement Mission
IMD Indian Meteorological Department
ASTER DEM Advanced Spaceborne Thermal Emission and Reflection Radiometer Digital Elevation Model
CUP Calibration and Uncertainty Programs
IHR Indian Himalayan Region

1. INTRODUCTION
The mountains covered with snow and glaciers are the early indicators of climate change (CC) (Kuniyal et al. 2021). The gla-
ciers in the Himalayas are one of the largest glacier-mountain systems in the world outside the polar region, with a length of
2,400 km and a width of 150–400 km (Reilly et al. 1996; Hasnain 2002; Bahuguna et al. 2014; Rautela et al. 2022a). There are
approximately 9,500 glaciers in the Indian Himalayan Region (IHR), covering a land area of about 40,000 km2 (Sangewar &
Shukla 2009). In the region of the Indian Himalayas, most rivers, streams, springs, and lakes are fed by the significant con-
tribution from the melting of glaciers and snow, and the basins of these hold a special place in the high mountain ecosystems
(Srivastava 2007; Scott et al. 2019; Rautela et al. 2022a). In the upstream catchments of the major rivers, such as Ganga,
Indus, Brahmaputra, and so on, the glacier and snow melt contribute most of the headwater (NRC 2012). It is usually
found that snow is temporarily stored in high mountains, and melted water is released into rivers later in the summer. As
glacier and snow runoff is necessary for major Himalayan river systems to remain perennial, rainfall volume made up in
the monsoon season is responsible for the high flow levels of these rivers (Tayal 2019). The snow accumulation in this
region will start from November to March, while this snow’s ablation will occur from April to September (Bisht et al.
2020; Rautela et al. 2020). From April to June, snowmelt runoff in the mountainous basins is a more dominant streamflow
component, and it accounts for a significant portion of streamflow from July to September (Jain et al. 2010). The snowmelt
runoff contributes 5% of streamflow as compared to the streamflow generated by the rainfall-runoff in the country (Schaner
et al. 2012; Raina & Srivastava 2014). This shows that snowmelt runoff is a good distributers of freshwater to the downstream
regions throughout the year (Ramanathan 2011; Rautela et al. 2022a). The process of snowmelt and snow accumulation is
largely affected due to CC and global warming. These processes will create a disturbance in the hydrological cycle of
upstream river basins of the Himalayan region due to the impact on precipitation patterns and temperature (Gebre and
Ludwig 2014; Kaini et al. 2019). The spatial and temporal variations in the precipitation and temperature pattern will signifi-
cantly affect the downstream regions in terms of water availability and associated water-related disasters (Nepal et al. 2014;
Kaini et al. 2020). The average annual inflows of various snow and glacier-fed rivers will increase with respect to CC since
2050, and consequently, the flow will decrease after 2050, and these perennial rivers will become seasonal rivers (IPCC 2018;
ICIMOD 2020; Kaini et al. 2020) The changes in the hydrological regime of these rivers will directly or indirectly affect the
1/5th of worlds populations those depends on these rivers (ICIMOD 2020).
In mountainous environments, hydro-meteorological conditions are highly variable over long periods and require physically
realistic and computationally efficient either distributed or semi-distributed modelling (Liston & Elder 2006). To understand the
hydrological characteristics of the mountainous basins, it is most common to describe the snowmelt distribution in the basin,
thus allowing for each watershed’s topography to be accounted for (Li et al. 2013, 2015, 2016). The various hydrological models
have been developed to simulate the hydrological response subject to both solid and liquid precipitation, such as Mike Zero
(NAM and SHE), and the public-domain models, such as HBV, Xinanjiang Model, HEC-HMS, SRM, and SWAT, and so on.
The SWAT model was used very commonly in water resources (Bergstrom 1992; Zhao & Liu 1995; Neitsch et al. 2011). The
benefit of utilizing the public-domain models is that these are freely available and simple to share model arrangements ( Jain
et al. 2017). SWAT provides a user-friendly interface for model setup in a GIS framework. Also, the SWAT model offers a
more extensive user base and a detailed user manual to the users for processing multiple processes (Jain et al. 2017).
SWAT is a continuous, semi-distributed model which has been used to simulate different hydrological responses using pro-
cess-based equations for daily, monthly, and yearly time series (Nasiri et al. 2020; Rautela et al. 2022b). SWAT is applied to
the catchment with an area of a couple of square kilometers to a thousand square kilometers (Spruill et al. 2000; Zhang et al.
2008). Several studies show the applications of SWAT for the modelling of snowmelt (Panhalkar 2014; Jain et al. 2017; Gupta

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 4

et al. 2018; Kumar & Bhattacharjya 2020), Rainfall–Runoff (Tripathi et al. 1999a; Shawul et al. 2013; Addis et al. 2016;
Himanshu et al. 2017), sediment transport (Tripathi et al. 1999b; Srivastava et al. 2004), and to estimate the hydropower
of a river-based in the results of SWAT (Pandey et al. 2014; Tamm et al. 2016). Many other model frameworks, such as
energy budget with distributed approach, have also been used to model snow-fed catchments. Still, this modelling requires
larger input datasets that are sometimes unavailable for Himalayan catchments. It is crucial to model the hydrological charac-
teristics of Himalayan river basins for various reasons (Kumar & Bhattacharjya 2021). These rivers provide water to nearly 2
billion people (Prakash 2020). These rivers have a high hydropower potential due to their perennial nature and steep terrain,
but developing that potential requires a solid knowledge of hydrologic response mechanisms (Pandey et al. 2015; Ghosh
2018). Moreover, these basins are prone to water-caused disasters like flash floods (Shrestha & Bajracharya 2013). Hydrolo-
gical data are scarce in the region due to its complicated topography.
Moreover, land-use and land-cover (LULC) changes could significantly impact the amount of snow and ice accumulated,
melting, and the hydrological response of these river basins (Miller et al. 2012). For the long-term planning of water resources,
it is necessary to study the effects of global warming on snow and glacier melt (NRC 2012). The hydrology of these rivers has
not been well studied, despite being common and highly important to human existence. In the present study, an attempt is
made to fill this void. The specific objective of the study is to simulate the long-term hydrological response of the Alaknanda
River Basin (ARB) and its effects on the basin’s water balance components on two different time scales, daily and monthly,
using SWAT. Since the river Alaknanda will be largely affected by CC and anthropogenic impacts. The findings of the present
study will provide a piece of important information about the contribution of the various hydrological components tothe gen-
eration of the streamflow.

2. STUDY AREA
The Alaknanda river system (Figure 1) is the significant upstream of the river Ganges that arises at the confluence and is fed
by Satopath and Bhagirathi Kharak glaciers in the Uttarakhand state of India. The Alaknanda river travels a distance of 195
km through the Chamoli, Rudraprayag, and Pauri districts of Uttarakhand and after that, confluences in the Bhagirathi river
and forms the Ganges and the river system drains an area of 11,063.68 km2. The river also plays a significant role from the
cultural point of view, and at the confluence of the major tributaries of the river, the Panchprayags (Vishnuprayag, Nand-
prayag, Karnaprayag, Rudraprayag, and Devprayad) is located. The main tributaries of the Alaknanda river are Saraswati
(meets at Mana), western Dhauliganga (meets at Vishnuprayag), Nandakini (meets at Nandprayag), Pinder (meets at Karna-
prayag), and Mandakini (meets at Rudraprayad). The combination of intense neotectonic activities and extreme rainfall has
shaped the geomorphology of the Alaknanda basin into steep slopes, high relief, and a high drainage density (Chopra et al.
2012; Rautela et al. 2022c). Mountainous terrain makes the basin subject to microclimates, and temperatures vary seasonally
and spatially (from river valleys to higher altitude regions). In ARB, Tungnath has the lowest average daily temperature of
0.5 °C in January and the highest average daily temperature of 30 °C in June in Srinagar (Panwar et al. 2017). The monsoon
brings more than 80% of the annual rainfall to India during the summer months of June to September (Kumar et al. 2010).
Consequently, the Alaknanda basin often witnesses cloud bursts, flash floods, and landslides due to heavy rainfall and
narrow valleys. The tributaries contribute a high streamflow of water to the river, including the western Dhauliganga, Nan-
dakini, Pinder, and Mandakini. Snowmelt, glacier melt, and seasonal rainfall are the main contributors to the perennial flows
in these tributaries. According to Strahler (1964) classification, the Alaknanda River makes a dendritic drainage pattern with
an order of 6th, and the average slope of the basin is approximately 30°C (Figure 1). In terms of LULC, water bodies, forest,
grass, agricultural land, residential area, barren land, and permanent snow covers an area of 0.41, 65, 2, 0.80, 1.70, 17.40, and
11%, respectively. Alaknanda basin has substantial hydropower potential from an economic perspective. Using the Ala-
knanda River and its tributaries as a renewable energy source, the SNDRP (2021) notes 37 hydropower dams are
operating, under construction, or planned.

3. MATERIALS AND METHODS

3.1. Data
3.1.1. Spatial data
Digital Elevation Model (DEM) is the essential input parameter for each hydrological model. The DEM stores information
about the geographic grid of a particular area, where each pixel of the grid describes a specific location with its elevation

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 5

Figure 1 | Location map of the study area.

(Figure 2(a)). In ArcSWAT, DEM is used to delineate the catchment boundary, sub-catchments boundary, stream network
generation, and identification of the catchment slope. In this study, the ASTER DEM for resolution 30 m has been used.
Also, the DEM is further used to classify the study area based on the elevation (Figure 2(b)). The LULC data for the catchment
area have been prepared using Sentinal-2 imaginary of 10 m resolution on Erdass imagine 2014 (Figure 2(c)). The study area

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 6

Figure 2 | (a) Digital elevation model, (b) elevation zones, (c) LULC classification, (d) soil classification, (e) slope classification, and (f) average
yearly precipitation (in mm) received by the Alaknanda river basin.

is further divided into 10 land-use classes using supervised classification. Soil data sets were acquired from Harmonized
World Soil Database (HSWDS) – Food and Agriculture (FAO), and the classification of the soil was done according to
the universal soil classification (Figure 2(d)). The catchment slope was classified into five categories such as 0–3, 3–10,
10–15, 15–30, and 30–9,999 m (Figure 2(e)).

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 7

3.1.2. Meteorological data

The meteorological parameters of the catchment are the most important dataset for modelling the hydrological processes.
The meteorological parameters such as solar radiation (S), relative humidity (RH), wind speed (W), and max–min temperature
(Tmax and Tmin) were taken on a daily time scale from MERRA-2 (Table 1), whereas the precipitation data (Figure 2(f)) will be
acquired from TRMM and IMD gridded weather data.

3.1.3. Hydrological data

The hydrological data of Alaknanda river at the outlet of the basin was acquired for a period of 1982–2016 on both a daily and
monthly period from the Central Water Commission (CWC) (Table 1). The gauging station was installed at the Devprayag
before the confluence of the Alaknanda and Bhagirathi. These measured data were used to correlate with the simulated
streamflow, and the model was calibrated and validated based on it.

3.2. Model setup

3.2.1. Snowmelt modelling
The SWAT model estimates snow accumulation and melting using a temperature-index approach. The melt received by snow
at the outlet of the basin is calculated based on the difference between the maximum and threshold temperature of the snow-
pack. During the calculation of infiltration and runoff, snowmelt is combined with rainfall. SWAT does not specifically
address snowmelt processes in frozen soils, but a provision is made for adjusting infiltration and estimating runoff when
soils are frozen (Neitsch et al. 2011). Despite this limitation, SWAT is still thought to be an appropriately integrated approach
for dealing with a wide range of problems. Many existing models are unable to simulate both snowmelt and rainfall–runoff
processes together. According to the temperature-index model, the temperature significantly affects snowmelt (Hock 2003).
The computation of the snowmelt is as follows (Equation (1)):


(Tsnowi þ Tmaxi
SNOmlti ¼ bmlti :SNO(covi)  SMTMP (1)
2

where bmlti is the melt factor (mm H2O-day), SNOmlti is the amount of snowmelt (in mm of H2O), Tmaxi is the maximum air
temperature (°C) on an ith day (mm of water), and SMTMP (°C) is allowable snowmelt base temperature.
The infiltration and runoff rates are computed using precipitation and snowmelt. In addition, precipitation is classified
according to the average air temperature as a threshold value. In the hydrological response unit (HRU), if the average
daily temperature falls below the SFTMP (also called the critical temperature), the precipitation is considered solid precipi-
tation and it is added to the snowpack. The process of snowmelt and sublimation causes depletion of the snowpack and the
mass balance for the HRUs is given in the following equation.

SNOi ¼ SNOi1 þ Ps  Esubi  SNOmlti (2)

Table 1 | Sources of the dataset used in this study

Spatio-temporal
S.No. Data type Source resolution Description

1. Topography USGS earth data 30 m ASTER DEM

2. Land-use land USGS earth explorer- Sentinel-2A satellite 10 m Land-use classification
cover imaginary
3. Soils Harmonized World Soil Database (HSWDS) – Soil Classification
– Food and Agriculture (FAO)
4. Meteorological MERRA-2 Daily Relative Humidity, Solar radiation, Wind
speed, Max & Min Temperature
5. Meteorological TRMM Daily Precipitation
6. Hydrological Central Water Commission (CWC) Daily Streamflow data obtained at the gauging
Monthly station

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 8

where SNOi is the water content of snowpack, Ps is the snow precipitation equivalent to water, Esubi snow sublimation, and
SNOmlti is the snowmelt equivalent to water (mm of water).
The area depletion curve is accounted for the HRUs by taking the variable snow coverage as shown in Equation (3)
(Anderson 1976).


1
SNOi SNOi SNOi
SNOcovi ¼ þ exp cov1  cov2 : (3)
SNOCOVMX SNOCOVMX SNOCOVMX

where SNOcovi is the fraction of HRU area covered by snow, SNOi is the water content of the snowpack on an ith day,
SNOCOVMX is the minimum snow water content that corresponds to 100% snow cover (mm H2O), and cov1 and cov2
are coefficients that control the shape of the curve.

3.2.2. Elevation bands

The elevation and temperature determine the snowpack and snowmelt caused by changes in the orographic variation of pre-
cipitation and temperature. Several studies have found that elevation plays a significant role in determining temperature and
precipitation variations (Zhang et al. 2008). A modified snowfall–snowmelt routine for mountainous terrain was introduced
into SWAT by Fontaine et al. (2002). Due to this modification, each sub-basin can be partitioned into 10 elevation bands, and
snowpack and snowmelt can be simulated based on elevation. To adjust for temperature and precipitation, the following fac-
tors were used:

PB ¼ P þ (ZB  Z):dP=dZ (4)

TB ¼ T þ (ZB  Z):dT=dZ (5)

where P and PB are the measured precipitation at stations and mean precipitation in the elevation band (mm), respectively, T
and TB are themeasured temperature at stations and mean temperature (°C) in the elevation band respectively, Z and ZB are the
elevations and mean elevation of the band (m), dP/dZ and dT/dZ are the precipitation and temperature lapse rate in mm/km
and °C/km, respectively.

3.2.3. Modelling of catchment hydrology

The most important inputs to the SWAT model are topography, vegetation, soil properties, and land management practices
(Figure 3). An exponential function of the soil depth and water content is used in SWAT to determine actual soil-water evap-
oration. Runoff is computed using the modified Soil Conservation Service (SCS) CN method. Also, the runoff calculation
takes into account both canopy infiltration and the amount of snow on the ground. To ensure that soil–water processes,
such as evaporation, infiltration, lateral flow, plant uptake, and percolation to lower layers are supported, numerous layers
are present in the soil profile. A downward flow occurs when the field capacity of a soil layer exceeds, and the layer
below is not saturated. Recharge of the shallow aquifer occurs through percolation from the soil profile. In parallel with
the percolation rate, the lateral subsurface flow is calculated. A shallow aquifer storage component is routed to a stream
to simulate the contribution of groundwater to total streamflow. The runoff is routed through the channel network using
either the variable storage routing technique or the Muskingum routing technique (Neitsch et al. 2011).
A main possible technique for assessing water yield, nutrient, and sediment circulation is to simulate the hydrologic cycle,
which incorporates the overall water circulation within the catchment (Equation (6)).

X
t
SWt ¼ SWo þ (Rv  Qs  Wseep  ET  Qlat  Qgw ) (6)
i¼1

where SWt is the Soil humidity/final water content, SW is the base humidity/initial soil–water content, Rv is the rainfall
volume, Qs is the surface runoff, Wseep is the seepage of water in the underlying soil layer, ET is the Evapotranspiration,
Qlat is the amount of lateral flow, Qgw is the amount of return flow on an ith day (mm water), and t is time in days.

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 9

Figure 3 | Flow chart of the methodology for simulation of streamflow.

3.2.4. Model simulation

The model simulation was done using the ArcSWAT 2012 interface using 35 years of streamflow data in two different time
scales, namely daily and monthly have been used for this study.
The number of years to skip (NYSKIP) also called the model warm up period has been taken as 4 years (1982–1985).

3.2.5. Model calibration and validation

Sequential Uncertainty Fitting (SUFI-2) algorithm has been adopted for calibration and validation of simulated streamflow
with the measured streamflow using SWAT-CUP 2012 for daily and monthly time step (Abbaspour et al. 2015). SWAT-
CUP is a computerized program used to calibrate, validate, and analyse sensitivity for various hydrological processes simu-
lated by SWAT. SWAT-CUP program uses various algorithms such as sequential uncertainty fitting (SUFI), particle swarm
optimization (PSO), generalized likelihood uncertainty estimation (GLUE), Parasol, and Markov chain Monte Carlo
(MCMC) to improve the model performance by changing the model parameters systematically. Different parameters such
as Soil parameters, groundwater, basin, HRU, and channel roughness parameters have been introduced to improve the
model performance. The calibration period was set from 1986 to 2007 for 22 years. Furthermore, optimized values of
these improved parameters are used for the validation of the streamflow data. The validation period was kept to be nine
years from 2008 to 2016. The assessment of strength of the simulated streamflow was measured by the p-factor and

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 10

r-factor. When the p-factor approaches unity and the r-factor approaches zero, the model is considered an ideal hydrological
model (Abbaspour et al. 2007). For the streamflow simulation, a value of p-factor greater than 0.7 is adequate (Abbaspour et al.
2015). Also, the model performance was measured by the coefficient of determination (R 2) (Equation (7)), Nash–Sutcliffe effi-
ciency (NSE) (Equation (8)), modified coefficient of determination (bR 2) (Equation (9)), ratio of the standard deviation of
observations to root mean square error (RSR) (Equation (10)), King–Gupta efficiency (KGE) (Equation (11)), and percentage
of bias (PBIAS) (Equation (12)). The simulated streamflow was optimized by setting Nash–Sutcliffe efficiency (NSE) as an
objective function.

 2
P
t
(Qmi  Qm )(Qsi  Qs )
i¼0
R2 ¼ (7)
P
t P
t
(Qmi  Qm ) (Qsi  Qs )
i¼0 i¼0

P
t
(Qmi  Qs )2
i¼0
NSE ¼ 1  (8)
P
t 2
(Qmi  Qm )
i¼0

lblR2 if lbl  1
bR2 ¼ (9)
lbl1 R2 if lbl . 1
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Pt
(Qmi  Qsi )2
RMSE i¼1
RSR ¼ ¼ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (10)
STDEVo Pt
(Qmi  Qm )
i¼1
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
KGE ¼ 1  (r  1)2 þ (a  1)2 þ (b  1)2 (11)

P
t
(Qmi  Qs )
PBIAS ¼ i¼0  100 (12)
P
t
Qmi
i¼0

where Qm and Qs are the mean measured and simulated streamflow during the period, Qmi and Qsi is the measured and
simulated streamflow in the ith day (cum/s), respectively, and n is the number of data points, R is the correlation coefficient,
b is the coefficient of regression, r is the linear regression coefficient between observed and simulated streamflow, a ¼ ss =sm
and b ¼ ms =mm , where σm and σs are the standard deviations of the observed and simulated data, respectively, and μm and μs
are the mean of observed and simulated data, respectively. The model is considered satisfactory if R 2 . 0.55, NSE .0.5,
bR 2  0.4, RSR  0.7, KGE  0.5, and PBIAS + 0.25 (Moriasi et al. 2007; Akhavan et al. 2010; Muleta 2012; Abbaspour
et al. 2015; Mehdi et al. 2015; Kouchi et al. 2017).

3.2.6. Sensitivity analysis

The SWAT is a continuous complex semi-distributed model that requires a large number of parameters. In the calibration of
streamflow, sensitivity analysis plays an important role in finding the suitable parameters (Imani et al. 2019). Sensitivity analy-
sis is a technique used to identify parameters which have a significant influence on model performance (Holvoet et al. 2005).
Local sensitivity analysis (LSA) uses the one-at-a-time (OAT) methodology, which analyses the impact of a single parameter at
a time, while keeping the others unchanged (Abbaspour et al. 2017). Whereas, using multiple regression analysis, the Latin
hypercube-generated parameters were regressed on the goal function values so that the global sensitivity (GS) of model par-
ameters could be estimated (Arnold et al. 2012). A GS analysis was performed to evaluate the sensitivity of the calibrated
model parameters. To assess the importance of each calibrated parameter, the statistical t-test and p-value have been con-
ducted. Sensitivity is considered to be higher for larger t-values. If the p-value approaches zero, the model is considered
significant (Abbaspour et al. 2015).

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 11

4. RESULT AND DISCUSSION

The basin of the Alaknanda river is sub-divided into five sub-basins, namely, Alaknanda Main Basin (AMB), Western Dhau-
liganga Basin (WDB), Nandakini Basin (NB), Pinder Basin (PB), and Mandakini Basin (MB) in the direction of the main river
(Figure 4). The model divides the land-use land cover into 10 sub-classes (Figure 2(c)). In addition, 0.42, 33.99, 0.92, 0.76,
32.98, 1.67, 17.39, 10.52, 1.34, and 0.01% of the catchment area is covered by water, evergreen forests, grassland, agricultural
land, mixed forest with small shrubs, built-up area, barren land, permanent snow and ice, pastureland and deciduous forest,
respectively. The watershed is divided into three sub-classes of soils. About 61.63% of the total catchment area is covered by
soil type I-Bh-U-C-3717 (Loam) (Soil hydrologic group C) followed by 23.63% area is covered by soil type Bd29-3c-3661
(clay_loam) (Soil hydrologic group C) (clay_loam) and 15.24% area is covered by soil type Glacier-6998 (UWB) (Soil hydro-
logic group D) (Figure 2(d)). The soils are categorized in group C and D which allows very less infiltration and are responsible
for the quick generation of surface runoff. In the last step, the model divides the catchment slope into the 5 classes as 0–3,
3–10, 10–15, 15–30, and 30–9,999 m with the help of DEM. About 47.96% of the total area is covered by the slope range
60–9,999 m followed by the 35.95% area is covered by a range 30–60 m (Figure 2(e)) which shows the catchment has very
steep slopes. A total of 122 nos. of HRU has been created by the unique combinations of 5 land-use classes, 3 soil classes,
and 5 slope classes for the 5 sub-watersheds of the Alaknanda River. Four land-use classes such as WATR (water), INDN
(grassland), AGRR (Agricultural land) and FRSD (Deciduous forests) have been neglected because their value is below
the threshold limit.
The components of the hydrological cycle obtained after the simulation are shown in Figure 5. The precipitation received
by the catchment is very high due to the circulation of the South-West monsoon. The Potential Evapotranspiration (PET) is
also quite high because of the high vegetative cover over the catchment area. The average CN for the catchment was found to
be 77.72 (Figure 5). The estimated surface runoff for the catchment was 160 mm. The streamflow/precipitation ratio

Figure 4 | Different sub-watersheds of the Alaknanda river along the mainstream.

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 12

Figure 5 | Water balance components after the initial simulation of the SWAT model.

(runoff–rainfall ratio) was 0.62, which is also satisfactory for this region. The surface runoff of the catchment is high due to the
high precipitation values received by the catchment. All the water balance ratios are under the permissible limits and are
shown in Figure 5.

4.1. Calibration and validation of the hydrological simulation

The model calibration was done by using the thumb rule. According to the thumb rule the number of simulations depends on
the number of parameters used in the calibration (Abbaspour et al. 2015). A minimum of 100 simulations was used for each
parameter during the calibration phase of the model. The number of iterations depends on the model evaluation parameters.
If the model satisfies the range of statistical indices the calibration process was stopped. The model evaluation parameters
such as R 2, NSE, bR 2, RSR, KGE and PBIAS for the initial calibration were found to be 0.20, 0.20, 0.08, 1.10, 0.25, 71.7
and 0.48, 0.26, 0.22, 0.86, 0.33 54.9, respectively, for daily and monthly time step (Table 3). The result of initial calibration
indicates there is a very high inconsistency between the observed and simulated streamflow. The SUFI algorithm uses five
elevation bands and 19 parameters to improve consistency, high flow, low flows, and model efficiency. Using the thumb
rule for the iterations, the optimum values of the calibrated model parameters have been adjusted. The range and the opti-
mum fitted values of the model parameters are shown in Table 2. The Precipitation lapse rate (PLAPS) was calibrated
within 1,000–2,000 mm, and it was fitted at 1,468.50 and 1,739.47 mm, respectively, for the daily and monthly time steps.
The temperature lapse rate (TLAPS) was calibrated within 7 to 6 °C and fitted at 6.25 °C for daily and monthly time
steps (Table 2). These two parameters are used to adjust the precipitation and temperature rates according to the elevation;
as a result, snowmelt contribution in the streamflow is improved. The Snowfall temperature (SFTMP) was calibrated within
5 to 5 °C and fitted at 4.06, and 2.29 °C, Snowmelt base temperature (SMTMP) was calibrated within 5 to 5 and fitted at
3.48 and 2.71 °C, maximum melt factor (SMFMX) was calibrated within 1–10 and fitted at 2.40 and 1.31, minimum melt
factor (SMFMN) was calibrated with and 1–10 and fitted at 6.91 and 1.54, and snowpack temperature lag factor (TIMP)
was calibrated within 1 to 1 and fitted at 0.19 and 0.47, respectively, for daily and monthly time step (Table 2). These
five parameters have also improved the snowmelt contribution in the streamflow. The minimum snow water content that

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 13

Table 2 | Description of the parameters with fitted values for daily and monthly simulated streamflow

S. No. Parameter with qualifier Minimum range Maximum range Fitted value (daily) Fitted value (monthly)

1 V__PLAPS.sub 1,000 2,000 1,468.75 1,739.47

2 V__TLAPS.sub 7 6 6.09 6.25
3 V__SFTMP.bsn 5 5 4.06 2.29
4 V__SMTMP.bsn 5 5 3.44 2.71
5 V__SMFMX.bsn 1 10 2.41 1.31
6 V__SMFMN.bsn 1 10 6.91 1.54
7 R__TIMP.bsn 1 1 0.19 0.47
8 R__SNO50COV.bsn 0 0.25 0.07 0.16
9 V__SNOCOVMX.bsn 0 200 156.25 60.53
10 R__CN2.mgt 0.3 0.3 0.08 0.10
11 R__SOL_AWC().sol 0 0.2 0.13 0.18
12 R__ESCO.hru 0.05 0.1 0.06 0.10
13 R__ALPHA_BF.gw 0 0.25 0.20 0.19
14 R__RCHRG_DP.gw 0 0 0.00 0.00
15 V__GWQMN.gw 0 20 9.38 19.00
16 V__GW_DELAY.gw 10 50 11.25 28.00
17 V__REVAPMN.gw 0 100 96.88 96.58
18 V__GW_REVAP.gw 0 0 0.00 0.00
19 V__CH_K2.rte 5 10 7.34 6.43

corresponds to 100% snow cover (SNOCOVMAX) was calibrated within 0–200 mm and fitted at 156 and 60.53 mm, and the
fraction of snow volume represented by SNOCOMX corresponds to 50% snow cover (SNO50COV) was calibrated within
0–0.25 and fitted at 0.07 and 0.16, respectively, for daily and monthly time step (Table 2). These two parameters show
there is significant snowfall in the basin. The curve number CN2 value was calibrated within 0.3 to 0.3, and it was fitted
at 0.09 and 0.10. The available water capacity of the soil layer (SOL_AWC) was calibrated within 0–0.2 mm water/mm
soil. It was fitted at 0.13 and 0.18 mm water/mm soil, and the soil evaporation compensation factor (ESCO) was calibrated
within 0.05–0.2. It was fitted at 0.06 and 0.10, respectively, for daily and monthly time steps (Table 2). These three parameters
indicate the peak surface runoff and streamflow were improved due to a small increase in the CN2 and higher water uptake
demand from the lower soil layer. The threshold water level in a shallow aquifer for base flow (GWQMN) was calibrated
within 0–20 mm. It was fitted at 9.37 and 19 mm, the groundwater ‘revap’ coefficient (GE_REVAP) was set to be zero 0.0,
and the percolation of the surface runoff to the deep aquifer (REVAPM) was calibrated within 0–100 mm. It was fitted at
96.87 and 96.58 mm. Re-evaporation through ground water (GW_REVAP) was set to be zero, delay in the ground water
(GW_DELAY) was calibrated within 10–50 days and fitted at 11 and 28 days, respectively, for daily and monthly time
step and base-flow recession coefficient (ALPHA_BF) was calibrated within 0–0.25. It was fitted at 0.19, respectively, for
both daily and monthly time steps (Table 2). The calibrated groundwater parameters indicate that a 19 mm water level is
required for base flow in the shallow aquifer while 96 mm is the threshold depth of water in the shallow aquifer percolate
to the deep aquifer, and the maximum amount of groundwater will transfer to the overlaying saturated zone from the shallow
aquifer whereas (Singh & Saravanan 2020), the very less response of groundwater flow to changes in recharge because of the
presence of impermeable rocks present in the Himalayan region. The effective hydraulic conductivity of the channel (CH_K2)
was calibrated within 5–10 mm/h, and it was fitted at 7.34 and 6.43 mm/h, respectively, for daily and monthly time steps
(Table 2) and shows there is a very less rate of surface water from the main channel. After calibration, the model evaluation
parameters such as R 2, NSE, bR 2, RSR, KGE, and PBIAS were found to be 0.60, 0.55, 0.46, 0.68, 0.75, 13.2 and 0.75, 0.74,
0.64, 0.51, 0.86 and 10.50, respectively, and also the p-factor and r-factor for the calibration period were found to be 0.79, 1.21
and 0.79, 1.30, respectively, for daily and monthly time step (Table 3) due to these, the peak flow, as well as base flow, has

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 14

Table 3 | Model performance evaluation for daily and monthly simulation

Performance evaluation parameters R2 NSE bR 2 RSR KGE PBIAS p-factor r-factor

Daily Initial calibration 0.20 0.20 0.08 1.10 0.25 71.7 – –

Calibration 0.60 0.54 0.46 0.68 0.75 13.2 0.79 1.21
Validation 0.65 0.59 0.50 0.56 0.79 7.5 0.73 1.30
Monthly Initial calibration 0.48 0.26 0.22 0.86 0.33 54.9 – –
Calibration 0.75 0.74 0.64 0.51 0.86 10.50 0.79 1.30
Validation 0.82 0.78 0.64 0.48 0.82 9.25 0.87 1.35

been improved (Figures 6(a) and 7(a)) also the correction of the observed streamflow and simulated streamflow is improved
(Figure 8(a) and 8(b)). Further, the calibrated parameters were used to validation of the streamflow for the period 2008–2016.
The R 2, NSE, bR 2, RSR, KGE, and PBIAS were found to be 0.65, 0.59, 0.50, 0.56, 0.79, 7.5 and 0.82, 0.78, 0.64, 0.51, 0.86,
9.25, respectively, for daily and monthly time step and also the p-factor and r-factor for the validation period was found to be
0.73, 1.30 and 0.87, 1.35, respectively, for daily and monthly time step (Table 3). In the validation, some peaks of the stream-
flow hydrograph were not properly captured by the model, whereas the model captured the lower peaks due to the lower
range of CN2 (Figures 6(b) and 7(b)), and the results show the observed and simulated discharge shows the good correlation

Figure 6 | (a) Calibration and (b) validation of streamflow on the daily time step.

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 15

Figure 7 | (a) Calibration and (b) validation of streamflow on monthly time step.

for the validation period (Figure 8(c) and 8(d)). The efficiency parameters are also improved for calibration as well as the
validation period.

4.2. Sensitivity analysis

For simulated streamflow, the calibrated parameters were used for the sensitivity analysis, and the GS analysis was used for
the monthly and daily periods. A higher t-value and lesser p-value indicate the most sensitivity at a 0.05 significance level
(Abbaspour et al. 2017). The calibrated SWAT parameters have been used for the sensitivity analysis, and apart from that,
six parameters such as SFTMP, TLAPS, SMTMP, CN2, SMFMX, and GW_DELAY are found most sensitive at a significance
level less than 0.05 during the daily and monthly simulated streamflow, respectively (Figure 9). Out of six sensitive par-
ameters, four parameters are snow parameters, and the result shows the contribution of the snowmelt is significant in
streamflow while the CN2 and delay in the groundwater will affect the contribution of surface runoff and groundwater in
the streamflow.

4.3. Water balance components of the basin

Due to the high elevation range in the Alaknanda basin, the snowmelt is found as a predominant factor in the streamflow.
Results of this study show the snowmelt’s contribution in the total streamflow ranges between 20 and 24% (Table 4), whereas
the contribution of rainfall is also high in the streamflow, which also ranges between 10 and 36% (Table 4). The loss of water
due to ET will range from 34 to 39% of the total precipitation received by the catchment (Figure 10(a) and 10(b)). The ground-
water contribution of the basin is also good due to the contribution of springs to the streamflow. These springs are recharged
during the monsoons and provide an adequate amount of water to the streamflow. A very little amount of water percolates
into the deep aquifer due to the presence of hard strata. The components of the water yield are shown in Figure 10(a) and
10(b). It also shows there is a heavy loss of water through ET. An adequate amount of runoff is found at the basin outlet
due to the combined effect of rainfall, snowmelt, and groundwater flow. Also, recharge of shallow aquifers in the basin
due to precipitation ranges from 15 to 17%, 2% of water is re-evaporated through groundwater, and only 1% of water is
used to recharge the deep aquifers (Figure 10(a) and 10(b)).

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 16

Figure 8 | Correlation between measured and simulated streamflow (a) for calibration on daily time step, (b) validation on daily time step,
(c) calibration on monthly time step, and (d) validation on monthly time step.

Figure 9 | Sensitive parameters used in the simulation of hydrological response of ARB.

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 17

Table 4 | Water balance components for Alaknanda basin

Components Initial simulation Final simulation

Precipitation 786.3 689.3

Snow fall 74.61 70.57
Snow melt 74.44 70.49
Sublimation 0.32 0.29
Surface runoff 160 121.95
Lateral soil 194.73 169.33
Groundwater (shallow aquifer) 135.78 106.31
Groundwater (deep aquifer) 8 6.38
Re-evaporation through Shallow aquifer 15.71 16.14
Deep aquifer recharge 7.97 6.44
Total aquifer recharge 159.42 128.82
Total water yield 498.52 403.98
Percolation out of soil 159.35 128.78
ET 272.9 270
PET 786 807
All units are in mm

Figure 10 | Water balance components for (a) initial simulation (b) final simulation.

Singh & Jain (2002) conducted a modelling analysis for the Satluj basin, using the SNOWMOD snowmelt runoff model to
cover the Satluj basin up to the Bhakra dam site, downstream of Rampur Site. They discovered that the average annual con-
tributions of snowmelt and rainfall to streamflow are around 68 and 32%, respectively. Singh & Jain (2002) determined that
the contribution from snow and glaciermelt is about 59% of the yearly flow and the contribution from rainfall is around 41%
in another study of this basin at the Bhakra Dam site (Downstream of Rampur). According to Khajuria et al. (2022), a sig-
nificant portion of streamflow is produced throughout the summer and monsoon seasons, with a contribution from
snowmelt ranging from 10 to 45%. Kumar et al. (2021) assess the CC impact on snowmelt runoff of the Alaknanda river
using SRM. They discovered a 20% and 2°C rise in the precipitation and temperature would result in a 37 and 53% rise in
the streamflow. Based on previous researches, the Alaknanda region experiences the lowest snow cover percentage during
the summer months, while February sees the highest snow cover area percentage of the entire year, at about 77%. Changes
in snow cover and snow depth are common events that have a significant impact on snowmelt flow. As the snow cover begins

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 18

to go, the snow’s depth also decreases, the volume of melting rises, and there is an increase in runoff from snowmelt. Snow
cover starts to thaw after February and persists through the end of November, which can be a major factor in increasing Snow
melt Runoff at this time of year. The Snowmelt runoff starts to grow from June, peaking inJuly to August, and decreases until
the end of December. However, the snowmelt contribution up to the Devprayag location is equivalent to previous research-
ers’ estimates in the current study (Singh & Jain 2002, 2003; Jain et al. 2010, 2017; Sukla et al. 2019).

4.4. Sustainable strategies for land and water resource management

Numerous studies from the last four decades have indicated the stress on land and water resource of the mountains under
increasing population coupled with changing land use, and CC. Development and management of water resources continue
to be central to the fight for sustainable growth, economic development, and reduced poverty. Results presented in this study
illustrate the significant contribution of snowmelt in the streamflow along with the preponderance of groundwater delay that
shall affect the inputs from surface runoff and groundwater towards streamflow in the Alaknanda basin. This commends to
very favourable condition(s) for artificial recharge of groundwater via site suitable Groundwater Augmentation Measures
(GAMs). A groundwater recharge potentiality map for the Area of Interest (AoI), considering most of the influencing par-
ameters like slope, soil, geology, lineaments, LULC, etc., can be prepared following RS and GIS techniques. This shall aid
in the identification of sites of various recharge capabilities within the AoI. Based on the developed map, attempt towards
delineating apposite GAMs like the use of recharge pits and trenches in areas of high recharge potentiality, masonry and
live check dams for the region of moderate potentiality, and seeding of grasses, stone bunds, plantation of broad-leaf hard-
wood, etc. over the regions of low potentiality which are usually accompanied by excessive slopes and relatively
impervious stratum, can be suggested. However, the downstream regions of Himalayan rivers largely depend on these
rivers in terms of irrigation, and due to CC the irrigation patterns and cropping intensity of these rivers are largely affected
(Suhardiman et al. 2018; Kaini et al. 2020). On the other hand, various socioeconomic factors such as household size, occu-
pation, landholding, and food sufficiently from land affect crop intensity (Kanin et al. 2020). The Indian state of
Uttarakhand’s government has acknowledged the need to use the potential for irrigation and hydropower fully. The aforesaid
work, therefore, not only aligns with the ethics of field applicability and practicality towards conserving groundwater and
surface water resources in mountains and fluvial valleys but also offers baseline data of seasonal water availability for
water resource planners and policymakers in the formulation of sustainable land and water management strategies, thus,
securing future water sustainability under changing land use and climate.

5. CONCLUSIONS
This study has attempted a long-term hydrological simulation to understand the hydrological response of a Himalayan River
viz; before the confluence of Alaknanda River at Devprayag. The model evaluation parameters for both calibration and vali-
dation would be considered as good for both periods. In general, the hydrograph shape was reproduced satisfactorily, except
for some peaks and recession limbs that were difficult to reproduce. Therefore, a basin situated in the Himalayas can thus be
considered to have a good water balance model through SWAT, which allows modelling of streamflow hydrographs and other
components in a basin. The water yield of the basin is found to be 40%, ET ranges between 29 and 33%, of the total precipi-
tation received by the basin. The contribution of the snowmelt in the total streamflow ranges between 20 and 24%, whereas
the contribution of rainfall is also high in the streamflow, which also ranges between 10 and 36%. In the lower Alaknanda
basin, interflow has a significant contribution to the streamflow. However, it is necessary to supplement these results with
more detailed hydrologic modelling of additional river basins to study their response mechanisms. The importance of upgrad-
ing spatial, soil, and hydro-meteorological databases and monitoring precipitation (rain and snow) and other climatic
variables at different elevations should be given greater consideration for the distributed hydrological modelling. Further-
more, isotope analysis can be performed and compared with hydrologic models to separate the components of runoff.
This study provides the baseline data for identifying the flood peaks and should be used to develop a model for flash
floods in the different sub-basins of the ARB with the availability of detailed hydro-meteorological data sets.

DATA AVAILABILITY STATEMENT

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

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 19

CONFLICT OF INTEREST
The authors declare there is no conflict.

REFERENCES

Abbaspour, K. C., Yang, J., Maximov, I., Siber, R., Bogner, K. & Mieleitner, J. et al. 2007 Modelling hydrology and water quality in the pre-
alpine/alpine Thur watershed using SWAT. Journal of Hydrology 333 (2–4), 413–430.
Abbaspour, K. C., Rouholahnejad, E., Vaghefi, S. R., Srinivasan, R., Yang, H. & Kløve, B. 2015 A continental-scale hydrology and water
quality model for Europe: calibration and uncertainty of a high-resolution large-scale SWAT model. Journal of Hydrology 524, 733–752.
Abbaspour, K., Vaghefi, S. & Srinivasan, R. 2017 A guideline for successful calibration and uncertainty analysis for soil and water assessment:
a review of papers from the 2016 International SWAT Conference. Water 10, 6. https://doi.org/10.3390/w10010006.
Addis, H. K., Strohmeier, S., Ziadat, F., Melaku, N. D. & Klik, A. 2016 Modeling streamflow and sediment using SWAT in Ethiopian
Highlands. International Journal of Agricultural and Biological Engineering 9 (5), 51–66.
Akhavan, S., Abedi-Koupai, J., Mousavi, S. F., Afyuni, M., Eslamian, S. S. & Abbaspour, K. C. 2010 Application of SWAT model to investigate
nitrate leaching in Hamadan–Bahar Watershed, Iran. Agriculture, Ecosystems & Environment 139 (4), 675–688.
Anderson, E. A., 1976 A point energy and mass balance model of snow cover. In: Development of a Snowfall-Snowmelt Routine for
Mountainous Terrain for the Soil Water Assessment Tool (SWAT) (Fontaine, T. A., Cruickshank, T. S., Arnold, J. G. & Hotchkiss, R. H.,
eds).
Arnold, J. G., Srinivasan, R., Muttiah, R. S. & Williams, J. R. 1998 Large area hydrologic modeling and assessment part I: model development
1. JAWRA Journal of the American Water Resources Association 34 (1), 73–89.
Arnold, J. G., Moriasi, D. N., Gassman, P. W., Abbaspour, K. C., White, M. J. & Srinivasan, R. et al. 2012 SWAT: model use, calibration, and
validation. Transactions of the ASABE 55, 1491–1508. https://doi.org/10.13031/2013.42256.
Bahuguna, I. M., Rathore, B. P., Brahmbhatt, R., Sharma, M., Dhar, S. & Randhawa, S. S. et al. 2014 Are the Himalayan glaciers retreating?
Current Science 1008–1013.
Bergstrom, S. 1992 The HBV Model – Its Structure and Applications.
Bisht, H., Kotlia, B. S., Kumar, K., Arya, P. C., Sah, S. K., Kukreti, M. & Chand, P. 2020 Estimation of suspended sediment concentration and
meltwater discharge draining from the Chaturangi glacier, Garhwal Himalaya. Arabian Journal of Geosciences 13 (6), 1–12.
Chopra, S., Kumar, V., Suthar, A. & Kumar, P. 2012 Modeling of strong ground motions for 1991 Uttarkashi, 1999 chamoli earthquakes, and
a hypothetical great earthquake in Garhwal–Kumaun Himalaya. Natural Hazards 64 (2), 1141–1159.
Fontaine, T. A., Cruickshank, T. S., Arnold, J. G. & Hotchkiss, R. H. 2002 Development of a snowfall–snowmelt routine for mountainous
terrain for the soil water assessment tool (SWAT). Journal of Hydrology 262 (1–4), 209–223.
Ghosh, N. 2018 Hydropower in the Himalayas: the economics that are often ignored. The Third Pole 29.
Gupta, C., Jain, S., Shivhare, V. & Jain, N. 2018 Hydrological stream flow modelling in himalayan basin under climate change. Journal of
Indian Water Resources Society 37 (4).
Hasnain, S. I. 2002 Himalayan glaciers meltdown: impact on South Asian Rivers. International Association of Hydrological Sciences,
Publication 274, 417–423.
Himanshu, S. K., Pandey, A. & Shrestha, P. 2017 Application of SWAT in an Indian river basin for modeling runoff, sediment and water
balance. Environmental Earth Sciences 76 (1), 1–18.
Hock, R. 2003 Temperature index melt modelling in mountain areas. Journal of Hydrology 282 (1–4), 104–115.
Holvoet, K., Griensven, A. V., Seuntjens, P. & Vanrolleghem, P. A. 2005 Sensitivity analysis for hydrology and pesticide supply towards the
river in SWAT. Physics and Chemistry of the Earth 30, 518–526.
ICIMOD 2020 Formation of Glacial Lakes in the Hindu Kush-Himalayas and GLOF Risk Assessment. Kathmandu, Nepal.
Imani, S., Delavar, M. & Niksokhan, M. H. 2019 Identification of nutrients critical source areas with swat model under limited data
condition. Water Resources 46 (1), 128–137.
Immerzeel, W. W., Pellicciotti, F. & Bierkens, M. F. P. 2013 Rising river flows throughout the twenty-first century in two Himalayan
glacierized watersheds. Nature Geoscience 6 (9), 742–745.
IPCC 2018 Global Warming of 1.5 °C. Special Report Intergovernmental Panel on Climate Change.
Jain, S. K., Goswami, A. & Saraf, A. K. 2010 Assessment of snowmelt runoff using remote sensing and effect of climate change on runoff.
Water Resources Management 24 (9), 1763–1777.
Jain, S. K., Jain, S. K., Jain, N. & Xu, C. Y. 2017 Hydrologic modeling of a Himalayan mountain basin by using the SWAT mode. Hydrology
and Earth System Sciences Discussions 1–26.
Kaini, S., Gardner, T. & Sharma, A. K. 2020 Assessment of socio-economic factors impacting on the cropping intensity of an irrigation
scheme in developing countries. Irrigation and Drainage 69 (3), 363–375.
Kaini, S., Nepal, S., Pradhananga, S., Gardner, T. & Sharma, A. K. 2021 Impacts of climate change on the flow of the transboundary Koshi
River, with implications for local irrigation. International Journal of Water Resources Development 37 (6), 929–954.
Khajuria, V., Kumar, M., Gunasekaran, A. & Rautela, K. S. 2022 Snowmelt runoff estimation using combined terra-aqua MODIS improved
snow product in Western Himalayan River Basin via degree day modelling approach. Environmental Challenges 8, 100585.

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 20

Kouchi, D. H., Esmaili, K., Faridhosseini, A., Sanaeinejad, S. H., Khalili, D. & Abbaspour, K. C. 2017 Sensitivity of calibrated parameters and
water resource estimates on different objective functions and optimization algorithms. Water 9 (6), 384.
Kumar, D. & Bhattacharjya, R. K. 2020 Evaluating two GIS-based semi-distributed hydrological models in the Bhagirathi-Alkhnanda River
catchment in India. Water Policy 22 (6), 991–1014.
Kumar, D. & Bhattacharjya, R. K. 2021 Change in rainfall patterns in the hilly region of Uttarakhand due to the impact of climate change.
Applied Environmental Research 43 (1), 1–3.
Kumar, V., Jain, S. K. & Singh, Y. 2010 Analysis of long-term rainfall trends in India. Hydrological Sciences Journal–Journal des Sciences
Hydrologiques 55 (4), 484–496.
Kumar, R., Manzoor, S., Vishwakarma, D. K., Al-Ansari, N., Kushwaha, N. L., Elbeltagi, A., Sushanth, K., Prasad, V. & Kuriqi, A. 2021
Assessment of climate change impact on snowmelt runoff in himalayan region. Sustainability 14 (3), 1150.
Kuniyal, J. C., Kanwar, N., Bhoj, A. S., Rautela, K. S., Joshi, P. & Kumar, K. et al. 2021 Climate change impacts on glacier-fed and non-glacier-
fed ecosystems of the Indian Himalayan Region: people’s perception and adaptive strategies. Current Science 120 (5), 888–899.
Li, H., Beldring, S., Xu, C. Y., Huss, M., Melvold, K. & Jain, S. K. 2015 Integrating a glacier retreat model into a hydrological model–Case
studies of three glacierised catchments in Norway and Himalayan region. Journal of Hydrology 527, 656–667.
Li, H., Xu, C. Y., Beldring, S., Tallaksen, L. M. & Jain, S. K. 2016 Water resources under climate change in Himalayan basins. Water
Resources Management 30 (2), 843–859.
Li, L., Engelhard, M., Xu, C. Y., Jain, S. K. & Singh, V. P. 2013 Comparison of satellite based and reanalysed precipitation as input to glacio
hydrological modeling for Beas river basin, Northern India. Cold and mountain region hydrological systems under climate change:
towards improved projections. IAHS Publication 360, 45–52.
Liston, G. E. & Elder, K. 2006 A distributed snow-evolution modeling system (SnowModel). Journal of Hydrometeorology 7 (6), 1259–1276.
Mehdi, B., Ludwig, R. & Lehner, B. 2015 Evaluating the impacts of climate change and crop land use change on streamflow, nitrates and
phosphorus: a modeling study in Bavaria. Journal of Hydrology: Regional Studies 4, 60–90.
Miller, J. D., Immerzeel, W. W. & Rees, G. 2012 Climate change impacts on glacier hydrology and river discharge in the Hindu Kush–
Himalayas. Mountain Research and Development 32 (4), 461–467.
Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D. & Veith, T. L. 2007 Model evaluation guidelines for systematic
quantification of accuracy in watershed simulations. Transactions of the ASABE 50 (3), 885–900.
Muleta, M. K. 2012 Model performance sensitivity to objective function during automated calibrations. Journal of Hydrologic Engineering
17 (6), 756–767.
Nasiri, S., Ansari, H. & Ziaei, A. N. 2020 Simulation of water balance equation components using SWAT model in Samalqan Watershed
(Iran). Arabian Journal of Geosciences 13, 1–15.
National Research Council 2012 Himalayan Glaciers: Climate Change, Water Resources, and Water Security. National Academies Press.
Neitsch, S. L., Arnold, J. G., Kiniry, J. R. & Williams, J. R. 2011 Soil and Water Assessment Tool Theoretical Documentation Version 2009.
Texas Water Resources Institute.
Pandey, A., Lalrempuia, D. & Jain, S. K. 2015 Assessment of hydropower potential using spatial technology and SWAT modelling in the Mat
River, southern Mizoram, India. Hydrological Sciences Journal 60 (10), 1651–1665.
Panhalkar, S. S. 2014 Hydrological modeling using SWAT model and geoinformatic techniques. The Egyptian Journal of Remote Sensing and
Space Science 17 (2), 197–207.
Panwar, S., Agarwal, V. & Chakrapani, G. J. 2017 Morphometric and sediment source characterization of the Alaknanda river basin,
headwaters of river Ganga, India. Natural Hazards 87 (3), 1649–1671.
Prakash, A. 2020 Retreating Glaciers and Water Flows in the Himalayas: Implications for Governance, Vol. 400. Observer Research
Foundation, New Delhi, India, pp. 1–14.
Raina, V. K. & Srivastava, D. 2014 Glacier atlas of India. GSI Publications 7 (1).
Ramanathan, A. L. 2011 Status report on Chhota Shigri Glacier (Himachal Pradesh), department of science and technology, ministry of
science and technology, New Delhi. Himalayan Glaciology Technical Report 1, 88.
Rautela, K. S., Kuniyal, J. C., Kanwar, N. & Bhoj, A. S. 2020 Estimation of stream hydraulic parameters and sediment load in river Neola in the
foothills of the panchchuli glacier during the ablation period. Journal of Himalayan Ecology and Sustainable Development 15, 114–125.
Rautela, K. S., Kuniyal, J. C., Alam, M. A., Bhoj, A. J. & Kanwar, N. 2022a Assessment of daily streamflow, sediment fluxes, and erosion rate of a pro-
glacial Stream Basin, Central Himalaya, Uttarakhand. Water, Air, & Soil Pollution 233, 136. https://doi.org/10.1007/s11270-022-05567-z.
Rautela, K. S., Kumar, M., Sofi, M. S., Kuniyal, J. C. & Bhat, S. U. 2022b Modelling of streamflow and water balance in the Kuttiyadi River
Basin using SWAT and remote sensing/GIS tools. International Journal of Environmental Research 16 (4), 1–4.
Rautela, K. S., Kumar, D., Gandhi, B. G., Kumar, A. & Dubey, A. K. 2022c Application of ANNs for the modeling of streamflow, sediment
transport, and erosion rate of a high-altitude river system in Western Himalaya, Uttarakhand. RBRH 27.
Reilly, J., Baethgen, W., Chege, F. E., Van De Geijn, S. C., Iglesias, A. & Kenny, G. et al. 1996 Agriculture in a changing climate: impacts and
adaptation. In: Climate Change 1995; Impacts, Adaptations and Mitigation of Climate Change: Scientific-Technical Analyses.
Cambridge University Press, pp. 427–467.
Sangewar, C. V. & Shukla, S. P. 2009 Inventory of the Himalayan Glaciers – A Contribution to International Hydrological Programme, An
Updated Edition. Geological Survey of India, Kolkata, p. 34.

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest
 Uncorrected Proof

AQUA — Water Infrastructure, Ecosystems and Society Vol 00 No 0, 21

Schaner, N., Voisin, N., Nijssen, B. & Lettenmaier, D. P. 2012 The contribution of glacier melt to streamflow. Environmental Research Letters
7 (3), 034029.
Scott, C. A., Zhang, F., Mukherji, A., Immerzeel, W., Mustafa, D. & Bharati, L. 2019 Water in The Hindu kush Himalaya. In: The Hindu Kush
Himalaya Assessment. Springer, Cham, pp. 257–299.
Shawul, A. A., Alamirew, T. & Dinka, M. O. 2013 Calibration and validation of SWAT model and estimation of water balance components of
Shaya mountainous watershed, Southeastern Ethiopia. Hydrology and Earth System Sciences Discussions 10 (11), 13955–13978.
Shrestha, A. B. & Bajracharya, S. R. 2013 Case Studies on Flash Flood Risk Management in the Himalayas: in Support of Specific Flash Flood
Policies. International Centre for Integrated Mountain Development (ICIMOD).
Shrivastava, P. K., Tripathi, M. P. & Das, S. N. 2004 Hydrological modelling of a small watershed using satellite data and GIS technique.
Journal of the Indian Society of Remote Sensing 32 (2), 145–157.
Shukla, S., Jain, S. K. & Kansal, M. L. 2021 Hydrological modelling of a snow/glacier-fed western Himalayan basin to simulate the current
and future streamflows under changing climate scenarios. Science of The Total Environment 795, 148871.
Singh, S. 2018 Alakhnanda–Bhagirathi River system. In: Book: The Indian Rivers, pp. 105–114.
Singh, P. & Jain, S. K. 2002 Snow and glacier melt in the Satluj River at Bhakra Dam in the western Himalayan region. Hydrological Sciences
Journal 47 (1), 93–106.
Singh, P. & Jain, S. K. 2003 Modelling of streamflow and its components for a large Himalayan basin with predominant snowmelt yields.
Hydrological Sciences Journal 48 (2), 257–276.
Singh, L. & Saravanan, S. 2020 Simulation of monthly streamflow using the SWAT model of the Ib River watershed. India. HydroResearch 3,
95–105.
SNDRP 2021 Hydroelectric Projects on Alaknanda River Basin, by South Asia Network on Dams Rivers and People. Available from: https://
sandrp.in/tag/alaknanda-river/.
Spruill, C. A., Workman, S. R. & Taraba, J. L. 2000 Simulation of daily and monthly stream discharge from small watersheds using the SWAT
model. Transactions of the ASAE 43 (6), 1431.
Srivastava, V. K. 2007 River ecology in India: present status and future research strategy for management and conservation. Proceedings-
Indian National Science Academy 73 (4), 255.
Strahler, A. N. 1964 Quantitative Geomorphology of Drainage Basins and Channel Networks. Handbook of Applied Hydrology. McGraw-
Hill, New York, pp. 4–39.
Suhardiman, D., Pavelic, P., Keovilignavong, O. & Giordano, M. 2018 Putting farmers’ strategies in the centre of agricultural groundwater use
in the Vientiane Plain, Laos. International Journal of Water Resources Development 36 (1), 149–169.
Tamm, O., Luhamaa, A. & Tamm, T. 2016 Modeling future changes in the North-Estonian hydropower production by using SWAT.
Hydrology Research 47 (4), 835–846.
Tayal, S. 2019 Climate change impacts on himalayan glaciers and implications on energy security of India. In: TERI Discussion Paper. The
Energy and Resources Institute New Delhi.
Tripathi, M. P., Panda, R. K. & Raghuwanshi, N. S. 1999a Estimation of sediment yield from a small watershed using SWAT model. In:
Proceedings of the Civil and Environmental Engineering Conference on ‘New Frontiers and Challenges, pp. 8–12.
Tripathi, M. P., Panda, R. K. & Raghuwanshi, N. S. 1999b Runoff estimation from a small agricultural watershed using SWAT model. In:
Hydrological Modelling: Proceedings on International Conference on Water, Environment, Ecology, Socio-Economics and Health
Engineering (WEESHE), pp. 143–152.
Zhang, X., Srinivasan, R., Debele, B. & Hao, F. 2008 Runoff simulation of the headwaters of the Yellow River using The SWAT model with
three snowmelt algorithms 1. JAWRA Journal of the American Water Resources Association 44 (1), 48–61.
Zhao, R. J. & Liu, X. R., 1995 The Xinanjiang model. In: Computer Models of Watershed Hydrology (Singh, V. P., ed.). Water Resources
Publication, pp. 215–232.

First received 11 October 2022; accepted in revised form 23 December 2022. Available online 25 January 2023

Downloaded from http://iwaponline.com/aqua/article-pdf/doi/10.2166/aqua.2023.176/1163795/jws2023176.pdf

by guest