case study Yamuna River Basin in Delhi

Journal of Hydrology: Regional Studies 56 (2024) 101960
Contents lists available at ScienceDirect
Journal of Hydrology: Regional Studies

journal homepage: www.elsevier.com/locate/ejrh
Hydrological and hydrodynamic modelling for flood management:

A case study of the Yamuna River Basin in Delhi
Jatin Anand a,b,* , A.K. Gosain b, R. Khosa b
a
Department of Civil Engineering, National Institute of Technology Rourkela, Rourkela, India
b
Department of Civil Engineering, Indian Institute of Technology Delhi, New Delhi 110016, India
A R T I C L E I N F O A B S T R A C T
Keywords: Study region: Yamuna River (Delhi), India.

Flood Study focus: The anthropogenic activities within the vicinity of the floodplain reduce the river’s
Hydrodynamic model margin and subsequently alter the magnitude of the river’s flow. The encroachment of riverbeds
IRIC
leads to waterlogging and flooding in urban areas, thereby causing damage to property, human
Water level
Yamuna River
life, etc. It necessitates a comprehensive study of the floodplain and changes in its proximity such
as encroachment of floodplains to carry out any further activities with certainty. This study
employs a two-dimensional model to simulate the Yamuna River’s (YR) hydrodynamic charac
teristics, focusing on India’s Delhi region.
New hydrological insights: Simulated flood flows are employed to evaluate floods of once in 10, 20,
25, and 30-year return periods using the flood frequency analysis for 1951–2013. The model
validation results indicated that the model could mimic the flood depth in YR. Simulation results
revealed that the floodplain’s encroachment had increased the severity of the floods. The increase
in the extremeness of flooding events, i.e., from once in a 10-year return period to a 30-year
return period event, is expected to increase the areas at risk of floods by 12 %. The model also
offers a potential platform for evaluating other alternatives, such as further encroachment, for a
business-as-usual scenario or for restoring the Yamuna floodplains. With such a comprehensive
perspective, floodplains’ role enhances river basin resilience to climate and anthropogenic
changes and increases flood safety.
1. Introduction
Rivers have been fundamental in sustaining human life and the development of civilizations throughout history. They provide
floodplains with essential sediments and nutrients that are crucial for agricultural and economic development (Gee et al., 1990; Krause
et al., 2007). However, river flooding has emerged as a significant problem worldwide, impacting vast areas and affecting millions of
people annually. In recent years, the severity and frequency of floods have generally worsened (Arrault et al., 2016; Gripp et al., 2014;
Soomro et al., 2024). According to reports by the United Nations, floods alone have affected approximately 2.3 billion people globally
over the past two decades (Wahlstrom and Guha-Sapir, 2015). This issue is particularly relevant in India, which has experienced severe
floods in recent years, leading to widespread devastation and loss of life (Mishra and Shah, 2018; Wahlstrom and Guha-Sapir, 2015).
* Corresponding author at: Department of Civil Engineering, National Institute of Technology Rourkela, Rourkela, India
E-mail addresses: anandj@nitrkl.ac.in, cez138430@civil.iitd.ac.in (J. Anand), gosain@civil.iitd.ac.in (A.K. Gosain), rakesh.khosa@gmail.com
(R. Khosa).
https://doi.org/10.1016/j.ejrh.2024.101960
Received 11 April 2024; Received in revised form 29 July 2024; Accepted 31 August 2024
Available online 17 September 2024
2214-5818/© 2024 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
J. Anand et al. Journal of Hydrology: Regional Studies 56 (2024) 101960
The increasing occurrence of floods highlights the urgent need for effective flood management and mitigation strategies to protect
vulnerable populations and sustain economic growth.
It is well-recognized that floodwater storage within floodplains can significantly reduce the magnitude of flood levels downstream
(Johnson et al., 2020; Schober et al., 2014; Sutcliffe and Parks, 1989). The interaction of flood flow with obstructed floodplains, often
due to anthropogenic activities, is a crucial factor influencing the frequency, extent, and severity of flooding events. This phenomenon,
known as the "levee effect," can exacerbate flood risks by preventing the natural dispersion of floodwaters (Domeneghetti et al., 2015;
Tobin, 1995; White, 1945). In recent decades, many urban and agricultural development projects have transformed the floodplains of
several rivers, intensifying the impact of floods (De Paiva et al., 2013; Entwistle et al., 2019; Lai et al., 2013; Rajib et al., 2021; Stevaux
et al., 2009). Consequently, floodplains have become some of the most complex and poorly understood areas in flood risk management.
Understanding the dynamics of floodplains and implementing effective management strategies are critical to mitigating flood risks and
enhancing the resilience of floodplain ecosystems and communities.
Similar to other rivers worldwide, the Yamuna River’s floodplain in India has faced significant anthropogenic pressures. Major
urban areas and townships, particularly around the Delhi region, have expanded within these floodplains. Over the years, these areas
have seen a steady increase in infrastructure development, including civic structures, roads, bridges, and notable constructions like the
Commonwealth Games Village (Kumar et al., 2019; Mehra et al., 2000; Singh et al., 2013). Rapid urbanization and encroachment
activities along these floodplains have dramatically altered their natural form and water surface profiles. This transformation has been
driven by the need to accommodate growing urban populations and associated infrastructure demands. However, this expansion has
significantly impacted the floodplain’s hydrodynamics, increasing the risk and severity of flood events. Several studies have outlined
these changes, highlighting the need for sustainable urban planning and floodplain management to mitigate the adverse effects of such
developments (Entwistle et al., 2019; Khan and Bajpai, 2014; Tang et al., 2023; Zope et al., 2016).
Urbanization in the floodplain has led to several significant hydrological impacts, including amplified runoff, increased maximum
discharge, and the accumulation of runoff volume. These changes ultimately shorten the time it takes for floodwaters to reach their
peak discharge (Campana and Tucci, 2001; Kulkarni et al., 2014; Schober et al., 2020; Sun et al., 2022; Teo, 2010). Additionally,
low-lying floodplain areas frequently experience flooding due to the lack of natural drainage systems. The flow regimes of the Yamuna
River have changed markedly, with an increase in the trend of minimum streamflow and a declining trend in the ratio of flood peak
discharge to the lowest flow rate of the year (Rai et al., 2012). This shift has led to a typical phenomenon known as "small floods but big
disasters" (D’Angelo et al., 2020; Lai et al., 2013). The altered flow regimes and inadequate drainage systems underscore the need for
comprehensive floodplain management and urban planning strategies to mitigate the adverse effects of urbanization on flood
dynamics.
Numerical models are essential tools for flood risk management, providing an effective and practical basis for simulating the
complex flow processes in rivers and floodplains (Acreman et al., 2003; Jodhani et al., 2022; Kim et al., 2022; Lin et al., 2016). In
recent decades, large-scale hydrological and hydrodynamic models have been extensively employed (Lin et al., 2016; Soomro et al.,
2022; Wing et al., 2019; Zhang et al., 2020). While past research has focused on regional river systems or global subsystems, it has yet
to develop an efficient numerical framework for performing flood routing in macro-scale basins like the Yamuna River catchment
(Kumar et al., 2019; Singh et al., 2013; Tomar et al., 2021; Vijay et al., 2007). Several studies have incorporated hydrological
modelling to study the Yamuna Basin’s water system (Anand et al., 2018c; Narula and Gosain, 2013), which helps in understanding the
basin’s inherent hydrologic characteristics (Anand et al., 2021a; Arnold et al., 2012; Remondi et al., 2016; Wang et al., 2014).
However, these hydrological models have inherent disadvantages as they cannot depict complex hydrodynamic processes. The
behaviour of composite flow dynamics in natural river reaches can be better understood through physically-based hydrodynamic
models (Anand et al., 2021b; De Paiva et al., 2013; Lai et al., 2013; Paz et al., 2010).
Recently, both one-dimensional (1D) and two-dimensional (2D) hydraulic models have been employed to simulate river hydro
dynamic processes, providing a deeper understanding of flow behaviour and flood dynamics (Grimaldi et al., 2013; Moral
es-Hernández et al., 2016; Nguyen et al., 2016). Grimaldi et al. (2013) focused on the hydraulic simulation component and the
associated sources of uncertainty, highlighting the challenges in accurately modelling river hydraulics. Another significant contri
bution is from Nguyen et al. (2016), who developed the coupled hydrologic-hydrodynamic model, Hires Flood-UCI. This model links
the HL-RDHM as a catchment hydrologic model with the 2D hydrodynamic Brazzo model, effectively tested in the US Illinois River
Basin to simulate complex flood scenarios. Grimaldi et al. (2019) further proposed an arithmetical experimentation method to measure
the distribution of errors when integrating hydrological and hydrodynamic models in large basins. Their study involved the coupled
HBV hydrological model and the LISFLOOD-FP hydrodynamic model, providing insights into error distribution and model perfor
mance. Morales-Hernández et al. (2016) adopted a hybrid approach for the Tiber River near the urban area of Rome (Italy), coupling a
1D river flow model with a 2D floodplain model to simulate real flood scenarios effectively. This approach demonstrated the utility of
hybrid models in capturing the complexities of urban flooding and providing robust flood risk assessments.
This study uses the International River Interface Cooperative (iRIC) modelling framework, which allows the modelling of two- and
three-dimensional flow, sediment transport, and comprehensive data and model output visualization (Dhanapala et al., 2022; Morid
et al., 2020; Nelson et al., 2015, 2010). Numerical modelling of fluid flow is based on solving the principal laws of conservation of mass
and momentum within the system. Therefore, a hydrodynamic model is useful in simulating flood inundation problems. The inte
gration of hydrological and hydrodynamic models presents an integrated approach to address the limitations inherent in using either
model type independently. By coupling these two types of models, it is possible to leverage the strengths of each. The hydrological
model can provide realistic, spatially distributed input data, such as discharge and sediment load, which serves as boundary conditions
for the hydrodynamic model. This approach ensures that the flow conditions simulated by the hydrodynamic model are grounded in a
comprehensive understanding of watershed processes. Conversely, the detailed flow patterns generated by the hydrodynamic model
2
can inform and refine the hydrological model, leading to more accurate predictions of flood extents and water levels. Thus, there is a
need to appropriately couple hydrological and hydraulic models to enhance the efficiency and accuracy of flood simulations. By
integrating surface runoff data from hydrological models into hydraulic models, we can achieve a more comprehensive understanding
of flooding dynamics, leading to better-informed flood management and mitigation strategies. In the present study, the SWAT model
was employed as it is a useful hydrologic model for simulating hydrologic components, especially surface runoff of large watersheds
(Abbaspour et al., 2015; Anand et al., 2018a, 2018c). The recently developed 2D iRIC model was employed to simulate hydraulic
components because it could mimic flood inundation and is capable of simulating large floods (Dhanapala et al., 2022; Nelson et al.,
2015). Due to its newness, it is not widely used, but it can be used to model floods in regions where data are inadequate (Dhanapala
et al., 2022; Nelson et al., 2015). The iRIC model was adopted by Wongsa (2014) to study the flooding due to the 2011 Tsunami in
Japan and by Rai et al. (2018) in the Brahmani and Baitarani river delta, which witnesses frequent flood events during the monsoon
season. While, Dhanapala et al., (2022) employed a 2D hydraulic model (iRIC) to simulate flooding in the Nilwala River basin which is
prone to recurrent flooding. It is worth mentioning that the two models employed in the present study are open-source models. The
open-source interface of the models provides financial benefits to the users, enables reproducibility of the scientific results, and allows
the user to develop additional tools based on their needs and use (Pontes et al., 2017).
Urban development within floodplains could greatly benefit from a thorough understanding of the interactions between floodplain
hydrodynamics and natural phenomena. This research aims to develop a coupled hydrological and hydrodynamic model for flood
inundation. By integrating the SWAT (Soil and Water Assessment Tool) model with the International River Interface Cooperative
(iRIC) hydraulic model, the study seeks to achieve high-resolution floodplain inundation modelling. Additional applications of the
coupled model are to simulate past severe floods and analyze floods during various scenarios. The study surpasses previous research by
analyzing multiple flood events (1997, 1988, 1995, and 1978) corresponding to various return periods (10, 15, 20, and 30 years),
rather than simulating a single flood event. This comprehensive analysis allows for a better understanding of flood dynamics and their
interaction with natural phenomena, providing a more robust risk and hazard assessment framework. Key aspects of the novelty
include the integration of the SWAT model with the iRIC hydraulic model, offering a more detailed and accurate simulation of
floodplain hydrodynamics, and addressing the limitations of single-event analyses in earlier studies. By evaluating four significant
Fig. 1. Index Map of Yamuna River in Delhi.
3
flood events, the research provides a broader perspective on flood behaviour across different return periods, enhancing the under
standing of flood frequency and severity. The model’s application to real-world scenarios enables a practical evaluation of its effec
tiveness in risk and hazard assessment, making it a valuable tool for urban planning and floodplain management. The study specifically
addresses the impact of urban development on floodplain hydrodynamics of the Yamuna river, providing insights into how such
Fig. 2. Major Construction Within the Flood Plain.
4
developments can alter flood water levels and exacerbate flood hazards. Additionally, the research simulates past severe floods and
explores various flood scenarios, helping to predict and manage future flood risks under different conditions. The present study intends
to assess the research gap by addressing how the magnitudes of flood water levels in the Yamuna River will evolve under the influence
of expected alterations in flood plains. This paper assesses how the loss in the flood plain would impact the flood hazard.
2. Study area
The Yamuna River is an important tributary of the Ganges and rises from Saptarishi Kund in the lower Himalayas in the Uttarkashi
region of Uttarakhand, India. It is the primary source of drinking water in Delhi, in addition to serving several other towns and villages
in Uttar Pradesh and Haryana (Fig. 1). The river enters Delhi near the village of Palla, 24 km upstream from the Wazirabad barrage,
and leaves Delhi downstream of the Okhla Barrage at Jaitpur. The floodplain’s width is 2–3 km, although the river channel is limited to
400–800 m (Dalai, Krishnaswami, and Sarin, 2002; Nielsen et al., 1991; Singh, 2013; Vijay et al., 2007). In this study, the emphasis is
restricted only to the Palla-Jaitpur stretch of 49 Kilometers, as a high-resolution digital elevation model is available only for this
stretch.
The Yamuna River covers a distance of 1376 km to its confluence with the Ganges at Allahabad in Uttar Pradesh (Asim and
Nageswara Rao, 2021). The catchment area of the Yamuna River is about 34.585 million hectares. The soil is predominantly loamy;
however, sandy loam and sandy clay loam are also prevalent (Fig. 5) (Narula and Gosain, 2013). A significant part of the basin is under
agricultural land use (80 %). A large part of this agricultural land use is under irrigated agriculture, with wheat and rice being the
predominant crops (Fig. 5). The water table depth ranges from 3 m to 100 m below ground level, resulting in shallow or deep
groundwater aquifer conditions in the area (Khan et al., 2020).
The Yamuna River in India is far from ’pristine’ due to extensive human exploitation through abstraction, discharges, and im
poundments. Delhi’s climate features extreme summer and winter temperatures, ranging from 27.6 ◦ C to 44.2 ◦ C and 22.2 ◦ C to 3.5 ◦ C,
respectively. The city’s mean annual rainfall is about 720 mm, predominantly during the monsoon months of July to September. The
area’s hydrogeology is defined by two distinct aquifers: quartzite rock and alluvial aquifer (National Institute of Disaster Management
Report, 2013; Vijay et al., 2007).
Delhi’s growing population, coupled with business and daily wage earners migrating to the city, has led to increased urban
development within the floodplains, significantly expanding encroachment levels. This urban expansion includes settlements, com
mercial developments, and infrastructure projects, raising the vulnerability to flooding. Encroachment in the Yamuna floodplain,
particularly near Shahadra and Wazirabad, has surged. Informal settlements, such as Batla House and parts of Sonia Vihar (Fig. 2), are
notable squatters.
Additionally, significant development projects within the floodplain, like the Akshardham Temple, Commonwealth Games
infrastructure, Delhi Secretariat, Millennium Bus depot, and Noida toll bridge, have further degraded these areas (Fig. 2). Structures
near Rajghat, including the Millennium Bus depot, Asia’s largest, constructed during the Commonwealth Games, exemplify this trend
(National Institute of Disaster Management Report, 2013; Singh et al., 2006).
The streamflow of the Yamuna within Delhi is significantly impacted by discharge from Hathnikund and Tajewala barrages, located
approximately 200 km upstream. Delhi has faced several major floods in recent years, notably in 1975, 1978, 1988, 1995, and 2010.
The 1978 flood was particularly severe, with water levels at Wazirabad Barrage reaching 208.17 m and a discharge rate of 7500
cumecs, causing widespread devastation. Breaches in rural embankments led to agricultural flooding, and urban areas such as Model
Town and Mukherjee Nagar experienced significant property damage (National Institute of Disaster Management Report, 2013).
Similarly, in 1995, heavy rains caused another major flood, and in 2010, extreme flooding resulted from heavy rainfall in the upper
basin. The construction for the Commonwealth Games village has reduced the available floodplain area, exacerbating flood risks.
These events highlight the need for effective flood management strategies within the Yamuna basin (National Institute of Disaster
Management Report, 2013; Singh et al., 2013; Vijay et al., 2007).
3. Methodology and data description
3.1. SWAT model
SWAT is a continuous, physically distributed, parameter-based model that can model the repercussions of the various land man
agement exercises upon hydrology, sediment, and pollutant transport in an agrarian watershed (Arnold et al., 1998). SWAT breaks up
basins into smaller sub-basins, interconnected by a network of streams, and subsequently redistributes catchment into HRUs (hy
drological response units) with specific slope, soil, and land cover characteristics within each sub-basin. The SWAT model is capable of
simulating the surface and subsurface flow, sediment generation and deposition, and nutrient transports. SWAT simulates various
hydrologic processes, including surface runoff generation, for which SWAT either employs the SCS curve number method or the Green
and Ampt infiltration equation. For the simulation of evapotranspiration, the SWAT model offers various methodologies such as
Hargreaves, Penman-Monteith methods or Priestley-Taylor method. Groundwater flow, lateral flow and percolation are assessed
through a mass balance of the underlying system. The Penman-Monteith method is adopted to compute potential evapotranspiration in
the study. The SWAT model has been intensively employed to compute surface runoff and evaluate the repercussions of climate and
land-use transformations on the distribution of water and erosion in large-scale basins (Li et al., 2016; Marhaento et al., 2017; Anand
et al., 2018b). The water balance components are computed for each HRUs and subsequently aggregated to the sub-basin scale (Arnold
and Fohrer, 2005). In addition, SWAT offers two water routing approaches, the variable storage (Williams, 1969) and Muskingum
5
(Cunge, 1969) methods. SWAT modeling begins by calculating the amount of water from the land of each sub-catchment to the main
channel and is routed through the streams to the outlet using the Muskingum Routing Method. The simulation of the hydrological cycle
by SWAT includes surface runoff, infiltration and canopy storage which is based on the water balance equation given below:
t
∑
SWt = SW0 + (Ri − Qi − ETi − Gi − Bi ) (1)
i=1
where SWt (mm) is the final soil water content, SW0 (mm) is the initial soil water content on day i, t (days) is time, Ri (mm) is the
precipitation amount on day i, Qi (mm) is the amount of surface runoff on day i, ETi (mm) is the evapotranspiration (ET) amount on day
i, Gi (mm) is the amount of water entering the vadose zone from the soil profile on day i, and Bi (mm) is the amount of return flow on
Fig. 3. The geographical context of Yamuna River basin and calibration locations till Agra.
6
day i.
The setup of the SWAT model for Yamuna River necessitates multiple data inputs, including digital elevation models (DEMs), land
use and land cover data, soil properties, weather data (precipitation, temperature, solar radiation, wind speed, and humidity) (Fig. 3),
and hydrological data (streamflow and reservoir operations) to accurately simulate watershed processes. The catchment of Yamuna
River is delineated until the Agra gauge and discharge site. Automatic watershed delineation was performed using the DEM to
delineate the Yamuna River basin and estimate watershed characteristics such as slope, topography, and flow direction. The watershed
was then delineated using the outlet at the CWC gauge stations. The delineation process was meticulously carried out to ensure that the
locations of major dams, and diversion structures were accurately incorporated. The land use map, soil data and slope of the watershed
were then defined into the SWAT model for use in HRU definition. The threshold value of the drainage network is a crucial parameter
in the SWAT model, particularly during the watershed delineation process. This value dictates the minimum drainage area required to
form a stream, thereby influencing the density and extent of the stream network within the watershed. Essentially, the threshold value
sets the flow accumulation point at which a stream is initiated. In the present study, a threshold value of 5000 has been used during the
delineation process using the SWAT model. The SWAT model includes an integrated snow module designed to simulate snowmelt
hydrology effectively. This module allows the delineation of up to ten elevation bands, each with its associated precipitation and
temperature lapse rates (Fontaine et al., 2002; Grusson et al., 2015; Neitsch et al., 2011; Qi et al., 2016; Rahman et al., 2013). To
accurately determine the water budget, SWAT distinguishes between liquid and solid precipitation based on near-surface air tem
perature. In SWAT, the model estimates snowpack, snowfall, and snowmelt events whenever the temperature falls below the desig
nated snowfall threshold. These processes are spatially differentiated as a function of elevation, allowing users to specify lapse rates for
both precipitation (PLAPS in mm H₂O/km/yr) and temperature (TLAPS in ◦ C/km). This feature is particularly crucial in regions with
significant elevation variations, as it enhances the accuracy of hydrological simulations. In the present study, the sub-basin has been
divided into ten elevation bands to capture the spatial variability of these processes effectively, ensuring a more precise simulation of
the hydrological dynamics in the study area.
Up to the Agra gauge and discharge site, the basin is discretized into 52 sub-basins (Fig. 3) with 1000 Hydrologic Response Units
(HRUs). The discretization of sub-basins into regions with unique combinations of slope, soil, and landuse makes it conceivable to
account for variances in ET and other hydrological components (Anand et al., 2018c; Arnold et al., 1998). Hathnikund barrage is in the
upstream part of the basin (Fig. 1). From the Hathnikund barrage, the water is diverted into the Western and Eastern Yamuna Canals.
The release of these canals is incorporated into the model to take care of the influence of the Hathnikund barrage. From the Hathnikund
barrage, the water is diverted into the Western and Eastern Yamuna Canals. The release of these canals is incorporated into the model
to take care of the influence of the Hathnikund barrage. In the present study, the SWAT model is calibrated and validated against
observed streamflow at seven-gauge stations: Tuini, Kalanaur, Karnal, Mawi, Delhi Railway Bridge, Mathura and Agra, as shown in
(Figs. 1 and 3).
First, the gauge station in the upstream section is calibrated, then the parameters conforming to the upstream section are kept the
same, and the parameter for the catchments between the upstream gauge station and the subsequent downstream gauge station is
calibrated further. A similar methodology is repeated for subsequent gauge gauges until we reach the Yamuna basin’s outlet (Agra).
Various parameters have been varied for the calibration of the SWAT model: Elevation band parameters, viz. precipitation Lapse rate
and temperature Lapse rate; snow parameters such as Minimum and Minimum melt factor and Temperature; and Hydrological pa
rameters, namely curve number, groundwater delay time, etc. (Arnold et al., 2013; Neitsch et al., 2002) (Table S1 in supplementary
material).
The SWAT model was then calibrated and validated at each discharge location within the basin using various statistical perfor
mance measures. These measures include:
Nash-Sutcliffe Efficiency (NSE): This metric evaluates the predictive accuracy of the model by comparing the observed and
simulated data. The Nash-Sutcliffe Efficiency (NSE) index is a statistical metric used to evaluate the predictive accuracy of hydrological
models by comparing observed and simulated data. The NSE value indicates how well the plot of observed versus simulated data aligns
with a 1:1 line, where a value of NSE = 1 signifies a perfect fit, indicating that the model predictions exactly match the observed values.
Conversely, NSE values range from -∞ to 1.0, with higher values (closer to 1.0) representing better predictive performance (Moriasi
et al., 2007, 2015).
⎡∑ n ⎤
(X − Yi )2
⎢i=1 i ⎥
NSE = 1 − ⎢⎣∑ n
⎥
⎦ (2)
2
(Xi − X)
i=1
Coefficient of Determination (R2): The R2 statistic, also known as the coefficient of determination, quantifies the proportion of the
variance in the observed data that is explained by the model’s predictions. It ranges from 0 to 1, where a value of 1 indicates a perfect
model fit, meaning that the model explains all the variability in the observed data. Conversely, a value of 0 implies that the model does
not explain any of the variability, indicating a poor fit. The closer the R2 value is to 1, the better the model’s predictive accuracy. This
metric is essential for evaluating the performance of hydrological models, as it provides a clear indication of how well the model
captures the observed data patterns and variability (Moriasi et al., 2007, 2015).
7
( n ) ( n )( n )
∑ ∑ ∑
n Xi Y i − Xi Yi
i=1 i=1 i=1
R2 = √ ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
√[ ( ][ (3)
n
) ( n )2 ( n ) ( n )2 ]
√
√ n ∑ X2 − ∑
X i n
∑ 2
Y −
∑
Yi
i i
i=1 i=1 i=1 i=1
Root Mean Square Error to Standard Deviation of Measured Data (RSR): RSR combines both error measures, providing a normalized
measure of the model’s performance by the ratio of the root mean square error to the standard deviation of the observed data.
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
∑ n
(Oi − Pi )2
i=1
RSR = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ (4)
∑ n
(Oi − P)2
i=1
Percent Bias (PBIAS): Another crucial performance metric is the Percent Bias (PBIAS), which assesses the tendency of the simulated
data to be consistently higher or lower than the observed values. PBIAS is calculated as the percentage difference between the observed
and simulated data, providing insight into the model’s bias. A PBIAS value of 0 indicates a perfect match between the simulated and
observed data, while positive values indicate model underestimation and negative values indicate overestimation. According to
Moriasi et al., (2015), the performance of hydrological models is deemed satisfactory when certain criteria are met for flow
predictions.
n
∑
Oi − Pi
i=1
PBIAS = n × 100 (5)
∑
Oi
i=1
Specifically, model predictions are considered acceptable if the daily Nash-Sutcliffe Efficiency (NSE) is greater than 0.50, indicating
that the model’s predictive power is more accurate than the mean of the observed data. Additionally, the coefficient of determination
(R2) should exceed 0.60, reflecting that the model explains at least 60 % of the variance in the observed data. Moreover, the Percent
Bias (PBIAS) should be within ±15 %, which means the simulated flow values should not deviate more than 15 % from the observed
values, either positively or negatively. Meeting these benchmarks ensures that the model’s predictions are reliable and can be
confidently used for hydrological assessments. Each of these metrics was applied to assess the model’s performance in simulating the
hydrological processes within the Yamuna River basin, ensuring the model’s robustness and reliability in predicting surface runoff and
discharge.
3.2. The Hydrodynamic Simulator (iRIC Model)
The International River Interface Cooperative (iRIC) software application offers an integrated river simulation interface. The iRIC
model offers a rich and unified interface where the users are able to compile the data essential for the river analysis, and subsequently,
the river networks could be easily replicated, and analytic outcomes from the model simulation could be visualized (Kumar et al.,
2021; Nelson et al., 2015). The critical aspect of the model is that the solvers were kept isolated from the interface. Thus, iRIC acts only
as a platform for different solvers that interact with the interface and instruct it in the model’s configuration. In the present study, a
Nays2D flood solver is used to simulate the unsteady 2-d flow of the river. Nays2D flood is a solver for analyzing flood flow that uses
boundary-fitted coordinates to simulate 2-dimensional unsteady flow (Nelson et al., 2015, 2010). The solver allows the user to have
multiple inflow points that enter from the river’s upstream or sides. The solver can be employed to analyze the flood flows of primitive
rivers since it does not need any data regarding the river channel. Apart from the inundation map, iRIC also provides a velocity profile
and streamflow path. The streamflow path, combined with the velocity profile, offers an insight into the flood propagation path.
French (1994) illustrated that the flowing water’s depth and its velocity have a maximum impact on embankments’ hydrodynamic
forces. Evaluating the flow path could help predict the critical places where the embankment should be placed and the probable lo
cations of cutoffs. Thus, providing an insight into the critical sites beforehand.
The iRIC model offers several advantages that make it a preferred choice for hydrodynamic studies. Firstly, its flexibility and
versatility allow it to support a wide range of applications, from simple 1D river flow simulations to complex 2D and quasi-3D
floodplain analyses, making it suitable for diverse hydrodynamic studies (Nelson et al., 2015, 2010). Secondly, being open-source,
iRIC eliminates the cost barriers associated with proprietary software, making advanced hydrodynamic modelling accessible to a
broader range of users, including academic institutions and resource-limited agencies (Nelson et al., 2015, 2010). Furthermore, the
open-source nature of iRIC fosters a collaborative development environment where users can contribute to and benefit from
continuous improvements and shared experiences within the community. Additionally, the intuitive graphical user interface (GUI) and
comprehensive documentation make iRIC user-friendly, reducing the time and effort required to learn and implement complex hy
drodynamic models (Nelson et al., 2015, 2010). Finally, iRIC includes advanced visualization tools that help in effectively analyzing
and presenting simulation results, enhancing the understanding and communication of complex hydrodynamic phenomena (Nelson
et al., 2015, 2010).
The necessary process involving the iRIC simulations, solver and input data, and the interface for each of the processes involved is
8
shown in Fig. 4 (Nelson et al., 2015, 2010);

The basic equations of continuity and momentum of 2-D unsteady flow in a rectangular coordinate system incorporated in the iRIC
model are expressed as:
Equation of Continuity
∂h ∂(hu) ∂(hv)
+ + = 0 (6)
∂t ∂x ∂y
Equation of Motion
[ ] [ ]
∂(hu) ∂(hu2 ) ∂(huv) ∂H τbx ∂ ∂(hu) ∂ ∂(hu)
+ + = − gh + + ν + ν (7)
∂t ∂x ∂y ∂x ρ ∂x ∂x ∂y ∂y
[ ] [ ]
∂(hv) ∂(huv) ∂(hv2 ) ∂H τby ∂ ∂(hv) ∂ ∂(hv)
+ + = − gh + + ν + ν (8)
∂t ∂x ∂y ∂y ρ ∂x ∂x ∂y ∂y
where, h = depth of water; u, v = depth-averaged component of velocity; τb = bed shear stress; ρ = density of water; H = stage height (H
= h + zb); zb = elevation of the bed; ν = eddy viscosity; t = time; and x, y = spatial coordinate components in the Cartesian system.
Bed shear stress components are expressed as follows:
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
τbx = ρCf u u2 + v2 (9)
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
τby = ρCf v u2 + v2 (10)
κ
v= u × h (11)
6
The input data required for the iRIC are DEMs, satellite images and runoff hydrographs at the inlet points (Dhanapala et al., 2022).
The Geospatial Delhi Limited (GSDL) 5 m DEM was used for modeling, and the necessary satellite images were obtained from Google
Earth. Afterwards, a design grid was created by defining a grid center line representing the river flow. The values for NI, NJ and W were
outlined, where NI = number of divisions in the longitudinal direction, NJ = number of divisions in the transverse division and W =
grid width in the transverse direction. Consequently, the calculation conditions were determined, which include boundary conditions,
initial conditions, roughness parameters and appropriate time steps for the calculations. The boundary conditions for the iRIC model
were obtained from the discharge data simulated by the SWAT model. Specifically, the SWAT model was used to generate the discharge
values at various points within the watershed. These discharge values were then employed as input boundary conditions for the iRIC
hydrodynamic model. This approach ensured that the hydrological processes simulated by the SWAT model, such as runoff generation
and flow routing, were accurately represented in the subsequent hydrodynamic simulations performed by iRIC. Whereas Roughness
parameters were determined through a meticulous calibration process that utilised historical flood event data and land use infor
mation. These parameters are crucial as they represent the surface characteristics influencing water flow resistance. The calibration
involved systematically adjusting the roughness coefficients until the model outputs closely matched the observed data. This iterative
Fig. 4. Flowchart of iRIC Simulation Procedure (Kumar et al., 2021).
9
Fig. 5. Data used in setting up of SWAT model for the Yamuna River basin.
process ensures that the model’s predictions are as accurate as possible, reflecting real-world conditions effectively. In the present
study, the iRIC model is calibrated and validated against observed streamflow at five-gauge stations: Jagatpur Bund; Wazirabad
Barrage; Old Railway Bridge; Yamuna Barrage, and Okhla Gauge stations, as shown in Fig. 1. The time step was selected to be suf
ficiently small to effectively capture the dynamics of fast-flowing water during flood events, while also ensuring computational
10
efficiency. This precise selection allows the model to accurately simulate the temporal evolution of water levels and flow velocities,
providing a detailed and reliable representation of flood dynamics. The integration of SWAT-simulated discharge as boundary con
ditions in iRIC facilitated a seamless coupling between the hydrological and hydrodynamic modelling components, thereby enhancing
the overall accuracy and reliability of the flood simulation outcomes.
3.3. Input data collection and processing of digital elevation model data
The model inputs used in this simulation study to set up the SWAT model include the Digital Elevation Model (DEM), a land-cover
map, a soil map, and hydro-meteorological data. For the discretization of the catchment area based on the DEM data and for the
generation and pre-processing of the input files, the ARC GIS interface of the SWAT model was employed. The elaborate descriptions of
the input information employed to set up the SWAT model are listed in Table 1.
The gridded daily weather data from India Meteorological Department (IMD) (rainfall and temperature) were employed to
generate meteorological input data for the model. IMD gridded daily precipitation at a resolution of 0.25◦ (~ 24 km × 26 km) (Fig. 3)
and maximum/minimum temperature at 0.5◦ (~ 48 km × 52 km) resolution has been used as hydro-meteorological data for the
model. In the absence of other daily weather data on relative humidity, solar radiation, and wind speed, long-term statistics were
adopted to produce a time series for the basin’s weather parameters.
For calibration (1990–1999) and validation (2000–2005) of the developed SWAT model, streamflow datasets from the Central
Water Commission (CWC), India, at different discharge locations for the period 1990–2005, are used. In the present study, the SWAT
model is calibrated and validated against observed streamflow at seven-gauge stations: Tuini, Kalanaur, Karnal, Mawi, Delhi Railway
Bridge, Mathura and Agra, as shown in (Fig. 6).
Furthermore, it should be noted that the use of high-quality data, such as the resolution of topographic data, does affect the hy
drodynamic model (Amanambu et al., 2024; Costabile et al., 2023; Yalcin, 2020). It has been determined that even a slight alteration in
the topographic data may have pronounced ramifications on the local hydraulics (Alipour et al., 2022; Marks and Bates, 2000).
Therefore, 5 m × 5 m resolution DEM sheets acquired from Geospatial Delhi Limited (GSDL) were used to create the topography of the
Delhi region for the hydrodynamic simulation. However, the original DEM obtained from the GSDL contained only the Delhi region
data and no data for the stretch parts that fall in the adjacent Uttar Pradesh region. Hence, the DEM needs some processing to fill the
missing area. For filling the voids, SRTM DEM resolution of 30 m × 30 m was first converted to DEM of resolution 5 m × 5 m and
merged to generate a new DEM (Franks and Rengarajan, 2023). In the present study, the iRIC model is calibrated and validated against
observed streamflow at five-gauge stations: Jagatpur Bund; Wazirabad Barrage; Old Railway Bridge; Yamuna Barrage, and Okhla
Gauge stations, as shown in (Fig. 1).
3.4. Flood frequency analysis
Flood Frequency Analysis is an effective technique to carry out risk analysis and is extensively employed to estimate extreme runoff
events which have a significantly low probability of occurrence. Thus, the flood frequency curve essentially gives the relation between
flood magnitude and return period (Carolan, 2007; Karmakar and Simonovic, 2008; Villarini et al., 2009).
The flood frequency analysis of the rainfall (1951–2013) and the simulated runoff from the SWAT (1963–2011) model was used to
attain the specific discharge conforming to distinctive return periods and subsequently corresponding to the subsequent floods’
conditions, the water surface profile was estimated for each of those extreme events. At first, the series of annual maximum rainfall
depths for the considered study period were probed for the possible presence of outliers, i.e., for the rainfall data that do not share
similar characteristics as compared to the observed characteristics of the rest of the rainfall values. However, the analysis does not
indicate the existence of outliers in the available rainfall time series. Following this, a comprehensive frequency analysis of the pre
cipitation time series data was conducted by fitting different statistical distributions, namely Extreme Value Type 1, Gumbel Distri
bution, Weibull Distribution, Generalized Extreme Value Distribution, Gamma distribution, Log-Pearson Type III distribution, and 2
and 3 parameters Log-normal distribution (Acreman and Sinclair, 1986; Ahmad et al., 1988; Rulfová et al., 2016; Strupczewski et al.,
2011).
The distributions have been fitted to the data using parameter estimation techniques, including the Method of Moments, Maximum
Likelihood, Probability Weighted Moments (PWM), and PWM-based L-Moments (Cunnane, 1988; Ul Hassan et al., 2019). As the
consideration of only 1-day rainfall or the peak flow could lead to a flood event of smaller magnitude, hence in this research, 5-day
cumulative rainfall and flood hydrograph for 5-days has been considered too. An elaborate frequency analysis of annual maximum
gridded rainfall corresponding to integer durations ranging from 1-day to 5-days suggested that EV1 (Gumbel’s Extreme Value Type 1)
distribution provides the best fit overall with reasonably small standard errors. Accordingly, the study exhibits that the EV1
Table 1
Input data for SWAT application to Yamuna River.
Data Spatial Resolution Source
Topography 30 m × 30 m Shuttle Radar Topography Mission (SRTM) (http://www2.jpl.nasa.gov/ srtm/)

5m×5m Geospatial Delhi Limited (GSDL)
Soil map 1:5000,000 scale Food and Agriculture Organization (FAO) (http://www.fao.org/nr/land/soils/digital-soil-map-of-the-world/en/)
Land-cover map 100 × 100 m ORNL DAAC (https://doi.org/10.3334/ORNLDAAC/1336)
11
Fig. 6. Values of different performance indicators: (a) NSE. (b) R2, (c) PBIAS, and (d) RSR indicate the SWAT model’s performance in simulating the
runoff at different gauge stations in the Yamuna River basin.
distribution best described the statistical properties of the observed extreme rainfall events for the study region and was adopted to
quantify the design storm event. Flood flows for 10, 20, 25, and 30-year return periods were computed (Table 2) and used as inflow for
hydrodynamic modelling.
12
Table 2
Rainfalls and Flood Flows of Yamuna corresponding to different Return periods.
Return Period Rainfall (5 Day) Flood Flow Year
1 in 10 years 156.67 mm 5497 m3/s 2009

1 in 20 years 173.4 mm 6594 m3/s 1988
1 in 25 years 179.35 mm 6942 m3/s 1995
1 in 30 years 180.97 mm 7571 m3/s 1978
3.5. Model conceptualization and setup
Precipitation data is a crucial element for hydrological modelling as it drives surface runoff and the overall mass balance of hy
drological systems. The spatial variability and accuracy of precipitation input are vital for the reliable application of hydrological
models (Cao et al., 2006). In mountainous basins like the Yamuna River basin, rainfall and temperature vary significantly with changes
in elevation (Rahman et al., 2013; Tahir et al., 2011). Typically, as elevation increases, temperature decreases, significantly influenced
by topographic shading and aspect. Concurrently, precipitation generally increases with elevation, leading to a higher proportion of
snowfall compared to rainfall (Abbaspour et al., 2007; Lu et al., 2010; Thanapakpawin et al., 2007). Due to the lack of precipitation
data for higher elevations, elevation corrections were applied to rainfall and temperature data from lower elevation stations to
simulate snow hydrology accurately. The SWAT model can adjust rainfall and temperature according to elevation changes using
elevation bands (Gassman et al., 2007; Narula and Gosain, 2013). For the Yamuna River basin, with elevations ranging from 2000 m to
8000 m, the area was divided into ten elevation bands to effectively account for variations in precipitation and temperature. A
temperature lapse rate ranging from − 3 to − 6.5◦ C/km and a precipitation lapse rate between 100 and 250 mm/km were applied based
on rigorous calibration for the sub-basins incorporating elevation bands.
The simulated surface runoff from the SWAT model and water level data at various gauge stations was taken as a boundary
condition to set up the iRIC model. Furthermore, the roughness parameters were adjusted to calibrate the hydrodynamic model so that
the model simulated results and the observed depth values are as close as possible (Dhanapala et al., 2022; Hosseiny et al., 2020;
Kumar et al., 2021). The determination of the channel bed roughness for the Yamuna River was based on Manning’s roughness co
efficient, which was meticulously calibrated through an iterative hit-and-trial method. This process involved adjusting the roughness
coefficient incrementally until a satisfactory agreement was achieved between the observed and simulated water levels at all gauging
stations along the river. This rigorous calibration ensured that the model accurately represented the river’s flow dynamics. A uniform
Manning’s roughness coefficient of 0.04 was ultimately adopted for the entire river. This value was chosen based on its ability to
consistently match observed data, providing a reliable representation of the hydraulic resistance within the Yamuna River’s channel
and floodplains. In addition, comparisons were made between the results of the IRIC model and the observational data using some
statistical metrics frequently used in hydrological research, such as Nash - Sutcliffe efficiency (NSE) and coefficient of determination
(R2) (Moriasi et al., 2007; Moriasi et al., 2015).
The model simulation and application of iRIC were performed for three scenarios (Table 3); (i) for the present scenario with
embankments, (ii) the flood plain without embankments, and (iii) flood plain with maximum encroachment. The first scenario is the
currently prevailing scenario, developed with a predefined embankment coinciding with the designated O-zone boundary which
comprises the entire floodplain along the stretch of the Yamuna River that falls in Delhi, from Palla to Jaitpur. In the second scenario,
the propagation of flood flows down the Yamuna course, and its movement across its flood plain was simulated with a model setup that
assumed the study reach to be in its pristine state. This simulation, therefore, ignores the presence of the present engineered infra
structure and embankments. The third scenario is a hypothetical scenario wherein the model setup comprises some probable and
realistic encroachments of the Yamuna floodplain that would cause the river’s waterway to shrink further, leading to a possible higher
afflux in water surface profiles. With the Yamuna flood plain being considered as an ’up for grab,’ Scenario 3 is, not in the least, an
unrealistic projection and is very desirable to demonstrate to the authorities the implications of such an attempt. Accordingly, in the
third scenario, Yamuna’s right of way was further restricted and limited to its full bank width, thus disconnecting the main waterway
from its floodplain altogether, as shown in Fig. 1.
With continuous stress on the flood plain because of urbanization and human activities, a hypothetical scenario was generated
where the river Yamuna flood plain of River Yamuna gets further encroached, thus limiting the river to its channel only. This scenario
was necessary in the sense that it depicts the severity of the flood event in case of further human encroachment of the flood plain.The
scenarios mentioned above were simulated for floods with return periods of once in 10, 20, 25, and 30 years obtained through SWAT-
based modelling.
Table 3
Details of Scenarios.
Sl. No. Scenario Details
1 Scenario 1 Present Scenario with all the embankments, residential areas within the floodplain, and structures
2 Scenario 2 No action or no anthropogenic influences
3 Scenario 3 Future scenario with further encroachment inside the floodplain, where the river is limited only to the river waterway
13
4. Results
The subsequent section explains the performance of the SWAT and the iRIC models in simulating the surface runoff of the Yamuna
River and in simulating the flood dynamics for different return periods, respectively. Moreover, a comparative analysis of flood depth
for different scenarios has also been done, such as for the present scenario without embankment scenario and for a more encroached
scenario.
4.1. Calibration and sensitivity analysis of the SWAT model for the Yamuna River
The preliminary setup of the SWAT model could not precisely compute the surface runoff at each gauge station, as the peak
discharge and base flow were underestimated. Therefore, it has become imperative to calibrate the parameters to simulate the surface
runoff properly. Model calibration is defined as a process that determines that a model can accurately simulate prevailing local
conditions (Refsgaard, 1997), while validation involves simulating the model with the same set of parameters and comparing the
models’ outputs with the observed data. In this study, a manual calibration approach was adopted to calibrate the developed SWAT
model for the Yamuna River Basin (Anand et al., 2018c; Mengistu et al., 2019; Neitsch et al., 2002; Sowah et al., 2020). The surface
runoff data for the Yamuna River was available at only a daily timescale; hence the model was calibrated at a daily timescale only.
Manual calibration is often tedious, time-consuming, and requires considerable experience (Muleta and Nicklow, 2005). Manual
calibration is an iterative approach by manually changing the model’s parameters until the best result is obtained (Arnold et al., 2012;
Bailey et al., 2022; Mengistu et al., 2019). This process demands a comprehensive understanding of the model’s response to parameter
changes and the nature of parameter interactions within the study area. Conversely, auto-calibration is more efficient and objective
(Anand et al., 2017, 2018b; Shawul et al., 2013). While auto-calibration primarily relies on quantitative comparisons, manual cali
bration integrates both quantitative and qualitative evaluations (Liew et al., 2005). Anand et al., (2018c) recommend leveraging the
modeler’s knowledge and experience through manual calibration. Despite its exhaustive nature, manual calibration offers a reliable
Fig. 7. Comparison of observed and simulated water levels for Scenario 1 at (a) Jagatpur Bund; (b) Wazirabad Barrage; (c) Old Railway Bridge; (d)
Yamuna Barrage, and (e) Okhla Gauge stations for the calibration scenario.
14
means to align observed and simulated values by systematically controlling mass balance and examining surface runoff using
hydrographs (Liew et al., 2005). Hence, although SWAT provides an auto-calibration option, a manual approach was chosen for this
study. This method enables site-specific calibration, allowing the SWAT model to closely replicate the hydrological processes of the
study area, thereby establishing a robust framework for future research.
The comparison of observed and simulated streamflow for the calibration period (1990–1999) at a monthly scale exhibited
Fig. 8. Contours of Inundation Depths Corresponding to Scenario 2 (Without Embankment in North Delhi for 25-Year Return Flood (1995).
15
excellent match at all the discharge locations (Fig. 6). Once the model had been calibrated, it was further simulated for validation for
the duration of 2000–2004. Fig. 6 demonstrates a synoptical illustration of the performance of the SWAT model at each of the
discharge locations adopted for the calibration and validation of the basin in terms of different statistical performance measures such
as Nash-Sutcliffe efficiency (NSE), coefficient of determination (R2), the ratio of the root mean square error to the standard deviation of
Fig. 9. Contours of Inundation Depths Corresponding to Scenario 2 (Without Embankment in South Delhi for 25-Year Return Flood (1995).
16
measured data (RSR), and percent bias (PBIAS).

The statistical performance measures to portray the performance of the developed model fall in a very good category with NSE >
0.75 and R2 > 0.85 at all the locations (Fig. 6). The examination of the performance measures exhibits that the SWAT model has been
able to capture the intrinsic characteristics of the Yamuna River basin quite precisely in terms of surface runoff.
4.2. Calibration of the iRIC model
Flood hydrographs with peak flow corresponding to different return periods of 10, 20, 25, and 30 years were considered for the
hydrodynamic modeling to visualize the effect of the selected extreme events. The iRIC model was used to simulate the Yamuna River
floodplain for the Delhi region, from Palla Village upstream to Jaitpur downstream (Fig. 1). The simulation was concentrated on the
flow of the water and upon the propagation of the flood water by considering flow conditions of different return periods. The model
was calibrated to accurately estimate the input parameters based on given local conditions, comparing the model output with the
observed data, thereby reducing the uncertainty associated with the prediction of the model outputs.
The observed and simulated flow depths at different gauge stations with a return period of 25 years (1995) were compared (Fig. 7).
The high values of NSE (0.95) and R2 (0.96) confirm that a satisfactory agreement was reached between the predicted and observed
models for flood depth. The model was then validated for the peak flows of 10-year (1997), 20-year (1988), and 30-year (1978) return
periods. From the validated sets, it can be deduced that the model has a strong predictive ability since all the events exhibit values of
NSE and R2 in the range of 0.94–0.96 (Fig. 7).
Validation of the flood depths at different gauge stations illustrates that the iRIC model is able to mimic the characteristics of the
Yamuna River and could be used to predict the flows of different inputs. Simulated water levels were mainly within 5–15 cm of the
observed value at all the locations, and about 60 % of assessments had a difference within 5 cm. The calibrated model is then employed
to simulate the inundation and areal extent for different flood flow conditions of different scenarios using the iRIC model.
A decline in river width prompts a substantial uprise in water depths and levels, while an inverse effect was observed when
Manning’s roughness coefficient was altered. As suggested in different literature works, Manning’s number of calibrated models was
much higher than expected (Vijay et al., 2007). Plausible explanations for these differences include the effect of piers of bridges, which
is not a part of the present study, and the impact of these piers was reflected with the help of a much higher Manning’s number.
Fig. 10. Comparison of Extent of Flood for Different Return Periods of 10 years (1997) and 30 Years (1978).
17
4.3. Without the embankment scenario
The iRIC model was used to generate situations prevailing in the Yamuna River during its pristine state, i.e. when no anthropogenic
activities were inside the flood plain. The flooded area is expected to extend slowly westward towards Palla Village, Bhalswa Lake, and
subsequently towards the Model Town in the upstream section (Fig. 8). The flood moved towards the eastward direction in the
downstream stretch. The flood continued beyond the New Delhi Railway bridge, towards Anand Vihar and Mayur Vihar, spreading as
far as Indrapuram. Moreover, it continued towards Ring Road and Raj Ghat in the westward direction, flooding the recently con
structed Millennium Bus Depot area and is expected to stop just ahead of National Zoological Park (Fig. 9). While the area developed
inside the flood plain, namely Sonia Vihar and Wazirabad, got flooded (Figs. 8 and 9).
The comparison of simulated flooding extent in 1978 (National Institute of Disaster Management Report, 2013) agrees with the
flooding pattern experienced in that year. The model indicated flooding in the urban colonies of North Delhi, particularly Model Town,
which experienced heavy flooding during the 1978 flood (Fig. 8). Space where structures were built, namely Millennium Bus depot,
Delhi Secretariat, and Commonwealth games infrastructures, were under flood (Fig. 8).
Furthermore, the flooding pattern analysis reveals that North Delhi and Trans Yamuna areas are at the maximum risk of flooding
(Figs. 8 and 9). The simulation results depict that these areas witnessed a flood inundation of around 3–4 m during the flood of 1978
(Figd. 8 and 9).
In addition, areas under flood inundation for different return period events were also compared. Model simulations acknowledge
that the extremeness of the flood event i.e., the return period of a particular event very much influences the severity of the flood.
Furthermore, the flood events also affect the overall extent, with flood events of higher return periods affecting a more significant
portion of the area than the flood of lower return periods (Fig. 10). The increase in the extremeness of flooding events, i.e., from once in
a 10-year return period to a 30-year return period event, is expected to increase the areas at risk of floods by 12 percent. The increment
in the flood depth is visible at all the gauge stations, with an increase in the extremeness of flood events.
The comparison of flood depth for the present scenario and the one without any encroachment indicates an increment in the flood
Fig. 11. Comparison of Water Levels for Scenario 1 (With Embankment), Scenario 2 (Without Embankment), and Scenario 3 (Future Encroachment)
at (a) Jagatpur Bund; (b) Wazirabad Barrage; (c) Old Railway Bridge; (d) Yamuna Barrage and (e) Okhla.
18
depth inside the flood plain at different sections which could be because of the shrinking of the flood plain (Fig. 11). The simulation
results also indicate that the area between Wazirabad Barrage and Old Railway Bridge has witnessed a maximum increase in flood
depth, acknowledging that a large portion of the region is already lost, and the river water now has a lesser area available to flow.
It can be noticed that the flood level was lower by 1 m to 1.5 m on average in the floodplain in comparison to the present situation
(Fig. 11) because the area under flood was higher in the no embankment scenario. The results corroborate the critical role floodplains
play in attenuating a flood. Apart from the inundation map, iRIC also provides a velocity profile and streamflow path. Therefore, the
streamflow path, combined with the velocity profile, has also been plotted for the outputs provided by the iRIC model, consequently
providing insight into the flood propagation path. This evaluation could help the policymakers in outlining the critical places where
the embankment should be placed and the probable locations of cutoffs, thus providing an insight into the critical sites beforehand
(Figure S1 in supplementary material).
4.4. Further encroachment scenario
The simulation indicates that, on average, an increase of 3.0 m in flood depth was witnessed. At the same time, some gauge stations
show an increase of about 5 m compared to the present situation (Fig. 11 & Fig. 12). The simulation has revealed that the upstream
portion of the study area (North Delhi) does witness a maximum increase in flood depth. The Jagatpur Bund and Wazirabad Barrage
witnessed a rise of almost 3.5 m in flood depth. At the same time, Old Railway Bridge and Yamuna Barrage experienced an increase of
2.1 m and 1.5 m, respectively (Figs. 11 and 12). Moreover, further analysis of the scenarios manifested that now even flood events of
the return period of once in 10 years would have the same flood level, which was for a 25-year return period earlier. Thus, demon
strating the notion that the severity of floods has increased because of the floodplain’s encroachment (Figs. 11 and 12).
Another important aspect of this study is the simulated depth and the extent of inundations expected to occur following a flood
event of different return periods in the Yamuna as it courses down from Palla to Jaitpur in different scenarios. The results are compared
for all the scenarios mentioned above for the flood event of the return period of 25-years (1995) (Fig. 12 & Fig. 13). It is observed that
the maximum simulated depth of inundation is restricted to a low value of 5.20 m when there is no embankment or physical restraint
on lateral overbank flood flow movement. However, in this scenario, the area under inundation is expected to reach a high value of
637.11 km2 (Fig. 12 & Figure S2 in supplementary material). Furthermore, for the present scenario, simulated depths are higher. Still,
more importantly, in this scenario, there is a considerable reduction of over 470 km2 in the areal spread (166.56 km2of the flooded
zone) (Fig. 12 & Figure S2 in supplementary material), which is a very significant change and requires immediate attention for the
current as well as proposed urban settlements (Figs. 12 and 13).
Additionally, corresponding to Scenario 3 (further encroachment), with lateral movement restricted just to the bank’s full width,
the simulation experienced an increase in the flow depth and, at places, is seen to be more than 9 m. At the same time, as a tradeoff, the
total area under flood inundation is just 85 km2 (Fig. 12 & Figure S2 in supplementary material). Another noticeable fact is that for
Fig. 12. Comparison of Water Levels for Scenario 1 (With Embankment), Scenario 2 (Without Embankment), and Scenario 3 (Future
Encroachment).
19
Fig. 13. Scale of impacts of 25-year return period flood event corresponding to three Scenarios.
Scenario 2 (without embankment), the maximum amount of areas falls in the lower half of the simulated depth range, viz., maximum
area corresponds to a depth within the range of 0.5–1.0 m. While for the present and further encroachment scenarios, a considerably
large portion of the area lies in the higher range of simulated depths (Fig. 13 & Fig. 12).
4.5. Discussions
The rapid land consumption has put significant pressure on the remaining floodplains, particularly in the Yamuna River region
around Delhi. Over the past two decades, there has been considerable development within these floodplains. It is crucial to assess their
hydrological and hydraulic effectiveness, especially for large rivers like the Yamuna, where potential flood damage is substantial.
Given the complex flow conditions, predicting these impacts is challenging. Hydrodynamic models play a crucial role in assessing the
impact of different floodplain characteristics and flood events on flood behaviour.
The study’s findings corroborate those of Schober et al. (2020) and Kiedrzyńska et al. (2015), which highlight that changes in
watersheds and floodplains can vary significantly between different rivers. Each river has unique characteristics, making it essential to
conduct explicit and detailed investigations for structural flood protection measures. Schober et al. (2020) noted that structural al
terations in floodplain topography significantly impact flood characteristics, while Kiedrzyńska et al. (2015) emphasized that
anthropogenic activities leading to floodplain loss negatively affect hydrological parameters like flood peak reduction and flood wave
transmission. This study supports these findings by revealing that floodplain losses have generally led to increased floodwater levels.
Floodplain encroachment often involves the construction of levees and other flood control structures. While these measures can
provide localized protection, they may also foster a false sense of security, leading to increased development in vulnerable areas.
Although levees are designed to reduce the probability of flood hazards, studies have increasingly alerted people about the residual
risks of levees (Ding et al., 2023; Heine and Pinter, 2012). Over the years, prominent cases of levee breaching have been reported. Due
to reduced awareness and preparedness, such events often result in high vulnerability and unexpected socio-economic losses. Notable
examples include the New Orleans levee failures in 2005 after Hurricane Katrina (Dixon et al., 2006), the Mississippi River levee
failures during the 1927 and 2008 floods (Bernhardt et al., 2011; Daniel, 1977), and the UK levee breaches in the 1953 North Sea flood
(Boulter et al., 2013). Our study’s findings are consistent with these observations. The simulations using the iRIC model for the Delhi
region of the Yamuna River demonstrated similar risks associated with floodplain encroachment and levee construction. The results
revealed significant increases in flood depths in encroached areas, with some regions experiencing depths exceeding 9 m. This aligns
with the increased flood stages and flow velocities observed in the Aono Research Forest in Japan and the Missouri River in the USA
(Asano and Uchida, 2016; Pinter et al., 2001). In addition, when these structures fail, the resulting floods can be devastating, affecting
both the encroached areas and adjacent regions. The present study corroborates these findings, showing that encroachments into the
Yamuna River floodplain have similarly heightened flood risks. Moreover, in Scenario 1, there is a significant reduction of over
470 km2 in the areal spread, resulting in 166.56 km2 of the flooded zone. Similar findings were observed by Remo et al., (2012), where
the comparison of Scenario 2 (no levees) with Scenario 1 (current configuration) demonstrated that the removal of all existing levees in
the study area led to a substantial increase in the extent of flooding by 1220 km2. This highlights the critical role of levees in mitigating
flood extent and protecting vulnerable areas from extensive inundation.
The study highlights the use of historical data for calibrating and validating hydrological and hydrodynamic models, recognizing
that these parameters might change in the future. While SWAT-based hydrological models show excellent performance through
thorough calibration, their accuracy might vary under changing conditions. The inherent uncertainty in water level and discharge data
further affects model reliability. Given the numerous structural modifications in the Yamuna River Basin and potential future changes
due to socio-economic development, the study acknowledges its limitations. Notably, the exclusion of physical features like bridge
piers, which significantly influence flood dynamics, may lead to underestimations or overestimations of flood risks. The research
underscores the need for continuous monitoring and adaptive management of floodplains to ensure accurate flood risk assessments and
effective mitigation strategies. Future studies should incorporate physical features and consider the impacts of land-use changes and
climate variability to enhance model robustness.
20
Although the current framework has some limitations, the methodology outlined in the present study can be considered suitable for
large-scale coupled hydrologic and hydrodynamic modeling. The proposed methodology has relatively low computational re
quirements, and several studies using integrated models have shown that it can provide acceptable results, especially for large rivers
(Farooq et al., 2019; Shrestha et al., 2020).
5. Conclusions
The present study presents a comprehensive evaluation of the integrated SWAT and iRIC models to simulate hydrological and
hydrodynamic processes in urban flood scenarios. The paper extensively validates the 2-D hydrodynamic model iRIC in the Yamuna
River segment, focusing on flood risk assessment within highly residential areas of Delhi, the capital of India. By employing the SWAT
model to capture the hydrological processes and the iRIC model to simulate detailed floodplain dynamics and analyzing changes in
flood frequency (i.e., flood hazard) and their corresponding effects, the study offers a robust framework for assessing flood risks and
planning mitigation strategies. The findings underscore the importance of preserving floodplains and managing urban growth sus
tainably to mitigate the impacts of extreme flood events in densely populated areas.
The calibration and validation results from the SWAT model demonstrated exceptional performance, showcasing NSE values
exceeding 0.75 and R2 values greater than 0.85 across all gauge stations. These high NSE values indicate that the model’s predictive
power closely aligns with the observed data, suggesting that the SWAT model can reliably simulate the hydrological processes within
the Yamuna River basin. These findings collectively underscore the model’s robustness and reliability in capturing the intricate hy
drological dynamics of the basin.
The iRIC model was calibrated and validated to simulate floodplain dynamics in the Delhi region of the Yamuna River across
various return periods, specifically 10, 20, 25, and 30 years. The calibration process achieved exceptional Nash-Sutcliffe Efficiency
(NSE) values of 0.95 and coefficient of determination (R2) values of 0.96, indicating a robust agreement between observed and
simulated flood depths. This high level of accuracy confirms the model’s capability to reliably predict flood behaviours in the region.
Following the successful calibration and validation, the iRIC model was employed to simulate flood inundation under different
scenarios, including a baseline scenario without embankments and a scenario with further encroachment on the floodplain. The
natural floodplain was shown to play a crucial role in attenuating floods, leading to lower flood levels. Maintaining these natural
buffers can significantly reduce the impact of flooding, highlighting their essential role in flood management strategies. However,
encroachment into the floodplain was found to drastically increase flood depths, with some areas experiencing depths exceeding 9 m.
This scenario revealed the severe consequences of unchecked urbanization and floodplain encroachment, demonstrating that such
developments can significantly exacerbate flood severity. The results highlight the need for stringent land-use policies and floodplain
protection measures to prevent such adverse impacts.
In addition, the study identified urban areas, particularly North Delhi and Trans Yamuna, as being at maximum risk of flooding,
with inundation levels ranging from 3 m to 4 m during major flood events. The simulations emphasized that while flood embankments
can offer protection, they also create a false sense of security. In the event of embankment failure, the resulting inundation could be
catastrophic. This finding stresses the need for comprehensive flood risk assessments and the development of more resilient infra
structure and emergency response plans to mitigate potential flood impacts in urban settings.
Overall, while the study provides valuable insights into flood dynamics and management, these limitations suggest a need for
further research incorporating high-resolution data and a broader range of physical features to enhance the model’s accuracy and
applicability. The insights gained from this study provide valuable information for urban planners, policymakers, and flood man
agement authorities. Future studies should also consider the impacts of climate change and ongoing urban development to provide
more robust and comprehensive flood risk assessments. The inundation scenarios simulated with the hydrodynamic model iRIC can
serve as a viable strategy for assessing the flood and can guide the Irrigation and Flood Control Department of Delhi, the authority for
flood management of river Yamuna within the Delhi region, to develop an effective flood mitigation plan, inform infrastructure
development within floodplains, and help formulate policies to manage and reduce flood risks. By utilizing these insights, communities
can enhance their resilience to flooding, ensuring safer and more sustainable urban development in flood-prone areas.
CRediT authorship contribution statement
R. Khosa: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. A.K. Gosain: Writing – review &
editing, Visualization, Supervision, Investigation, Funding acquisition. Jatin Anand: Writing – original draft, Visualization, Valida
tion, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.
Data availability
The data that has been used is confidential.
21
Acknowledgments
The authors would like to thank the Indian Institute of Technology Delhi (India) and Geo-Spatial Delhi Limited (GSDL) for sup
porting this study.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ejrh.2024.101960.
References
Abbaspour, K.C., Rouholahnejad, E., Vaghefi, S., 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. J. Hydrol. 524, 733–752. https://doi.org/10.1016/j.jhydrol.2015.03.027.
Abbaspour, K.C., Yang, J., Maximov, I., et al., 2007. Modelling hydrology and water quality in the pre-alpine/alpine Thur watershed using SWAT. J Hydrol 333,
413–430. https://doi.org/10.1016/j.jhydrol.2006.09.014.
Acreman, M.C., Riddington, R., Booker, D.J., 2003. Hydrological impacts of floodplain restoration: a case study of the River Cherwell, UK. Hydrol. Earth Syst. Sci. 7,
75–85. https://doi.org/10.5194/hess-7-75-2003.
Acreman, M.C., Sinclair, C.D., 1986. Classification of Drainage Basins according to their Physical Characteristics; An Application for Flood Frequency Analysis in
Scotland. J. Hydrol. 84, 365–380. https://doi.org/10.1016/0022-1694(86)90134-4.
Ahmad, M.I., Sinclair, C.D., Werritty, A., 1988. Log-logistic Flood Frequency Analysis. J. Hydrol. 98, 205–224. https://doi.org/10.1016/0022-1694(88)90015-7.
Alipour, A., Jafarzadegan, K., Moradkhani, H., 2022. Global sensitivity analysis in hydrodynamic modeling and flood inundation mapping. Environ. Model. Softw.
152, 105398 https://doi.org/10.1016/j.envsoft.2022.105398.
Amanambu, A.C., Mossa, J., Chen, Y.H., Deitch, M., Alruzuq, A., 2024. Damming consequences: Quantifying the effects on channel geometry and floodplain
inundation. Catena 235, 107634. https://doi.org/10.1016/j.catena.2023.107634.
Anand, J., Devak, M., Gosain, A.K., Khosa, R., Dhanya, C.T., 2017. Spatial Extent of Future Changes in the Hydrologic Cycle Components in Ganga Basin using Ranked
CORDEX RCMs. Hydrol. Earth Syst. Sci. Discuss. 1–34. https://doi.org/10.5194/hess-2017-189.
Anand, J., Devak, M., Gosain, A.K., Khosa, R., Dhanya, C.T., 2021a. Spatio-temporal effect of climate and land-use change on water balance of the Ganga river basin.
J. Hydro-Environ. Res. 36, 50–66.
Anand, J., Gosain, A.K., Khosa, R., 2018a. Prediction of land use changes based on Land Change Modeler and attribution of changes in the water balance of Ganga
basin to land use change using the SWAT model. Sci. Total Environ. 644, 503–519. https://doi.org/10.1016/j.scitotenv.2018.07.017.
Anand, J., Gosain, A.K., Khosa, R., 2018b. Optimisation of Multipurpose Reservoir Operation by Coupling Soil and Water Assessment Tool ( SWAT) and Genetic
Algorithm for Optimal Operating Policy ( Case Study: Ganga River Basin. Sustain. https://doi.org/10.3390/su10051660.
Anand, J., Gosain, A.K., Khosa, R., 2021b. Impacts of climate and land use change on hydrodynamics and sediment transport regime of the Ganga River Basin. Reg.
Environ. Chang. 21, 79. https://doi.org/10.1007/s10113-021-01806-x.
Anand, J., Gosain, A.K., Khosa, R., Srinivasan, R., 2018c. Regional scale hydrologic modeling for prediction of water balance, analysis of trends in streamflow and
variations in streamflow: The case study of the Ganga River basin. J. Hydrol. Reg. Stud. 16, 32–53. https://doi.org/10.1016/j.ejrh.2018.02.007.
Arnold, J.G., Kiniry, J.R., Srinivasan, R., Williams, J.R., Haney, E.B., Neitsch, S.L., 2013. Soil & Water Assessment Tool: Input/output documentation. version 2012.
Texas Water Resour. Institute. TR 439, 650.
Arnold, J.G., Moriasi, D.N., Gassman, P.W., Abbaspour, K.C., White, M.J., Srinivasan, R., Santhi, C., Harmel, R.D., Griensven, A.Van, VanLiew, M.W., Kannan, N.,
Jha, M.K., 2012. Swat: Model Use, Calibration, and Validation. Am. Soc. Agric. Biol. Eng. 55, 1491–1508.
Arnold, J.G., Srinivasan, R., Muttiah, R.S., Williams, J.R., 1998. Large area hydrologic modeling and assesment Part I: Model development. JAWRA J. Am. Water
Resour. Assoc. 34, 73–89. https://doi.org/10.1111/j.1752-1688.1998.tb05961.x.
Arrault, A., Finaud-Guyot, P., Archambeau, P., Bruwier, M., Erpicum, S., Pirotton, M., Dewals, B., 2016. Hydrodynamics of long-duration urban floods: experiments
and numerical modelling. Nat. Hazards Earth Syst. Sci. Discuss. 1–29. https://doi.org/10.5194/nhess-16-1413-2016.
Asano, Y., Uchida, T., 2016. Detailed documentation of dynamic changes in flow depth and surface velocity during a large flood in a steep mountain stream. J. Hydrol.
541, 127–135. https://doi.org/10.1016/j.jhydrol.2016.04.033.
Asim, M., Nageswara Rao, K., 2021. Assessment of heavy metal pollution in Yamuna River, Delhi-NCR, using heavy metal pollution index and GIS. Environ. Monit.
Assess. 193, 103.
Bailey, R.T., Bieger, K., Flores, L., Tomer, M., 2022. Evaluating the contribution of subsurface drainage to watershed water yield using SWAT+ with groundwater
modeling. Sci. Total Environ. 802, 149962 https://doi.org/10.1016/j.scitotenv.2021.149962.
Bernhardt, M., Briaud, J.-L., Kim, D., Leclair, M., Storesund, R., Lim, S.-G., Bea, R.G., Rogers, J.D., 2011. Mississippi river levee failures: June 2008 flood. ISSMGE Int.
J. Geoengin. Case Hist. 2, 127–162.
Boulter, S., Palutikof, J., Karoly, D.J., Guitart, D., 2013. Natural Disasters and Adaptation to Climate Change. Cambridge University Press.
Campana, N.A., Tucci, C.E., 2001. Predicting floods from urban development scenarios: case study of the Dilúvio Basin, Porto Alegre, Brazil. Urban Water 3, 113–124.
https://doi.org/10.1016/S1462-0758(01)00004-8.
Cao, W., Bowden, W.B., Davie, T., Fenemor, A., 2006. Multi-variable and multi-site calibration and validation of SWAT in a large mountainous catchment with high
spatial variability. Hydrol Process 20, 1057–1073. https://doi.org/10.1002/hyp.5933.
Carolan, M.S., 2007. One step forward, two steps back: Flood management policy in the United States. Environ. Polit. 16, 36–51. https://doi.org/10.1080/
09644010601073507.
Costabile, P., Costanzo, C., Kalogiros, J., Bellos, V., 2023. Toward Street-Level Nowcasting of Flash Floods Impacts Based on HPC Hydrodynamic Modeling at the
Watershed Scale and High-Resolution Weather Radar Data. Water Resour. Res. 59 https://doi.org/10.1029/2023WR034599.
Cunnane, C., 1988. Methods and merits of regional flood frequency analysis. J. Hydrol. 100, 269–290. https://doi.org/10.1016/0022-1694(88)90188-6.
D’Angelo, C., Fiori, A., Volpi, E., 2020. Structural, dynamic and anthropic conditions that trigger the emergence of the levee effect: insight from a simplified risk-based
framework. Hydrol. Sci. J. 65, 914–927. https://doi.org/10.1080/02626667.2020.1729985.
Dalai, T.K., Krishnaswami, S., Sarin, M.M., 2002. Major ion chemistry in the headwaters of the Yamuna river system. Geochim. Cosmochim. Acta 66, 3397–3416.
https://doi.org/10.1016/S0016-7037(02)00937-7.
Daniel, P., 1977. Deep’n as it Come: the 1927 Mississippi River Flood. University of Arkansas Press.
De Paiva, R.C.D., Buarque, D.C., Collischonn, W., Bonnet, M.-P., Frappart, F., Calmant, S., Bulhões Mendes, C.A., 2013. Large-scale hydrologic and hydrodynamic
modeling of the Amazon River basin. Water Resour. Res. 49, 1226–1243. https://doi.org/10.1002/wrcr.20067.
Dhanapala, L., Gunarathna, M.H.J.P., Kumari, M.K.N., Ranagalage, M., Sakai, K., Meegastenna, T.J., 2022. Towards Coupling of 1D and 2D Models for Flood
Simulation—A Case Study of Nilwala River Basin, Sri Lanka. Hydrology 9, 1–14. https://doi.org/10.3390/hydrology9020017.
Ding, M., Lin, P., Gao, S., Wang, J., Zeng, Z., Zheng, K., Zhou, X., Yamazaki, D., Gao, Y., Liu, Y., 2023. Reversal of the levee effect towards sustainable floodplain
management. Nat. Sustain. 6, 1578–1586. https://doi.org/10.1038/s41893-023-01202-9.
22
Dixon, T., Amelung, F., Ferretti, A., Novali, F., Rocca, F., Dokka, R., Sella, G., Kim, S., Wdowinski, S., Whitman, D., 2006. Subsidence and flooding in New Orleans.
Nature 441, 587–588.
Domeneghetti, A., Carisi, F., Castellarin, A., Brath, A., 2015. Evolution of flood risk over large areas: Quantitative assessment for the Po river. J. Hydrol. 527, 809–823.
https://doi.org/10.1016/j.jhydrol.2015.05.043.
Entwistle, N.S., Heritage, G.L., Schofield, L.A., Williamson, R.J., 2019. Recent changes to floodplain character and functionality in England. Catena 174, 490–498.
https://doi.org/10.1016/j.catena.2018.11.018.
Farooq, M., Shafique, M., Khattak, M.S., 2019. Flood hazard assessment and mapping of River Swat using HEC-RAS 2D model and high-resolution 12-m TanDEM-X
DEM (WorldDEM). Nat. Hazards 97, 477–492. https://doi.org/10.1007/s11069-019-03638-9.
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). J. Hydrol. 262, 209–223. https://doi.org/10.1016/S0022-1694(02)00029-X.
Franks, S., Rengarajan, R., 2023. Evaluation of Copernicus DEM and Comparison to the DEM Used for Landsat Collection-2 Processing. Remote Sens 15. https://doi.
org/10.3390/rs15102509.
Gassman, P.W., Reyes, M.R., Green, C.H., Arnold, J.G., 2007. The soil and water assessment tool: Historical development, applications, and future research directions.
Trans Asabe 50, 1211–1250 https://doi.org/10.1.1.88.6554.
Gee, D.M., Anderson, M.G., Baird, L., 1990. Large-scale floodplain modelling. Earth Surf. Process. Landf. 15, 513–523. https://doi.org/10.1002/esp.3290150604.
Grimaldi, S., Petroselli, A., Arcangeletti, E., Nardi, F., 2013. Flood mapping in ungauged basins using fully continuous hydrologic-hydraulic modeling. J. Hydrol. 487,
Grimaldi, S., Schumann, G.J.P., Shokri, A., Walker, J.P., Pauwels, V.R.N., 2019. Challenges, Opportunities, and Pitfalls for Global Coupled Hydrologic-Hydraulic
Modeling of Floods. Water Resour. Res. 55, 5277–5300. https://doi.org/10.1029/2018WR024289.
Gripp, L., Alves, S., Assemany, P.P., Paula, A., Faria, M. De, Campos, L., 2014. Evaluation of Alternatives for Reducing Flow Peaks in an Urban Watershed, in: 6th
International Conference on Flood Management. pp. 1–11.
Grusson, Y., Sun, X., Gascoin, S., Sauvage, S., Raghavan, S., Anctil, F., Sáchez-Pérez, J.M., 2015. Assessing the capability of the SWAT model to simulate snow, snow
melt and streamflow dynamics over an alpine watershed. J. Hydrol. 531, 574–588. https://doi.org/10.1016/j.jhydrol.2015.10.070.
Heine, R.A., Pinter, N., 2012. Levee effects upon flood levels: An empirical assessment. Hydrol. Process. 26, 3225–3240. https://doi.org/10.1002/hyp.8261.
Hosseiny, H., Nazari, F., Smith, V., Nataraj, C., 2020. A Framework for Modeling Flood Depth Using a Hybrid of Hydraulics and Machine Learning. Sci. Rep. 10, 1–14.
https://doi.org/10.1038/s41598-020-65232-5.
Jodhani, K.H., Patel, D., Madhavan, N., 2022. A review on analysis of flood modelling using different numerical models. Mater. Today Proc. https://doi.org/10.1016/
j.matpr.2021.07.405.
Johnson, K.A., Wing, O.E.J., Bates, P.D., Fargione, J., Kroeger, T., Larson, W.D., Sampson, C.C., Smith, A.M., 2020. A benefit–cost analysis of floodplain land
acquisition for US flood damage reduction. Nat. Sustain. 3, 56–62. https://doi.org/10.1038/s41893-019-0437-5.
Karmakar, S., Simonovic, S.P., 2008. Bivariate flood frequency analysis: Part 1. Determination of marginals by parametric and nonparametric techniques. J. Flood
Risk Manag. 1, 190–200. https://doi.org/10.1111/j.1753-318X.2008.00022.x.
Khan, A., Bajpai, V.N., 2014. Constriction of the Yamuna River Floodplains Within Delhi Region Since 19Th Century: a Serious Concern. Int. J. Res. Eng. Technol. 03,
430–436. https://doi.org/10.15623/ijret.2014.0309068.
Khan, A., Govil, H., Taloor, A.K., Kumar, G., 2020. Identification of artificial groundwater recharge sites in parts of Yamuna River basin India based on Remote
Sensing and Geographical Information System. Groundw. Sustain. Dev. 11, 100415 https://doi.org/10.1016/j.gsd.2020.100415.
Kiedrzyńska, E., Kiedrzyński, M., Zalewski, M., 2015. Sustainable floodplain management for flood prevention and water quality improvement. Nat. Hazards 76,
955–977. https://doi.org/10.1007/s11069-014-1529-1.
Kim, J., Lee, J., Hwang, S., Kang, J., 2022. Urban flood adaptation and optimization for net-zero: Case study of Dongjak-gu, Seoul. J. Hydrol. Reg. Stud. 41, 101110
https://doi.org/10.1016/j.ejrh.2022.101110.
Krause, S., Jacobs, J., Bronstert, A., 2007. Modelling the impacts of land-use and drainage density on the water balance of a lowland-floodplain landscape in northeast
Germany. Ecol. Modell. 200, 475–492. https://doi.org/10.1016/j.ecolmodel.2006.08.015.
Kulkarni, A.T., Mohanty, J., Eldho, T.I., Rao, E.P., Mohan, B.K., 2014. A web GIS based integrated flood assessment modeling tool for coastal urban watersheds.
Comput. Geosci. 64, 7–14. https://doi.org/10.1016/j.cageo.2013.11.002.
Kumar, S., Agarwal, A., Villuri, V.G.K., Pasupuleti, S., Kumar, D., Kaushal, D.R., Gosain, A.K., Bronstert, A., Sivakumar, B., 2021. Constructed wetland management in
urban catchments for mitigating floods. Stoch. Environ. Res. Risk Assess. 35, 2105–2124. https://doi.org/10.1007/s00477-021-02004-1.
Kumar, M., Sharif, M., Ahmed, S., 2019. Flood risk management strategies for national capital territory of Delhi, India. ISH J. Hydraul. Eng. 25, 248–259. https://doi.
org/10.1080/09715010.2017.1408434.
Lai, X., Jiang, J., Liang, Q., Huang, Q., 2013. Large-scale hydrodynamic modeling of the middle Yangtze River Basin with complex river–lake interactions. J. Hydrol.
492, 228–243. https://doi.org/10.1016/j.jhydrol.2013.03.049.
Liew, M.W.Van, Arnold, J.G., Bosch, D.D., 2005. Problems and potential of autocalibrating a hydrologic model. Trans. ASAE 48, 1025–1040. https://doi.org/
10.13031/2013.18514.
Lin, E., Shaad, K., Girot, C., 2016. Developing river rehabilitation scenarios by integrating landscape and hydrodynamic modeling for the Ciliwung River in Jakarta,
Indonesia. Sustain. Cities Soc. 20, 180–198. https://doi.org/10.1016/j.scs.2015.09.011.
Lu, E., Takle, E.S., Manoj, J., 2010. The Relationships between Climatic and Hydrological Changes in the Upper Mississippi River Basin: A SWAT and Multi-GCM
Study. J Hydrometeorol 11, 437–451. https://doi.org/10.1175/2009JHM1150.1.
Marks, K., Bates, P., 2000. Integration of high-resolution topographic data with floodplain flow models. Hydrol. Process. 14, 2109–2122 https://doi.org/10.1002/
1099-1085(20000815/30)14:11/12<2109::AID-HYP58>3.0.CO;2-1.
Mehra, A., Farago, M.E., Banerjee, D.K., 2000. A Study of Eichhornia crassipes Growing in the Overbank and Floodplain Soils of the River Yamuna in Delhi, India.
Environ. Monit. Assess. 60, 25–45. https://doi.org/10.1023/A:1006181516828.
Mengistu, A.G., van Rensburg, L.D., Woyessa, Y.E., 2019. Techniques for calibration and validation of SWAT model in data scarce arid and semi-arid catchments in
South Africa. J. Hydrol. Reg. Stud. 25, 100621 https://doi.org/10.1016/j.ejrh.2019.100621.
Mishra, V., Shah, H.L., 2018. Hydroclimatological Perspective of the Kerala Flood of 2018. J. Geol. Soc. India 92, 645–650. https://doi.org/10.1007/s12594-018-
1079-3.
Morales-Hernández, M., Petaccia, G., Brufau, P., García-Navarro, P., 2016. Conservative 1D–2D coupled numerical strategies applied to river flooding: The Tiber
(Rome). Appl. Math. Model. 40, 2087–2105. https://doi.org/10.1016/j.apm.2015.08.016.
Moriasi, D., Arnold, J., Liew, M.W.Van, Bingner, R.L., Harmel, R.D., Veith, T.L., 2007. Model evaluation guidelines for systematic quantification of accuracy in
watershed simulations. Am. Soc. Agric. Biol. Eng. 50, 885–900. https://doi.org/10.13031/2013.23153.
Moriasi, D.N., Gitau, M.W., Pai, N., Daggupati, P., 2015. Hydrologic and water quality models: Performance measures and evaluation criteria. Trans. ASABE 58,
1763–1785. https://doi.org/10.13031/trans.58.10715.
Morid, R., Shimatani, Y., Sato, T., 2020. An integrated framework for prediction of climate change impact on habitat suitability of a river in terms of water
temperature, hydrological and hydraulic parameters. J. Hydrol. 587, 124936 https://doi.org/10.1016/j.jhydrol.2020.124936.
Muleta, M.K., Nicklow, J.W., 2005. Sensitivity and uncertainty analysis coupled with automatic calibration for a distributed watershed model. J. Hydrol. 306,
Narula, K.K., Gosain, A.K., 2013. Modeling hydrology, groundwater recharge and non-point nitrate loadings in the Himalayan Upper Yamuna basin, 468-469 Su,
S102-16 Sci. Total Environ.. https://doi.org/10.1016/j.scitotenv.2013.01.022.
National Institute of Disaster Management Report, 2013. National disaster risk reduction policy approved, Business Recorder.
Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Srinivasan, R., Williams, J.R., 2002. Soil and Water Assessment Tool User’s Manual, GSWRL report.
23
Neitsch, S., Arnold, J., Kiniry, J., Williams, J., 2011. Soil & Water Assessment Tool Theoretical Documentation Version 2009, Texas Water Resources Institute. https://
doi.org/10.1016/j.scitotenv.2015.11.063.
Nelson, J.M., Shimizu, Y., Abe, T., Asahi, K., Gamou, M., Inoue, T., Iwasaki, T., Kakinuma, T., Kawamura, S., Kimura, I., Kyuka, T., McDonald, R.R., Nabi, M.,
Nakatsugawa, M., Simões, F.R., Takebayashi, H., Watanabe, Y., Simoes, F.J., Takebayashi, H., Watanabe, Y., 2015. The international river interface cooperative:
Public domain flow and morphodynamics software for education and applications. Adv. Water Resour. 93, 62–74. https://doi.org/10.1016/j.
advwatres.2015.09.017.
Nelson, J.M., Shimizu, Y., Takebayashi, H., McDonald, R.R., 2010. The international river interface cooperative (iRIC)- Public domain software for river modeling, in:
Joint Federal Interagency Conference, 2nd. p. 8.
Nguyen, P., Thorstensen, A., Sorooshian, S., Hsu, K., AghaKouchak, A., Sanders, B., Koren, V., Cui, Z., Smith, M., 2016. A high resolution coupled
hydrologic–hydraulic model (HiResFlood-UCI) for flash flood modeling. J. Hydrol. 541, 401–420. https://doi.org/10.1016/j.jhydrol.2015.10.047.
Nielsen, S.A., Refsgaard, J.C., Mathur, V.K., 1991. Conceptual Modelling of Water Loss on Flood Plains and its Application to River Yamuna Upstream of Delhi.
Hydrol. Res. 22, 265–274.
Paz, A.R., Bravo, J.M., Allasia, D., Collischonn, W., Tucci, C.E.M., 2010. Large-Scale Hydrodynamic Modeling of a Complex River Network and Floodplains. J. Hydrol.
Eng. 15, 152–165. https://doi.org/10.1061/(ASCE)HE.1943-5584.0000162.
Pinter, N., Thomas, R., Wlosinski, J.H., 2001. Assessing flood hazard on dynamic rivers. Eos, Trans. Am. Geophys. Union 82, 333–339.
Pontes, P.R.M., Fan, F.M., Fleischmann, A.S., de Paiva, R.C.D., Buarque, D.C., Siqueira, V.A., Jardim, P.F., Sorribas, M.V., Collischonn, W., 2017. MGB-IPH model for
hydrological and hydraulic simulation of large floodplain river systems coupled with open source GIS. Environ. Model. Softw. 94, 1–20. https://doi.org/10.1016/
j.envsoft.2017.03.029.
Qi, J., Li, S., Li, Q., Xing, Z., Bourque, C.P.A., Meng, F.R., 2016. Assessing an Enhanced Version of SWAT on Water Quantity and Quality Simulation in Regions with
Seasonal Snow Cover. Water Resour. Manag. 30, 5021–5037. https://doi.org/10.1007/s11269-016-1466-8.
Rahman, K., Maringanti, C., Beniston, M., Widmer, F., Abbaspour, K., Lehmann, A., 2013. Streamflow Modeling in a Highly Managed Mountainous Glacier Watershed
Using SWAT: The Upper Rhone River Watershed Case in Switzerland. Water Resour. Manag. 27, 323–339. https://doi.org/10.1007/s11269-012-0188-9.
Rai, P.K., Dhanya, C.T., Chahar, B.R., 2018. Coupling of 1D models (SWAT and SWMM) with 2D model (iRIC) for mapping inundation in Brahmani and Baitarani river
delta. Nat. Hazards 92, 1821–1840. https://doi.org/10.1007/s11069-018-3281-4.
Rai, R.K., Upadhyay, A., Ojha, C.S.P., Singh, V.P., Rai, R.K., Upadhyay, A., Ojha, C.S.P., Singh, V.P., 2012. Salient Features of the Yamuna River Basin. Yamuna River
Basin Water Resour. Environ. 13–25.
Rajib, A., Zheng, Q., Golden, H.E., Wu, Q., Lane, C.R., Christensen, J.R., Morrison, R.R., Annis, A., Nardi, F., 2021. The changing face of floodplains in the Mississippi
River Basin detected by a 60-year land use change dataset. Sci. Data 8, 1–11. https://doi.org/10.1038/s41597-021-01048-w.
Refsgaard, J.C., 1997. Parameterisation, calibration and validation of distributed hydrological models. J. Hydrol. 198, 69–97. https://doi.org/10.1016/S0022-1694
(96)03329-X.
Remo, J.W.F., Carlson, M., Pinter, N., 2012. Hydraulic and flood-loss modeling of levee, floodplain, and river management strategies, Middle Mississippi River, USA.
Nat. Hazards 61, 551–575. https://doi.org/10.1007/s11069-011-9938-x.
Remondi, F., Burlando, P., Vollmer, D., 2016. Exploring the hydrological impact of increasing urbanisation on a tropical river catchment of the metropolitan Jakarta,
Indonesia. Sustain. Cities Soc. 20, 210–221. https://doi.org/10.1016/j.scs.2015.10.001.
Rulfová, Z., Buishand, A., Roth, M., Kyselý, J., 2016. A two-component generalized extreme value distribution for precipitation frequency analysis. J. Hydrol. 534,
Schober, B., Hauer, C., Habersack, H., 2014. A novel assessment of the role of Danube floodplains in flood hazard reduction (FEM method). Nat. Hazards 75, 33–50.
https://doi.org/10.1007/s11069-013-0880-y.
Schober, B., Hauer, C., Habersack, H., 2020. Floodplain losses and increasing flood risk in the context of recent historic land use changes and settlement developments:
Austrian case studies. J. Flood Risk Manag. 13, 1–18. https://doi.org/10.1111/jfr3.12610.
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. Hydrol. Earth Syst. Sci. Discuss. 10, 13955–13978. https://doi.org/10.5194/hessd-10-13955-2013.
Shrestha, S., Imbulana, N., Piman, T., Chonwattana, S., Ninsawat, S., Babur, M., 2020. Multimodelling approach to the assessment of climate change impacts on
hydrology and river morphology in the Chindwin River Basin, Myanmar. Catena 188, 104464. https://doi.org/10.1016/j.catena.2020.104464.
Singh, A.P., Ghosh, S.K., Sharma, P., 2006. Water quality management of a stretch of river Yamuna: An interactive fuzzy multi-objective approach. Water Resour.
Manag. 21, 515–532. https://doi.org/10.1007/s11269-006-9028-0.
Singh, S., Nair, S.S., Gupta, A.K., 2013. Ecosystem Services for Disaster Risk Reduction: A Case Study of Wetland in East Delhi Region. India Glob. J. Hum. Soc. Sci.
Geogr. Geo-Sci., Environ. Disaster Manag 13, 37–47.
Soomro, S. e hyde, Boota, M.W., Zwain, H.M., Soomro, G. e Z., Shi, X., Guo, J., Li, Y., Tayyab, M., Aamir Soomro, M.H.A., Hu, C., Liu, C., Wang, Y., Wahid, J.A.,
Bai, Y., Nazli, S., Yu, J., 2024. How effective is twitter (X) social media data for urban flood management? J. Hydrol. 634, 131129 https://doi.org/10.1016/j.
jhydrol.2024.131129.
Soomro, S. e hyde, Hu, C., Boota, M.W., Ahmed, Z., Chengshuai, L., Zhenyue, H., Xiang, L., Soomro, M.H.A.A., 2022. River Flood Susceptibility and Basin Maturity
Analyzed Using a Coupled Approach of Geo-morphometric Parameters and SWAT Model. Water Resour. Manag. 36, 2131–2160. https://doi.org/10.1007/
s11269-022-03127-y.
Sowah, R.A., Bradshaw, K., Snyder, B., Spidle, D., Molina, M., 2020. Evaluation of the soil and water assessment tool (SWAT) for simulating E. coli concentrations at
the watershed-scale. Sci. Total Environ. 746, 140669 https://doi.org/10.1016/j.scitotenv.2020.140669.
Stevaux, J.C., Latrubesse, E.M., Hermann, M.L. de P., Aquino, S., 2009. Floods in Urban Areas of Brazil. Dev. Earth Surf. Process. 13, 245–266. https://doi.org/
10.1016/S0928-2025(08)10013-X.
Strupczewski, W.G., Kochanek, K., Bogdanowicz, E., Markiewicz, I., 2011. On seasonal approach to flood frequency modelling. Part I: Two-component distribution
revisited. Hydrol. Process. 26, 705–716. https://doi.org/10.1002/hyp.8179.
Sun, Y., Liu, C., Du, X., Yang, F., Yao, Y., Soomro, S. e hyde, Hu, C., 2022. Urban storm flood simulation using improved SWMM based on K-means clustering of
parameter samples. J. Flood Risk Manag 15, 1–15. https://doi.org/10.1111/jfr3.12826.
Sutcliffe, J.V., Parks, Y.P., 1989. Comparative water balances of selected African wetlands. Hydrol. Sci. J. 34, 49–62. https://doi.org/10.1080/02626668909491308.
Tahir, A.A., Chevallier, P., Arnaud, Y., et al., 2011. Modeling snowmelt-runoff under climate scenarios in the Hunza River basin, Karakoram Range, Northern Pakistan.
J. Hydrol. 409, 104–117. https://doi.org/10.1016/j.jhydrol.2011.08.035.
Tang, Y., Sun, Y., Han, Z., Soomro, S. e hyde, Wu, Q., Tan, B., Hu, C., 2023. Flood Forecasting Based on Machine Learning Pattern Recognition and Dynamic Migration
of Parameters. J. Hydrol. Reg. Stud. 47, 101406 https://doi.org/10.1016/j.ejrh.2023.101406.
Teo, F.Y., 2010. Study of the hydrodynamic processes of rivers and floodplains with obstructions. Cardiff University (United Kingdom).
Thanapakpawin, P., Richey, J., Thomas, D., et al., 2007. Effects of landuse change on the hydrologic regime of the Mae Chaem river basin, NW Thailand. J Hydrol 334,
Tobin, G.A., 1995. The levee love affair: a stormy relationship? JAWRA J. Am. Water Resour. Assoc. 31, 359–367.
Tomar, P., Singh, S.K., Kanga, S., Meraj, G., Kranjčić, N., Ðurin, B., Pattanaik, A., 2021. Gis-based urban flood risk assessment and management—a case study of delhi
national capital territory (NCT), India. Sustain 13, 1–20. https://doi.org/10.3390/su132212850.
Ul Hassan, M., Hayat, O., Noreen, Z., 2019. Selecting the best probability distribution for at-site flood frequency analysis; a study of Torne River. SN Appl. Sci. 1, 1–10.
Vijay, R., Aabha, S., Gupta, A., 2007. Hydrodynamic simulation of river Yamuna for riverbed assessment: A case study of Delhi region. Environ. Monit. Assess. 130,
381–387. https://doi.org/10.1007/s10661-006-9405-4.
Villarini, G., Smith, J.A., Serinaldi, F., Bales, J., Bates, P.D., Krajewski, W.F., 2009. Flood frequency analysis for nonstationary annual peak records in an urban
drainage basin. Adv. Water Resour. 32, 1255–1266. https://doi.org/10.1016/j.advwatres.2009.05.003.
Wahlstrom, M., Guha-Sapir, D., 2015. The human cost of weather-related disasters 1995–2015. Geneva, Switz. UNISDR.
24
Wang, G., Yang, H., Wang, L., Xu, Z., Xue, B., 2014. Using the SWAT model to assess impacts of land use changes on runoff generation in headwaters. Hydrol. Process.
28, 1032–1042. https://doi.org/10.1002/hyp.9645.
White, G.F., 1945. Human adjustment to floods: Department of geography research paper No. 29. Chicago, Univ. Chicago.
Wing, O.E.J., Sampson, C.C., Bates, P.D., Quinn, N., Smith, A.M., Neal, J.C., 2019. A flood inundation forecast of Hurricane Harvey using a continental-scale 2D
hydrodynamic model. J. Hydrol. X 4. https://doi.org/10.1016/j.hydroa.2019.100039.
Wongsa, S., 2014. Simulation of Thailand Flood 2011. Int. J. Eng. Technol. 6, 452–458. https://doi.org/10.7763/IJET.2014.V6.740.
Yalcin, E., 2020. Assessing the impact of topography and land cover data resolutions on two-dimensional HEC-RAS hydrodynamic model simulations for urban flood
hazard analysis. Nat. Hazards 101, 995–1017. https://doi.org/10.1007/s11069-020-03906-z.
Zhang, W., Zhang, X., Liu, Y., Tang, W., Xu, J., Fu, Z., 2020. Assessment of flood inundation by coupled 1d/2d hydrodynamic modeling: A case study in mountainous
watersheds along the coast of southeast China. Water (Switz. ) 12, 1–14. https://doi.org/10.3390/w12030822.
Zope, P.E., Eldho, T.I., Jothiprakash, V., 2016. Impacts of land use–land cover change and urbanization on flooding: A case study of Oshiwara River Basin in Mumbai,
India. CATENA 145, 142–154. https://doi.org/10.1016/j.catena.2016.06.009.
25

case study Yamuna River Basin in Delhi

Uploaded by

Copyright:

Available Formats

case study Yamuna River Basin in Delhi

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

case study Yamuna River Basin in Delhi

Uploaded by

Copyright:

Available Formats

Journal of Hydrology: Regional Studies 56 (2024) 101960

Contents lists available at ScienceDirect

Journal of Hydrology: Regional Studies

Hydrological and hydrodynamic modelling for flood management:

Keywords: Study region: Yamuna River (Delhi), India.

Fig. 1. Index Map of Yamuna River in Delhi.

Fig. 2. Major Construction Within the Flood Plain.

3. Methodology and data description

3.1. SWAT model

3.2. The Hydrodynamic Simulator (iRIC Model)

shown in Fig. 4 (Nelson et al., 2015, 2010);

Fig. 4. Flowchart of iRIC Simulation Procedure (Kumar et al., 2021).

3.4. Flood frequency analysis

Topography 30 m × 30 m Shuttle Radar Topography Mission (SRTM) (http://www2.jpl.nasa.gov/ srtm/)

1 in 10 years 156.67 mm 5497 m3/s 2009

3.5. Model conceptualization and setup

measured data (RSR), and percent bias (PBIAS).

4.2. Calibration of the iRIC model

4.3. Without the embankment scenario

4.4. Further encroachment scenario

CRediT authorship contribution statement

Declaration of Competing Interest

The data that has been used is confidential.

Appendix A. Supporting information

You might also like