flood modelling: recent indian contributionscivil.iisc.ernet.in/~pradeep/chandra insa 2.pdf ·...
TRANSCRIPT
Review Article
Flood Modelling: Recent Indian Contributions
R CHANDRA RUPA1 and PP MUJUMDAR1,2,3,*
1Department of Civil Engineering, Indian Institute of Science, Bangalore, Karnataka 560 012, India2Interdisciplinary Centre for Water Research (ICWaR), Indian Institute of Science, Bangalore, India3Divecha Center for Climate Change, Indian Institute of Science, Bangalore, India
(Received on 03 April 2018; Revised on 03 January 2019; Accepted on 30 April 2019)
Flood modelling is an important first step towards understanding and managing floods, at all spatial scales, from a large
catchment of a river to highly developed urban areas and to individual property and infrastructure. Recent advances in
modelling techniques along with sophisticated computational tools and data products have facilitated a rapid progress in
flood modelling. Despite this progress, an accurate assessment and forecasting of floods with the associated risk is still
elusive due to uncertainties at each stage of the modelling. This paper presents a review of flood modelling with specific
focus on India. In recent years, significant contributions have been made by Indian researchers to the hydrologic science in
general, and to flood modelling in particular. Indian contributions to the areas of hydrologic modelling of floods, flood
forecasting, flash flood modelling, urban floods, and risk assessment and mitigation are reviewed in the paper. Since the
emphasis is on flood modelling, the related topics of flood frequency analysis are excluded, although some representative
studies on extreme rainfall frequencies are briefly mentioned. The paper concludes with some perspectives on future
directions in flood modelling and actions needed for sustainable flood management practices, in the country.
Keywords: Flood Modelling; Flood Forecasting; Uncertainties; Risk Assessment; Flash Floods; Urban Floods
*Author for Correspondence: E-mail: [email protected]
Proc Indian Natn Sci Acad 85 No. 4 December 2019 pp. 705-722
Printed in India. DOI: 10.16943/ptinsa/2019/49648
Introduction
Floods are the most common and third most damaging
natural hazards globally after storms and earthquakes
(Wilby and Keenan, 2012). Flooding is a major concern
in India, as indeed it is in the other parts of the world.
Consequently, a great effort has been put by Indian
research community in developing models and
methodologies to understand floods at various spatial
scales. Significant contributions have been made by
the community in recent years, in the areas of flood
estimation, multivariate hydrologic modelling,
uncertainty quantification, climate change impacts,
flood hazard and risk assessment. Large river basins
such as the Ganga, Mahanadi, Godavari and Krishna
basins have been studied by several researchers to
understand the nature of river flooding. Several studies
also focus on modelling urban floods. A main objective
of the brief review presented in this paper is to compile
information on the Indian research on flood modelling
and to provide perspectives on implementable actions
to mitigate the impacts of floods. Modelling aspects
of the floods, including hydrologic issues, data
requirements and uncertainty quantification are first
discussed, followed by modelling for climate change
impact assessment. Flood forecasting, with an
emphasis on forecasting of flash floods is discussed
next. Urban flooding, which is known to be increasing
in frequency and spatial extent in the country, is
reviewed next. The concluding remarks of the paper
provide some thoughts on future directions for flood
modelling and implementable policies.
Flood Modelling
Flood modelling and creating flood inundation maps
are essential to understand the possible impacts of
floods of a given magnitude and to initiate actions on
the ground to minimize the damages. Hydrodynamic
modelling plays an important role in obtaining the flood
characteristics, i.e., the magnitude, duration and the
spatial distribution of flooding. Progress in
hydrodynamic modelling during the last decade has
led to considerable improvements in ability to simulate
706 R Chandra Rupa and PP Mujumdar
flooding scenarios. Models may be classified
depending on how the catchment processes are
represented (deterministic or stochastic) or on how
the catchment is discretized spatially (lumped or
distributed). Routing models that estimate the flood
wave propagation along a river channel, have been
developed with the continuity, momentum and data
driven approaches (Perumal and Price, 2013;
Gopakumar and Mujumdar, 2008; Perumal and Sahoo,
2007; Mujumdar, 2001). Studies on flood frequency
analysis, based on statistical models, are in general
used for understanding the long-term changes in flood
magnitudes and frequencies (Guru and Jha, 2015;
Kamal et al., 2017; Kumar et al., 2003; Basu and
Srinivas, 2014; Santhosh and Srinivas, 2013).
Mujumdar and Nagesh Kumar (2012) discussed
various aspects of floods including the hydrological
modelling for floods, climate change impacts
assessment, remote sensing and GIS for modelling
floods, along with case studies.
Different hydrological modelling software such
as HEC-HMS, SWAT, MIKE, VIC and SWMM are
used for modelling flows in the river basins. It is
instructive to note that most hydrologic models
reproduce the average conditions very well but fail to
simulate the extreme conditions. The models also
perform badly in catchments with significant structural
interventions such as dams, unless the flows are
normalised. For example, Nandi and Reddy (2017)
used a VIC model for simulating the hydrological
variables over Krishna River Basin and concluded
that the VIC model can handle large-scale variability
but overestimates the stream flow in the downstream
portion as the effect of storage structures are not
considered in their model. Chawla and Mujumdar
(2015) used VIC model to simulate the Upper Ganga
Basin (UGB) and analysed the stream flow patterns
and magnitude of runoff. Fig. 1 shows the calibration
and validation results from VIC model. They have
also presented optimum set of parameters for the two
regions (upstream and midstream regions of UGB)
along with their performance measures during
calibration. On the other hand, hydrodynamic models
specifically tailored for flood flow simulations would
serve better for the purpose. Patro et al. (2009) used
a one-dimensional hydrodynamic model to simulate
the discharges in the Mahanadi River basin for the
monsoon period (June-September) of the year 2004.
They developed the model by integrating information
from various sources i.e., refining the cross-sections
derived from SRTM DEM along with the measured
river cross-sections, available river discharge as well
as water-level data at different gauging sites.
Efforts are currently being put to improve the
use of hydrologic models for simulating floods at river
basin scales by driving them with state of the art
numerical weather prediction (NWP) models and/or
with improved rainfall data products (e.g., Chawla
and Mujumdar, 2019). Shah and Mishra (2016)
evaluated the performance of VIC model with several
rainfall data sets to simulate the daily discharges in
the Mahanadi river basin and concluded that bias in
real-time precipitation products affects the initial
condition and precipitation forcing, which in turn
affects flood peak timing and magnitudes.
Hydrologic models, meant primarily for stream
flow simulation, are applied commonly in flood
Fig. 1: Calibration and validation results of (a) upstream
and (b) midstream regions (Source: Chawla and
Mujumdar, 2015). The figure shows that the average
conditions are in general simulated well by the model,
but the extreme flood flows are poorly simulated
Flood Modelling: Recent Indian Contributions 707
modelling studies as well, and from the applications it
is in general concluded that the models are able to
reproduce the flood flows with ‘acceptable accuracy’.
However, acceptable accuracy involves subjectivity
and to overcome this limitation, the statistical goodness-
of-fit measures are used to estimate the model
efficiency. Even the use of such measures does not
necessarily allow an objective judgment of model
performance, which depends on the specific objectives
of the model application for given a catchment and
observed data. Also, each group working on a basin
develops its own model, resulting in an excess of such
tools for each major basin. For example, the Mahanadi
basin is modelled by many researchers (Mondal and
Mujumdar, 2012; Ghosh et al., 2010; Pattanayak et
al., 2017; Asokan and Dutta 2008; Gosain et al., 2006;
Jena et al., 2014; Patro et al., 2009). Johnston and
Smakhtin (2014) discussed on how much modelling is
enough for a river basin. Their study reviewed
hydrological modelling in four large basins -Nile,
Mekong, Ganges and Indus –and suggested four areas
for action to improve effectiveness and reduce
duplication in hydrological modelling. The input data,
model details and the results, should be published on
a common platform, to allow more coordinated
approaches and benefit from past modelling
experiences. For each major basin, an appropriate
agency should be identified to take responsibility for
coordination and data sharing, to reduce redundancy
of effort and promote collaboration among
researchers. Initiatives are required to improve the
quality and quantity of data for the models (for
calibration and validation purpose) and novel
approaches are essential for data collection (remote
sensing, crowd-sourcing and community-based
observations) and data assimilation from different
sources.
Data Requirements
The efficiency and usefulness of any model depends
on the quantity and quality of the data, it is presented
with during the calibration and validation stages. The
data requirements of flood modelling for river basins
fall into three distinct categories, (i) topographic data
of the basin and the channel and/or the river, (ii)
hydrometeorological data including precipitation,
temperature, solar radiation, land use-land cover and
sub-surface hydraulic properties, and (iii) time series
of flow rates and stage data to provide model input
and output boundary conditions for model calibration
and validation. Acquisition of part of this data has
been made possible by developments in the field of
remote sensing. Remote sensing, from both satellites
and aircraft, allows the collection of spatially distributed
data over large areas and reduces the need for costly
ground survey. However, the ground measurements
complement the data from the satellites, and are useful
for validation of the satellite products. Due to the
advances in satellite remote sensing in the last few
decades, more and more geo-spatial datasets related
to hydrology - such as topography, soil and land use-
have become available through several open sources.
Durga Rao et al. (2011) developed a flood forecast
model for the Godavari Basin considering a distributed
modelling approach with space inputs. They have
computed the topographic and hydraulic parameters
using the land use/land cover grid that is derived from
the Indian Remote Sensing Satellite (IRS-P6) AWiFS
sensor data (56 m resolution), Shuttled Radar
Topographic Mission (SRTM) Digital Elevation Model
(DEM), and the soil textural grid. The model is
calibrated and validated using the field
hydrometeorological data. The model was tested
during the 2010 floods with real-time 3-hour interval
hydrometeorological and daily evapotranspiration data
and found that the accuracy in estimating the peak
flood discharge and lag time was good. Mondal et al.
(2016) have reviewed the progress in hydrological
modelling achieved in India during the last about five
years and discussed the data requirements of the
hydrological modelling. Despite the availability of
various open sources for useful data, there is still a
lack of good quality and quantity of data at required
spatial and temporal, specifically when extremes are
modelled, leading to uncertainties in the model results.
Due to resource constraints and limited ground
measurements, often the modeller needs to extrapolate
information from the available measurements in space
and time. In addition, uncertainties in the climate
models, which are used to assess the likely
hydrological impact of future system response, (for
example to climate and land management change),
further compounds the uncertainties in the projections.
It is important to quantify and communicate such
uncertainties for use in policy making.
Uncertainty Quantification
Uncertainties are inherent in any modelling process
708 R Chandra Rupa and PP Mujumdar
and originate from a wide range of sources, from model
formulation to the quality and quantity of data.
Uncertainties cannot be eliminated, but their amplitude
should be estimated andtheir sources should be
identified to understand the impact on modelling
(Deletic et al., 2012). Beven (2006) stated that there
are many sources of uncertainty that interact non-
linearly in the modelling process. However, not all
uncertainty sources can be quantified with acceptable
levels of accuracy, and the proportion of uncertainty
sources being ignored may be high in environmental
(including hydrological) modelling investigations
(Deletic et al., 2012, Harremoës, 2003; Doherty and
Welter, 2010). The uncertainties should be quantified
wherever possible and propagate them to the decisions
of risk. Specifically, in the precipitation data,
quantifying uncertainties is crucial in obtaining the risks
and resilience of the hydrologic system.
Precipitation is one of the hydrologic variables
with a potential to impart the highest amount of
uncertainty due to high variability in both space and
time. Mondal et al. (2016) reported that although,
Indian Meteorological Department (IMD) has a wide
network of rain gauges there are still data sparse
regions. Afew studies have explored using rainfall data
from satellite data such as Tropical Rainfall
Measurement Mission (TRMM) to overcome this gap
(Indu and Nagesh Kumar, 2014; Mondal et al., 2018).
Attempts have also been made to integrate the satellite
data with forecasting systems to improve the
performance of the models in predicting floods
(Sharma et al., 2017; Mitra et al., 2013). However,
there are high uncertainties associated with
precipitation, and still gaps exist in precisely quantifying
the uncertainties in literature. Quantification of
uncertainties especially in the extreme rainfall that
causes flooding can be quantified using statistical
methods (assimilating the data from various sources)
and the Bayesian techniques (e.g., Chandra Rupa et
al., 2015).
In addition to the uncertainty in precipitation,
there are several uncertainties in the other inputs to a
hydrologic model. The upstream discharge and the
roughness relationship for the main channel, for
example, havea major influence on the uncertainty in
the modelled water levels. Floodplain bathymetry, weir
formulation and discretization of floodplain topography
add most to the uncertainties in model outcomes
(Warmink et al., 2011). Yaduvanshi et al. (2018)
analysed the runoff response during extreme rain
events over the Basin of Subarnarekha River in India
using soil and water assessment tool (SWAT).They
analysed the sensitivity of basin parameters to improve
the flood runoff simulation efficiency of the model.
Climate Change Impacts
Changing climate increases the risks associated with
floods. The magnitude and the frequency and of flood
discharges are affected due to climate change, and
there is a clear indication that the changes in the
magnitude and frequency will continue in the future
due to continuous increase in the concentration of
greenhouse gasses (GHGs) in the atmosphere (IPCC,
2012). The streamflow variation is region or basin
specific and is not uniform across the world. Mujumdar
and Ghosh (2008) presented an overview of the
methodologies developed for assessing hydrologic
impacts of climate change with an emphasis on
statistical techniques for regional impact assessment
and modelling of uncertainty, resulting from the use
of multiple climate models. They have demonstrated
the methodologies with the case study of Orissa
meteorological subdivision and Mahanadi river basin,
which shows a possible decreasing trend in rainfall
and monsoon streamflow of the region in future.
Gosain et al. (2006) quantified the climate change
impacts on the water resources of Indian River basins.
Their study included an assessment of impact on
floods. They have considered more than 12 river
basins of the country, simulated 40 years (20 years
present and 20 years future) of weather data, and
concluded that under the Green House Gas (GHG)
scenario, severity of droughts and intensity of floods
in various parts of the country may get deteriorated.
Hirabayashi et al. (2013) used a state-of-the-art global
river routing model with an inundation scheme to
compute river discharges and flooded areas. An
ensemble of projections under a new high-
concentration scenario shows a large increase in flood
frequency in Southeast Asia, Peninsular India, eastern
Africa and the northern half of the Andes, with small
uncertainty in the direction of change.
Asokan and Dutta (2008) analysed the water
resources in the Mahanadi river basin under the
current and projected climate conditions and
concluded that the basin is expected to experience
Flood Modelling: Recent Indian Contributions 709
progressively increasing intensities of flood in
September and drought in April in future. Also, they
analysed the water demand estimation considering
sectors of domestic, irrigation and industry and
concluded that the future water demand shows an
increasing trend until 2050, beyond which the demand
will decrease owing to the assumed regulation of
population explosion.
Precipitation is a basic and important input to
the hydrological models for flood estimations. There
are studies showing the increases in the extreme
precipitation due to climate change (e.g., Guhathakurta
et al., 2011). Mondal and Mujumdar (2015) analysed
changes in extreme rainfall characteristics over India
using a high-resolution gridded dataset. Intensity,
duration and frequency of precipitation over a
threshold in the summer monsoon season are modelled
by non-stationary distributions whose parameters vary
with physical covariates like the El-Nino Southern
Oscillation index (ENSO-index - indicator of large-
scale natural variability), global average temperature
(indicator of human-induced global warming) and local
mean temperatures (indicator of localized changes).
Fig. 2 shows the grid-wise best statistical models for
intensity, duration and frequency of extreme rainfall.
They concluded that the intensity, duration and
frequency exhibit non-stationarity due to different
drivers and no spatially uniform pattern is observed in
the changes in them across the country.
However, an extreme precipitation event may
not be responsible for an adverse flood situation. There
are several other factors including the antecedent
moisture conditions and basin conditions. Jena et al.
(2014) analysed floods in Mahanadi basin in eastern
India and examined if the increase in flooding in the
basin is only due to increase in extreme rainfall. They
Fig. 2: Grid-wise best statistical models for (i) extreme rainfall intensity (ii) extreme rainfall duration and (iii) extreme
rainfall frequency: (a) spatial patterns (b) percentage of locations falling under each category of models (Source:
Mondal and Mujumdar, 2015)
710 R Chandra Rupa and PP Mujumdar
carried out a region based analysis of extreme
precipitation and concluded that the frequencies of
high floods in Mahanadi basin is due to an increase in
extreme precipitation in the middle reaches of the
basin.
In the context of climate change, even if the
observed past is stationary, there can be
nonstationarities in future hydrologic extremes. In such
cases, possible realizations of future can be obtained
from the climate model driven physically based
hydrologicmodel streamflow projections. For
nonstationary future realizations, the research problem
is to investigate how long the stationary historical
return levels of floods will remain valid, considering
uncertainties in the estimation of observed and
projected return levels (Mondal and Mujumdar, 2016).
In the transient case, the effective return levels (Katz
et al. 2002) are considered by holding the probability
of exceedance fixed at eachyear, because the one-
to-one relation between return period and probability
of exceedance is no longer valid as the probability of
exceedance changes from year to year. Mondal and
Mujumdar (2016) detected the changes in the flood
return levels and concluded that coherent change in
flood return level across the projections is not detected
in the Columbia River using streamflow projections.
Though there are studies reported for detection and
attribution of hydrological extremes for the Indian
region (Jain and Kumar, 2012; Sonali and Nagesh
Kumar, 2016), no explicit research is carried out for
studying the changes in the flood return levels. Met
Office Hadley Center, UK, complied robust
information on the physical impacts of climate change
for more than 20 countries, including India (http://
eprints.nottingham.ac.uk/2040/12/India.pdf, assessed
on 27.03.2018). The Center has studied the climate
change impacts in the terms of varying return periods
for various aspects including water stress, drought,
precipitation, pluvial and fluvial flooding, cyclones and
crop yields. A research report prepared by Joint Global
Change Research Institute and Battelle Memorial
Institute, Pacific Northwest Division, US, notified that
climate projections indicate several changes in India’s
future climate, given inherent uncertainties (NIC,
2009). Their report indicated that a warming of 0.5oC
is likely over all India by the year 2030, which is
approximately equal to the warming over the 20th
century. And a warming of 2-4oC by the end of this
century is expected, with the maximum increase over
northern India. Also, increase in extreme precipitation
events, both in magnitude and frequency, is likely
leading to significant flooding. These changes will
have implications on increases in flood frequencies
and intensities.
Several uncertainties abound, in addition to the
data uncertainties, when one is interested in projecting
the flood discharges using a hydrological model.
Atmosphere-Ocean Global Climate Models
(AOGCMs) are credible and reliable tools for global
scale climate analyses. Downscaling methods are used
for transferring coarse-scale climate information to
regional scale. Projections of hydro-climatic variables
using downscaling comprises several sources of
uncertainties. Uncertainties may arise from (i) the
selection of the climate model, (ii) the choice of carbon
emission scenarios, (iii) the choice of downscaling
methods, (iv) the selection of hydrological model and
model parameters and (v) the internal variability of
the climate system. Prudhomme and Davies (2009)
reported that selection of climate models createsa
higher uncertainty in the downscaling process
compared to the choice of emission scenarios or model
parameterization. However, it is also concluded that
significant source of uncertainty in hydrologic
projections is due to downscaling methods, compared
to the choice of climate models and emission
scenarios. (Bürger et al., 2012; Mandal et al., 2016).
Significant progress has been made by researchers
in developing methods for quantifying uncertainties
with respect to the Indian context (Mujumdar and
Ghosh, 2008; Ghosh et al., 2010; Raje and Mujumdar,
2010a, Raje and Mujumdar, 2010b; Mondal and
Mujumdar, 2012).
The structural difference in the uncertainties in
parameter estimation and the hydrological models can
have a significant effect on the spatial and temporal
distribution of runoff. Limited literature is available in
India, which investigates all sources of uncertainty
(including (i) input uncertainty, e.g., sampling and
measurement errors in rainfall observations; (ii) output
uncertainty, e.g., rating curve errors affecting runoff
estimates; (iii) structural uncertainty, arising from
formulation of hydrologic models; and (iv) parametric
uncertainty, reflecting the incompetence to specify
exact values of model parameters due to uncertainties
in the calibration data, model approximations, etc.) in
streamflow projections under climate change. Chawla
Flood Modelling: Recent Indian Contributions 711
and Mujumdar (2018) segregated the uncertainties in
streamflow projections arising from (i) General
Circulation Models (GCMs), (ii) emission scenarios,
(iii) land use scenarios, (iv) stationarity assumption of
the hydrologic model, and (v) internal variability of
the processes, using analysis of variance (ANOVA)
approach. They have used Variable Infiltration
Capacity (VIC) model over the Upper Ganga Basin
(UGB) and concluded that model parameters vary
with time, annulling the often-used assumption of
model stationarity. They found that the streamflow
reduces in future in the UGB under the nonstationary
model condition, and the model stationarity assumption
and GCMs along with their interactions with emission
scenarios, act as dominant sources of uncertainty.
Several researchers have studied quantifying the
uncertainties in different river basins of India (e.g.,
Yaduvanshi et al., 2018; Uniyal et al., 2015; Singh et
al., 2014; Srivastav et al. 2007). However, studies
are required to understand the sources of uncertainties
and methods to segregate them in modelling the
streamflow.
Flood Forecasting
Flood forecasting systems can be developed over a
wide range of temporal scales ranging from hours to
days, or even months, depending on the spatial scale
of interest and/or the availability of reliable weather
forecasts. The ability of weather forecasts has
improved considerably over the last 3 decades.
Improvements in weather forecasts and advances in
monitoring, remote sensing, data collection and models
have led to simultaneous improvements in the flood
forecasting ability. Depending on the availability of
the data (both hydrological and meteorological), basin
characteristics, computational facilities, lead time of
forecasts, and the purpose for which the forecast is
used, different flood-forecasting techniques are being
used in India. Some of the usually employed
techniques include: (i) simple relations based on stage-
discharge relationships, (ii) co-axial correlation
diagrams developed utilizing the stage, discharge and
rainfall data, etc., (iii) event based hydrological system
models for small to moderate-sized catchments, (iv)
network models consisting of the sub-basins and sub-
reaches for the large-sized catchments, and (v)
deterministic hydrologic models (at selected places).
In addition to the above, stochastic models have
also been applied for real-time flood forecasting. In
India, the statistical approach is most widely used to
formulate real-time flood forecasts, in addition to the
computing techniques such as ANN and fuzzy logic
(Lohani et al., 2014; Mukerji et al., 2009; Lohani et
al., 2006, Thirumalaiah and Deo, 2000). Event-based
network models are applied along with the multi-
parameter hydrological models (Rahman et al., 2012)
to some pilot projects. Singh (2008) consolidated the
Indian experiences in real-time flood forecasting
highlighting the flood problems, data requirements,
methods employed for issuing flood forecasts and
research developments in the area of real-time flood
forecasting. He reported that the flash floods are the
most severe and there is no effective system
implemented for flash-flood forecasting for arid and
semi-arid regions. Some flood-warning arrangements
exist in the country, but these are largely aimed at
transmitting limited information on flood levels from
upstream points to the areas lower down. Such
warnings have limited utility in as much as they do
not indicate the likely levels and the time of arrival of
floods at the vulnerable places. Further, they do not
often give adequate advance notice. To improve the
flood forecasting, an integrated approach combining
the real-time observations as an input to the
hydrological model and an effective communication
systems is required. Several if-then scenarios and
algorithms should be analysed and an efficient warning
system based on the analysis should be setup.
Communication plays an important role in sending the
warnings to the concerned authorities and therefore
a good communication system is required in
disseminating the flood information.
Flash Flood Forecasting
Flash floods are characterized by rapid occurrence,
with very limited opportunity for issuing warnings.
They are often accompanied by other natural hazards
(such as landslides and mud flows), causing damage
to buildings and businesses, collapse of hydraulic
infrastructure and in extreme situations, loss of life.
Recent floods in Kedarnath, Uttarakhand (during 2013
and 2015) are a classic example of flash floods in the
Mandakini River that killed thousands of people and
livestock, devastating the country. Though the duration
of the events was small compared to other flood
disasters in the country, they resulted in severe
damage to property and life because they were
712 R Chandra Rupa and PP Mujumdar
accompanied by simultaneous landslides and debris
flow. Post-disaster satellite images depict that the river
banks were eroded completely along the Kedarnath
valley due to the flash floods. Therefore, it is necessary
to identify the likely places of flash flooding and
vulnerable areas.
Durga Rao et al. (2014) concluded that an
extreme erosion took place in the upstream portion of
Kedarnath, besides the breach of Chorabari Lake and
deposition of debris/sediments in the valley. They have
carried out hydrological and hydraulic simulation study
of the Mandakini River using space-based inputs to
quantify the causes of the flash floods and their
impact. Flood inundation simulations were done using
CARTO DEM of 10 m posting in which the combined
effect of lake breach and high-intensity rainfall flood
was examined. As the slopes are very steep in the
upstream catchment area, lag-time of the peak flood
was found to be less and the Kedarnath valley was
washed away without any alert. The study reveals
quantitative parameters of the disaster which was due
to an integrated effect of high rainfall intensity, sudden
breach of Chorabari Lake and very steep topography.
Mandal and Chakrabarty (2016) developed a
simulation model of surface runoff in upper Teesta
basin using HEC-RAS and HEC-HMS by integrating
meteorological and morphological data in the
geospatial environment. Most recently, Chawla et al.
(2018) implemented a WRF model to investigate the
impact of different processes on extreme rainfall
simulation, by considering a representative event that
occurred during 15–18 June 2013 over the Ganga
Basin in India. They have improved the model
performance through incorporation of detailed land
surface models (LSMs) and analysed the effects of
model grid spacing with two sets of downscaling
ratios. They concluded that higher downscaling ratio
causes higher variability and consequently large errors
in the simulations.
Over the recent years, there has been an
increase in attention to improve flash flood warnings
in India. GCM predictions of climate change show an
increase in extreme precipitation events, which may
lead to more severe flash flooding. In addition, more
areas will be prone to flash flooding as a result of
increased urbanization (Jha et al., 2012). To address
hazards due to flash floods, there is a need for flash
flood forecasting with high spatial resolution and
adequate lead-time. Advances in flash flood
forecasting have been achieved through a range of
improvements in observing capabilities, modelling
techniques, and decision support systems
(Hapuarachchi et al., 2011). The most notable
improvements are satellite and radar observations and
associated techniques for use of these data. However,
India needs larger attention in the area of radar
hydrology, which is lacking at present. Advanced
techniques have been developed for deriving very high
resolution (spatial and temporal) real-time rainfall
estimates from weather radar data merged with
gauged rainfall. Also, a number of satellite-based
precipitation products with high temporal and spatial
resolution (near real-time) have recently been
developed. New techniques have been developed for
merging multiple sources of information to produce
rainfall forecasts with extended lead-times. These
techniques combine satellite-based rainfall, radar
rainfall, gauged rainfall and Numerical Weather
Prediction (NWP) model outputs to produce rainfall
forecasts (Xie and Arkin, 1996; Huffman et al., 1997;
Alberoni et al., 2000; Sinclair and Pegram, 2005;Dutta
et al., 2017; Hayden and Liu, 2018). Applications of
these techniques to Indian situations are still to be
developed and demonstrated.
Challenges in Flash Flood Forecasting
Developing a fully functional end-to-end forecasting
system ranging from data observations to public
evacuation involves many challenges. Lead time of
forecast is the most critical factor in such a system
as the time-lag between rainfall event and flash flood
is short. The forecasts should be made with adequate
lead-time to provide effective warnings, and therefore,
obtaining reliable forecasts of precipitation with
adequate lead-time are quite important. With advances
in science and computational facilities, precipitation
forecasting methods need to be improved through
better understanding of meteorological and
climatological processes and through better
observation networks.
Complex phenomena are involved for triggering
flash floods and to understand the underlying complex
natural hydrological processes, high quality observed
data are required including land use, soil properties,
soil moisture, morphology, and upstream conditions.
With the availability of reliable data, hydrologic/
Flood Modelling: Recent Indian Contributions 713
hydraulic models can be used for flash flood
forecasting and the uncertainty estimates of forecasts
improve the credibility of a forecast system. However,
the development of methodologies incorporating
uncertainties into the decision-making process is a
major challenge in flash flood forecasting and warning.
Further research is required for developing
dynamically varying probabilistic risks, which can be
used in the decision-making process.
For using a reliable forecast, even if the warnings
are accurate and on time, another major challenge is
to disseminate the alerts to general public to take proper
action. Though, with the advancements in
communication technologies, dissemination of flood
warnings to the public is possible, a knowledge gap is
often observed between the risk understood by the
public and the risk communicated by the authorities.
The improvement of knowledge and public awareness
of flash floods is therefore a very important aspect.
The research community must therefore work
together with the stakeholders including government
and the public to ensure that the science and
technology developed by them is put to right use for
societal benefit.
Urban Flooding
Recent increase in magnitude and frequency floods
in Indian cities of Mumbai, Chennai, Bengaluru,
Hyderabad, Vadodhara, Delhi and other cities have
posed a great challenge to the hydrologic modelling
community. Urban flooding is, in a hydrologic sense,
a case of flash flooding with high discharges
accumulating within a very short duration due to
increased impervious areas. The expansion of urban
areas due to population growth combined with climate
change increases the risk of frequent and severe urban
flash floods. Unfortunately, at the present stage, there
is no integrated model or system capable of dealing
with the complex hydrological behaviour of urban flash
floods. New modelling techniques, along with
extensive instrumentation and communication
infrastructure are needed for forecasting and
management of urban flash flood (WMO, 2012; WMO
2008).
The devastating floods that hit Chennai city and
other parts of Tamil Nadu during November-
December 2015 claimed more than 400 lives and
caused enormous economic damages. Such a
mammoth loss to life and property posed a challenge
to the scientific community in developing a
comprehensive understanding of the event. Answers
to a number of pressing questions related to the
conditions prevailing during and immediately preceding
the flooding period are necessary towards developing
such an understanding. These include, among others:
(a) what were the atmospheric conditions that caused
the high intensity rainfall, (b) how was the rainfall
distributed spatially and temporally, (c) how much flow
occurred in the three rivers passing through the city–
the Kosasthalaiyar River, the Cooum River and the
Adyar River - and the Buckingham canal, (d) how
were the two reservoirs upstream of the city, viz., the
Chembarambakkam reservoir and the Poondi
reservoir, operated, (e) how much overland flow was
generated in the city due to rainfall over the city alone,
(f) how did the storm water drainage system respond,
(g) which areas in the city were inundated and for
how long, (h) how did the waters recede after the
rains ceased, (i) what were the health implications of
the event, (j) did the land use change in the city over
the years exacerbate the flooding, and, most
importantly, (k) what actions need to be taken so that
for similar rainfall patterns repeating in future, the
city would not face such a devastating deluge?. A
rapid assessment report of the event, prepared by a
voluntary team of researchers (Narasimhan et al.,
2016), addresses these issues in a great detail.
In case of urban flood modelling, extreme
precipitation analysis takes an important place.
Quantification of extreme precipitation and associated
uncertainties at short duration is important. Chandra
Rupa et al. (2015) concluded that Bayesian analysis
quantifies uncertainties in the parameters of an
extreme value distribution fitted to the data in a better
way, when compared to standard parameter estimation
methods like Maximum Likelihood Estimation (MLE).
Bayesian analysis enables reliable estimates of short
duration extreme precipitation events.Fig.3a shows
the posterior distribution of return levels for 10-year
return period and for durations of 15-min, 30-min, -
hour, 3-hour, 6-hour, 12-hour and 24-hour, obtained
from the Bayesian approach and using MLE
approach. Fig. 3b shows the posterior distribution of
return levels for 15-min duration for different return
periods. The return levels obtained from the estimates
of parameters using MLE method are marked (black
circles) in Fig. 3. From the figure, it is seen that the
714 R Chandra Rupa and PP Mujumdar
return levels obtained from the MLE method shifts
from the mode of the posterior distribution of return
levels obtained from Bayesian analysis as the duration
and the return period reduces. Typically, in urban
areas, the infrastructure is built for short duration
precipitation events and for return periods of order 2-
year and 5-year. At these short durations and low
return periods, the standard statistical methods like
MLE underestimates the return levels.
In addition to the precipitation, other factors such
as changes in land use, climate change has impact on
the urban flooding (Praskievicz and Chang, 2009).
Zope et al. (2016) investigated the impact of land
use–land cover (LULC) change and urbanization on
floods for an urban catchment of the Oshiwara River
in Mumbai using HEC-GeoHMS and HEC-HMS
models. They have concluded that the flood
inundation area is increased by 5.61% for the 100-
year return period and 6.04% for the 10-year return
period. They showed that the lower return periods
led to a maximum change in peak discharge/volume
of runoff compared to higher return periods for change
in land use conditions. Need for an integrated
approach to flood management, considering the land
use change into the hydrological model is emphasised
by Suriyaand Mudgal (2012).Considering several
factors affecting the extreme precipitation, Agilan and
Umamahesh (2015) analysed changes in daily and
sub-daily (4-h) extreme rainfall using various climate
change detection indices for Hyderabad city. They
concluded that non-stationarity in daily extreme rainfall
is associated with global processes (ENSO cycle and
global warming) and non-stationarity in sub-daily (4-
h) extreme rainfall is associated with local processes
(urbanization and local temperature changes).These
findings have a great implication in urban flood
modelling. Other cities suchas Delhi (Kovats and
Akhtar, 2008), Chennai and Kolkata (Sen, 2013) are
also studied to understand the impacts on urban floods
due to changes in precipitation, land-use and climate.
In addition, improper maintenance and management
of hydraulic infrastructure often causes flooding in
urban areas (Gupta, 2007; Narasimhan et al., 2016).
Risk Assessment and Mitigation Measures
Around the world, acceleration in population growth
and changes in land-use patterns and climate have
increased human vulnerability to floods. Harmful
impacts of floods include mortality, morbidity and
widespread damage of crops, infrastructure and
property (Doocy et al., 2013; IPCC, 2007). However,
if the floods are modelled accurately and if risk
assessment studies are conducted, the impacts can
be minimised by pre-emptive actions. A general
(A) (B)
Fig. 3. (A) Probability density function (pdf) of return levels (mm/hr) for 10-year return period for various durations in
comparison with return levels from Maximum Likelihood Estimation (MLE) method. Black circle is the return level
obtained from MLE method at the corresponding duration. (B) pdf of return levelsfor15-min duration at different
return periods in comparison with return levels obtained from MLE method (Black circles). As the duration and the
return period reduces the return levels obtained from the MLE method shifts away from the mode of the posterior
distribution of return levels obtained from Bayesian approach
Flood Modelling: Recent Indian Contributions 715
approach to define flood risk is by considering the
product of the flood hazard (i.e., the physical and
statistical aspects of flooding, like the magnitude,
extent and hours of flooding, return period of the event
etc.) and the vulnerability (i.e., the exposure of floods
to people and assets, and the liability of elements at
risk to suffer from flood damage etc.). Flood risk
management involves not only managing the prevailing
flood risk situation, but also planning for a system that
aims to minimize flood risk (WMO 2009a). This
process involves risk analysis on a regular basis and
evaluation of hazard based on the latest information,
hydrometeorological data, technical developments and
altered conditions due to urbanization and land use
change. In this context, web based GIS tools have
proven to be useful for flood mapping as well as risk
mapping (Sanyal and Lu, 2006; Kulkarni et al., 2014,
Shivaprasad Sharma et al., 2018; Hazarika et al.,
2018). Artificial intelligence and computational
intelligence methods based on big data analysis are
also applied for early flood detection (Fotovatikhah et
al., 2018).
Bajracharya et al. (2007) assessed glacial lake
outburst flood hazard in the Sagarmatha region using
dam break and hydrodynamic modelling. They have
prepared a glacial lake outburst flood vulnerability
rating map to identify vulnerable settlements. Sindhu
and Durga Rao (2016) modelled the Brahmani-
Baitarani River Basin for flood damage mitigation
assessment. Abbas et al. (2015) assessed the policy
and planning processes and flood-related scientific
research in India, Pakistan and Bangladesh. A
comparison of the existing flood management systems
of the three countries is undertaken based on a
systematic review, and a framework for sustainable
flood management in the region is suggested. Their
results of the literature analysis reveal poor support
from scientific research focusing on flooding issues
in the case of Pakistan, while Bangladesh and India
seem to have benefited from research support in
formulating their flood management strategies. In
India, the Natural Disaster Monitoring Agency
(NDMA) has prepared guidelines for flood
management and suggested structural and non-
structural measures (NDMA, 2008).
Flood risk reduction and management strategies
in urban context with example of the Chennai city
are discussed by Gupta and Nair (2010). The Chennai
rapid assessment report, prepared by Narasimhan et
al. (2016), is a voluntary work by researchers from
different institutes in the country to provide an
understanding of various factors that influenced the
devastating floods in Chennai during Nov. and Dec.
2015. Kumar et al. (2017) analysed the existing flood
control measures. Based on their analysis, they
suggested several implementable structural and non-
structural measures for alleviating the problem of
riverine as well as urban flooding in the national capital
territory of Delhi. In case of Bangalore, an Integrated
Urban Flood Management (Mujumdar et al., 2017)
project is carried out by different research
organizations and risk maps showing the areas of
flooding for a pilot study area in Bengaluru city have
been developed (Fig. 4).
Concluding Remarks
The brief review provided in the paper highlights the
contributions of the Indian research community to the
area of flood modelling. While the scientific
contributions have been significant, the corresponding
transfer of the knowledge to policy and implementation
has been poor. This section brings out perspectives
on directions to be pursued both by the research
community and the policy makers to ensure that the
results of research are useful in better understanding
the science of floods and mitigating the impacts of
floods.
Changing rainfall patterns, due to both natural
and anthropogenic causes, have exacerbated the
flooding problem in India. The frequencies, magnitudes
and the spatial extent of the floods are known to be
increasing in the country. Unfortunately, infrastructure
development has lagged behind the economic and
population growth, resulting in increasing losses and
damages due to floods. Poorly conceived drainage
infrastructure, coupled with other problems like
encroachments, have significantly contributed to the
flood deluge year after year, in the face of increasing
rainfall intensities, making flash floods a common
occurrence. Capacity to deal with rapid changes -
such as increase in in extreme rainfall events and
rapid urbanisation and the ability to anticipate and
adapt to slow changes and trends (population increase,
climate change) is very minimal in the country, which
poses new challenges for flood management. To
overcome thesechallenges and to aim towards a flood
716 R Chandra Rupa and PP Mujumdar
resilient society the following specific actions are
proposed that integrate the scientific knowledge with
technologies and administrative policies:
l Data Issues: There are several sources of data
such as ground observations, satellite imagery,
radar data and histories of past flood situations,
which provides a valuable information and helps
in modelling the floods precisely. Integrated
approaches, merging all the relevant data, is
required to overcome data crunch and avoid
uncertainties arise due to insufficient quality and
quality of data. It is also the responsibility of the
custodians of the data (typically, the government
agencies that collect and archive the data) to
make the data available for research studies.
Events such as the Kedarnath floods of 2013
and 2015 could lend a great deal of insight if
models are developed to reconstruct those
events. While the scientific capability to do this
exists in the country, the efforts are greatly
handicapped by a lack of useful data – both
because of an absence of data and because of
inaccessibility to the available data.
l Integrated modelling: In general, flood
forecasting and risk management is quite
challenging for many reasons, including those
due to uncertaintiesarising out of lack of data,
calibration of modelparameters etc. Many
researchers have worked on quantifying
uncertainties in extreme precipitation, modelling
of drainage system using hydrologic models,
quantification of uncertainties in using different
hydrologic models, risk and reliability analysis,
creation of hazard and vulnerability maps and
measures for sustainable development.
However, an integrated approach is lacking, and
it is an important aim to pursue by the
researchers as high uncertainties are associated
in each step of modelling. It is also necessary
that the large number of tools and models
developed for a given river basin (e.g., the
Mahanadi river basin) are compared for their
effectiveness and the results synthesised to
make them useful for policy makers. It is the
responsibility of the research community to
communicate the uncertainty in the results in a
way that can be understood and used by the
policy makers.
l Flood inundation maps: Generation of what-if
scenarios based on probability of flooding and
the resilience of the infrastructure system and
development of new methodologies for planning
under uncertainty for flood management is
required. Probabilistic flood inundationmaps –
similar to the seismic maps developed for the
country - should be developed for each basin,
Fig. 4: Flood inundation maps of the Hulimavu-Madivala catchment in Bengaluru (A) under normal conditions; (B) with
different land-use characteristics (increased urbanization) and (C) with 30% reduction in runoff due to rainwater
harvesting
(C)(B)(A)
Flood Modelling: Recent Indian Contributions 717
and must be revisited frequently to account for
the rapidly changing landuse, climate,
demography etc.
l Collaboration between the Research Community
and Policy Makers: A significant amount of
internationally recognised research is
conductedin the country today. Several advances
have been made in developing reliable models,
improving flood forecasts and quantifying and
reducing uncertainties. However, this knowledge
is not transferred to the policy and administrative
decision making. The government bodies and
the policy makers should involve researchers
while developing contingency planning for
disasters. Also, advanced forecasting and
warning systems with sufficient lead times
should be developed with the help of researchers
to disseminate flood warnings to the local
communities under threat. The National Disaster
Management Agency (NDMA) has played an
important role in bringing the two – apparently
disparate communities – together when
guidelines were developed by the Agency. Such
efforts should be pursued with vigour at all
administrative levels. It is also incumbent on the
research community to respond positively to such
initiatives.
l Flood risk management as part of IWRM: The
failure of localised attempts to deal with fluvial
flooding has stressed the need to take a strategic
approach to catchment management. Also, there
are compelling arguments to deal with other
catchment functions at a river basin scale,
including water supply and hydro-ecology.
Therefore, flood risk management needs to
become an integral aspect of a multi-functional
approach to river basins as it is addressed in
IWRM (Integrated Water Resources
Management). Adaptive management
approaches including the interdisciplinary,
system-oriented and trans-disciplinary research
are required to cope up with the increasing
uncertainties due to global and climate changes
and the fast changing socio-economic conditions.
In essence, IWRM approaches combining the
hydrological models, flood risk management and
socio-economic conditions while preserving the
ecosystems are much commanded (Grabs et al.,
2007, WMO 2009b, WMO 2011; WMO 2017).
l Flood as a resource: With severe water stresses
occurring year after year in several parts of the
country – and simultaneously, the floods causing
a huge damage every year – it is necessary to
evolve a new paradigm of floods as a resource
and not as a disaster. Newer ways and methods
of storing and using the flood waters need to be
developed. This may require identifying
groundwater potential zones to recharge the
groundwater, analysing quality of flood water
and assessing the possibility of implementation
of green infrastructure and rainwater harvesting
units to harvest/reuse the flood water. Flood
modelling helpsin identifying the potential areas
of flooding, flood magnitudes and the associated
time of flooding. With this information, the flood
volumes can be utilized to either percolate into
the ground depending on the ground water
conditions or for storing by linking the storage
units in particular catchments and/or designing
the storage units depending on the amount of
flood.
l Training and Capacity Development: As in many
other fields, there is a very high need of trained
manpower in the country to provide a scientific
support to policy makers on large number of
issues related to floods. Besides the
contemporary approaches to mitigating flood
effects, there is a vast repository of traditional/
indigenous knowledge in the country useful in
local adaptation practices. Such conventional
practices are to be integrated with advanced
scientific knowledge to provide adaptive
responses to the flooding problem at all scales.
References
Abbas A, Amjath-Babu TS, Kächele H, Usman M and Müller K
(2015) An overview of flood mitigation strategy and
research support in South Asia: implications for sustainable
flood risk management International Journal of Sustainable
Development & World Ecology 23: 1, 98-111, doi: 10.1080/
13504509.2015.1111954
Agilan V and Umamahesh NV (2015), Detection and attribution
718 R Chandra Rupa and PP Mujumdar
of non-stationarity in intensity and frequency of daily and
4-hour extreme rainfall of Hyderabad, India Journal of
Hydrology 530 677-697, doi: https://doi.org/10.1016/
j.jhydrol.2015.10.028.
Alberoni PP, Ducrocq V, Gregori G, Haase G, Holleman I, Lindskog
M, Macpherson B, Nuret M and Rossa A (2000)
Assimilation of Radar Precipitation Data in NWP Models-
A Review Physics and Chemistry of The Earth Part B-
hydrology Oceans and Atmosphere 25 10.1016/S1464-
1909(00)00185-4
Asokan AM and Dutta D (2008) Analysis of water resources in
the Mahanadi River Basin, India under projected climate
conditions Hydrol Process 22 3589-3603 doi: 10.1002/
hyp.6962
Bajracharya B, Shrestha AB and Rajbhandari L (2007) Glacial
Lake Outburst Floods in the Sagarmatha Region Mountain
Research and Development 27 336-344 doi: 10.1659/
mrd.0783
Basu B and Srinivas VV (2014) Flood frequency analysis using
a novel mathematical approach International Journal of
Engineering Research Innovative Research Publications 3
209-213 ISSN: 2319-6890
Beven K (2006) On undermining the science? Hydrological
Processes 20 3141-3146 https://onlinelibrary.wiley.com/
doi/pdf/10.1002/hyp.6396
Bürger G, Sobie SR, Cannon AJ, Werner AT and Murdock TQ
(2012) Downscaling extremes: An intercomparison of
multiple methods for future climate Journal of Climate 26
3429-3449 http://dx.doi.org/10.1175/JCLI-D-12-00249.1
Chandra Rupa R, Saha U and Mujumdar PP (2015) Model and
Parameter Uncertainty in IDF Relationships under Climate
Change Advances in Water Resources 79 127-139 https://
www.sciencedirect.com/science/article/pii/S03091708
15000433
Chawla I and Mujumdar PP (2015) Isolating the Impacts of land
Use and Climate Change on Streamflow Hydrol Earth Syst
Sci 19 3633-3651 doi: 10.5194/hess-19-3633-2015
Chawla I and Mujumdar PP (2019) Evaluating, Rainfall Datasets
to reconstruct Floods in Data Sparse Himalayan Region,
Journal of Hydrology (Manuscript no. HYDROL 33962,
under review)
Chawla I and Mujumdar PP (2018) Partitioning Uncertainty in
Streamflow Projections under Nonstationary Model
Conditions Advances in Water Resources 112 266-282
https://www.sciencedirect.com/science/article/pii/
S0309170817300179
Chawla I, Osuri KK, Mujumdar PP and Niyogi D (2018)
Assessment of the Weather Research and Forecasting
(WRF) model for simulation of extreme rainfall events in
the upper Ganga Basin Hydrol Earth Syst Sci 22 1095-
1117 (https://doi.org/10.5194/hess-22-1095-2018)
Deletic A, Dotto CBS, McCarthy DT, Kleidorfer M, Freni G,
Mannina G, Uhl M, Henrichs M, Fletcher TD, Rauch W,
Bertrand-Krajewski JL and Tait S (2012) Assessing
uncertainties in urban drainage models, Physics and
Chemistry of the Earth 42 3-10 https://www.sciencedirect.
com/science/article/pii/S1474706511000623
Doherty J and Welter D (2010) A short exploration of structural
noise Water Resources Research 46 W05525 doi:10.1029/
2009WR008377
Doocy S, Daniels A, Murray S and Kirsch TD (2013) The Human
Impact of Floods: A Historical Review of Events 1980-
2009 and Systematic Literature Review PLOS Currents
Disasters Edition 1. doi: 10.1371/currents.dis.
f4deb457904936b07c09daa98ee8171a
Durga Rao KHV, Rao VV, Dadhwal VK and Diwakar PG (2014)
Kedarnath flash floods: A hydrologicaland hydraulic
simulation study Current Science 106 4598-603
Durga Rao KHV, Rao VV, Dadhwal VK, Behera G and Sharma JR
(2011) A Distributed Model for Real-Time Flood
Forecasting in the Godavari Basin Using Space Inputs Int
J Disaster Risk Sci 2 31-40, doi: 10.1007/s13753-011-
0014-7
Dutta D, KasimahanthiA J, MallickS, GeorgeJ P, and DevarajanP
K, 2017: Quality assessment of VVP winds from Indian
Doppler weather radars: a data assimilation perspective.
Journal of applied remote sensing11(3) 036021
Fotovatikhaha F, Herrera M, Shamshirband S, Chaue K, Ardabili
SF and Piran Md. J (2018) Survey of computational
intelligence as basis to big flood management: challenges,
research directions and future work Engineering
Applications of Computational Fluid Mechanics 12 411-
437, https://doi.org/10.1080/19942060.2018.1448896
Ghosh S, Raje D and Mujumdar PP (2010) Mahanadi streamflow:
climate change impact assessment and adaptive strategies
Current Science 98 1084-1091
Gopakumar R and Mujumdar PP (2008) A Fuzzy Dynamic Wave
Routing Model, Hydrological Processes 22 1564-1572,
doi:10.1002/hyp.6727
Gosain AK, Rao S and Basuray D (2006) Climate change impact
assessment on hydrologyof Indian river basins, Current
Science Special Section: Climate Change and India 90 346-
353
Guhathakurta P, Sreejith OP and Menon PA (2011) Impact of
climate change on extreme rainfall events and flood risk in
India J Earth Syst Sci 120 359-373 https://link.springer.
Flood Modelling: Recent Indian Contributions 719
com/article/10.1007/s12040-011-0082-5
Gupta AK and Nair SS (2010) Flood risk and context of land-
uses: Chennai city case Journal of Geography and Regional
Planning 3 365-372
Gupta K (2007) Urban flood resilience planning and management
and lessons for the future: a case study of Mumbai, India
Urban Water Journal 4 183-194, https://www.tandfonline.
com/doi/abs/10.1080/15730620701464141
Guru N and Jha R (2015) Flood Frequency Analysis of Tel Basin
of Mahanadi River System, India using Annual Maximum
and POT Flood Data Aquatic Procedia 4 427-434, https:/
/www.sciencedirect.com/science/article/pii/S2214241X
15000589
Hapuarachchi HA, Wang QJ and Pagano TC (2011) A review of
advances in flash flood forecasting Hydrol Process 25 2771-
2784. doi:10.1002/hyp.8040
Harremoës P (2003) The role of uncertainty in application of
integrated urban water modeling, In: International IMUG
Conference. Tilburg, Netherlands
Hayden L and Chuntao L (2018) A Multiyear Analysis of Global
Precipitation Combining CloudSat and GPM Precipitation
Retrievals Journal of Hydrometeorology 19 1935-1952
Hazarika N, Barman D, Das AK, Sarma AK and Borah SB (2018),
Assessing and mapping flood hazard, vulnerability and
risk in the Upper Brahmaputra River valley using
stakeholders’ knowledge and multicriteria evaluation
(MCE) J Flood Risk Management 11 S700-S716, https://
onlinelibrary.wiley.com/doi/pdf/10.1111/jfr3.12237
Hirabayashi Y, Mahendran R, Koirala S, Konoshima L, Yamazaki
D, Watanabe S, Kim H and Kanae S (2013) Global flood
risk under climate change Nature Climate Change3 816-
821; https://www.nature.com/articles/nclimate1911
Huffman G, Adler RF, Arkin P, Chang A, Ferraro R, Gruber A,
Janowiak J, McNab A, Rudolf B and Schneider (1997)
The Global Precipitation Climatology Project (GPCP)
Combined Precipitation Dataset Bulletin of the American
Meteorological Society 78 5-20 https://doi.org/10.1175/
1520-0477(1997)078<0005:TGPCPG>2.0.CO;2
Indu J and D Nagesh Kumar (2014) Evaluation of TRMM
PRsampling error over a subtropical basin using
bootstraptechnique IEEE Transactions on Geoscience and
Remote Sensing 52 art. no. 6757000, pp. 6870-6881 doi:
10.1109/TGRS.2014.2304466
IPCC - Intergovernmental Panel on Climate Change (2012) Special
report on Managing the risks of extreme events and
disasters to advance climate change adaptation (SREX), a
special report of Working groups I and II of the
Intergovernmental Panel on climate change, (ed.) Field,
CB, Barros V, Stocker TF, Qin D, Dokken DJ, Ebi KL,
Mastrandrea MD, Mach KJ, Plattner G-K, Allen SK,
Tignor M and Midgley PM, Cambridge University Press,
Cambridge and New York
IPCC (2007) Contribution of Working Group II to the Fourth
Assessment Report of the Intergovernmental Panel on
Climate Change. In: Parry ML, Canziani OF, Palutikof JP,
van der Linden PJ and Hanson CE (Eds.). Cambridge
University Press, Cambridge, United Kingdom and New
York, NY, USA
Jain SK and Kumar V (2012) Trend analysis of rainfall and
temperature data for India Current Science 102 37-49
Jena PP, Chatterjee C, Pradhan G and Mishra A (2014) Are recent
frequent high floods in Mahanadi basin in eastern India
due toincrease in extreme rainfalls? Journal of Hydrology
517 847-862, https://www.sciencedirect.com/science/
article/pii/S0022169414004818.
Jha AK, Bloch R and Lamond J (2012) Cities and Flooding, A
Guide to Integrated Urban Flood Risk Management for
the 21st Century, Published by World Bank, doi:10.1596/
978-0-8213-8866-2, https://doi.org/10.1596/978-0-8213-
8866-2
Johnston R and SmakhtinV (2014) Hydrological modeling of
large river basins: How much is enough? Water Resour
Manage 28 2695-2730, doi 10.1007/s11269-014-0637-8
Kamal V, Mukherjee S, Singh P, Sen R, Vishwakarma CA, Sajadi
P, Asthana H and Rena V (2017) Flood frequency analysis
of Ganga river at Haridwar and Garhmukteshwar Appl
Water Sci 7 1979-1986, doi 10.1007/s13201-016-0378-3
Katz RW, Parlange MB and Naveau P (2002) Statistics of extremes
in hydrology Adv Water Resour 25 1287-1304, https://
www.sciencedirect.com/science/article/pii/S030917080
2000568
Kovats S and Akhtar R (2008) Climate, climate change and human
health in Asian cities Environment and Urbanization 20
165-175, https://doi.org/10.1177/0956247808089154.
Kulkarni AT, Mohanty J, Eldho TI, Rao EP and Mohan BK
(2014) A web GIS based integrated flood assessment
modelling tool for coastal urban watersheds Computers &
Geosciences 64 7-14 https://www.sciencedirect.com/
science/article/pii/S0098300413002896
Kumar S, Sarkar A, Thakur SK and Shekhar S (2017)
Hydrogeological characterization of aquifer in palla flood
plain of Delhi using integrated approach Journal of the
Geological Society of India 90 459-466, https://link.
springer.com/article/10.1007/s12594-017-0739-z
Kumar R, ChatterjeeC, KumarS, Lohani A K and Singh R D
720 R Chandra Rupa and PP Mujumdar
(2003), Development of Regional Flood Frequency
Relationships Using L-moments for Middle Ganga Plains
Subzone 1(f) of IndiaWater Resources Management17 243–
257, https://link.springer.com/article/10.1023/A:1024770
124523
Lohani AK, Goel NK and Bhatia KKS (2006) Takagi-Sugeno
fuzzy inference system for modelling stage–discharge
relationship Journal of Hydrology 331 146-160 https://
www.sciencedirect.com/science/article/pii/S002216940
6002733
Lohani AK, Goel NK and Bhatia KKS (2014) Improving real
time flood forecasting using fuzzy inference system Journal
of Hydrology 509 25-41 https://www.sciencedirect.com/
science/article/pii/S0022169413008378
Mandal S, Srivastav RK and Simonovic SP (2016) Use of Beta
Regression for Statistical Downscaling of precipitation in
the Campbell River Basin, Brithish Columbia, Canada
Journal of Hydrology 538 49-62,https://www.
sciencedirect.com/science/article/pii/S002216941630 1901
Mandal SP and Chakrabarty A (2016) Flash flood risk assessment
for upper Teesta river basin: usingthe hydrological
modeling system (HEC-HMS) software Model Earth Syst
Environ 2 59, doi 10.1007/s40808-016-0110-1
Mitra AK, IM Momin, EN Rajagopal, S Basu, MN Rajeevan and
TN Krishnamurti (2013) Gridded Daily Indian Monsoon
Rainfall for 14 Seasons: Merged TRMM and IMD Gauge
Analyzed Values Journal of Earth System Science 122
1173-1182, doi:10.1007/s12040-013-0338-3
Mondal A and PP Mujumdar (2012) On the basin-scale detection
and attribution of human induced climate change in monsoon
precipitation and streamflow Water Resources Research
48 W10520, doi:10.1029/2011WR011468
Mondal A and PP Mujumdar (2015) Modeling Non-Stationarity
in intensity, duration and frequency of extreme rainfall
over India Journal of Hydrology 521 217-231, https://
www.sciencedirect.com/science/article/pii/S002216941400
9937
Mondal A and PP Mujumdar (2016) Detection of Change in
Flood Return Levels under Global Warming, ASCE Journal
of Hydrologic Engineering, 10.1061/(ASCE)HE.1943-
5584.0001326, 04016021, http://dx.doi.org/10.1061/
(ASCE)HE.1943-5584.0001326
Mondal A, B Narasimhan, S Muddu and PP Mujumdar (2016)
Hydrologic modelling, Proc. Indian National Science
Academy, 82, 3, 817-832 doi:10.16943/ptinsa/2016/48487
Mondal A, Lakshmi V, Hashemi H (2018) Intercomparison of
trend analysis of Multisatellite Monthly Precipitation
Products and Gauge Measurements for River Basins of
India Journal of Hydrology 565 779-790
Mujumdar PP and Nagesh Kumar D (2012) Floods in a Changing
Climate: Hydrologic Modeling, International Hydrology
Series, Cambridge University Press, Cambridge, U.K.,
ISBN-13: 9781107018761
Mujumdar PP, Mohan Kumar MS, Srinivasa Raju K, Umamahesh
NV, Valsalam R, Srinivasa Reddy GS and Niyogi D (2017)
Integrated Urban Flood Management inIndia – Technology
Driven Solutions, Monograph submittedto ITRA,
Government of India Jul. 2017
Mujumdar PP (2001) Flood Wave Propagation - The Saint Venant
Equations Resonance 6 66-73
Mujumdar PP and Ghosh S (2008) Climate change impact on
hydrology and water resources Journal of Hydraulic
Engineering 14 1-17
Mukerji A, Chatterjee C and Raghuwanshi NS (2009) Flood
Forecasting Using ANN, Neuro-Fuzzy, and Neuro-GA
Models ASCE Journal of Hydrologic Engineering 14 647-
652, doi: 10.1061/(ASCE)HE.1943-5584.0000040
NandiS and Reddy MJ (2017) Distributed rainfall runoff modeling
over Krishna river basin European Water 57 71-76, http:/
/www.ewra.net/ew/pdf/EW_2017_57_10.pdf
Narashimhan B, Bhallamudi SM, Mondal A, Ghosh S and
Mujumdar PP (2016) Chennai floods 2015 – A rapid
assessment, Technical Report, Interdisciplinary Centre for
Water Research, IISc Bangalore, May 2016 (available
online, http://www.icwar.iisc.ernet.in/wp-content/uploads/
2016/06/Chennai-Floods-Rapid-Assessment-Report.pdf)
NDMA (2008) National DisasterManagement Guidelines-
Management of Floods, National Disaster Management
Authority, Government of India
NIC (2009) India: The Impact of Climate Change to 2030, A
Commissioned Research Report prepared by Joint Global
Change Research Institute and Battelle Memorial Institute,
Pacific Northwest Division, USA, National Intelligence
Council, NIC 2009-03D
Patro S, Chatterjee C, Singh R and Raghuwanshi NS (2009)
Hydrodynamic modelling of a large flood-prone river
systemin India with limited data Hydrol Process 23 2774-
2791 doi: 10.1002/hyp.7375
Pattanayak S, Nanjundiah RS and Nagesh Kumar D (2017) Linkage
between global sea surface temperature and
hydroclimatology of a major river basin of India before
and after 1980 Environ Res Lett 12 124002, https://doi.org/
10.1088/1748-9326/aa9664
Perumal M and B Sahoo (2007) Applicability criteria of the
Flood Modelling: Recent Indian Contributions 721
variable parameter Muskingum stage and discharge routing
methods Water Resour Res 43 W05409, doi: 10.1029/
2006WR00490
Perumal M and Price RK (2013) A fully mass conservative variable
parameter McCarthy-Muskingum method: Theory and
verification Journal of Hydrology 502 89-102 https://
www.sciencedirect.com/science/article/pii/S002216941300
601X
Praskievicz S and Chang H (2009) A review of hydrological
modelling ofbasin-scale climate change and
urbandevelopment impacts Progress in Physical
Geography 33 650-671 doi: 10.1177/0309133309348098
Prudhomme C and Davies H (2009) Assessing Uncertainties in
Climate Change Impact Analyses on the River Flow
Regimes in the UK. Part 1: Baseline Climate Climatic
Change 93 177-195, http://dx.doi.org/10.1007/s10584-
008-9464-3
Raje D and Mujumdar PP (2010a) Hydrologic Drought Prediction
under Climate Change: Uncertainty Modeling with
Dempster-Shafer and Bayesian Approaches. Advances in
Water Resources, doi: 10.1016 / j.advwatres.2010.08.001
(Pub: Elsevier, Netherlands pdf
Raje D and Mujumdar PP (2010b) Constraining Uncertainty in
Regional Hydrologic Impacts of Climate Change:
Nonstationarity in Downscaling, Water Resources
Research, 46, W07543, doi:10.1029/2009WR008425.
(Pub: American Geophysical Union)
Rahman MM, Goel NK and Arya DS (2012) Development of
the Jamuneswari Flood Forecasting System: Case Study
in Bangladesh ASCE Journal of Hydrologic Engineering
17 1123-1140, doi: 10.1061/(ASCE)HE.1943-5584.
0000565
Santhosh D and Srinivas VV (2013) Bivariate frequency analysis
of floods using a diffusion based kernel density estimator
Water Resources Research American Geophysical Union
& Wiley 49 8328-8343, doi:10.1002/2011WR010777
Sanyal J and Lu XX (2006) GIS-based flood hazard mapping at
different administrative scales: A case study in Gangetic
West Bengal, India, Singapore J Tropical Geography 27
207-220 doi:10.1111/j.1467-9493.2006. 00254.x
Sen DJ (2013) Real-time rainfall monitoring and flood inundation
forecasting for the city of Kolkata, ISH Journal of Hydraulic
Engineering 19 137-144, http://dx.doi.org/10.1080/
09715010.2013.787718
Shah HL and Mishra V (2016), Uncertainty and Bias in Satellite-
Based Precipitation Estimates over Indian Subcontinental
Basins: Implications for Real-Time Streamflow Simulation
and Flood Prediction J Hydrometeorology, American
Meteorological Society, doi : 10.1175/JHM-D-15-0115.1
Sharma VK, Mishra N, Shukla AK, Yadav A, Rao GS and
Bhanumurthy V (2017) Satellite data planning for flood
mapping activities based on high rainfall events generated
using TRMM, GEFS and disaster news Annals of GIS, 23
131-140, doi: 10.1080/19475683.2017.1304449
Shivaprasad Sharma SV, Roy PS, Chakravarthi V and Srinivasa
Rao G (2018) Flood risk assessment using multi-criteria
analysis: a case study from Kopili River Basin, Assam,
India Geomatics, Natural Hazards and Risk 9 79-93, https:/
/doi.org/10.1080/19475705.2017.1408705
Sinclair S and Pegram G (2005) Combining radar and rain gauge
rainfall estimates using conditional merging Atmosph Sci
Lett 6 19-22 doi: 10.1002/asl.85
Sindhu S and Durga Rao KHV (2016) Hydrological and
Hydrodynamic Modeling for Flood Damage Mitigation in
Brahmani-Baitarani River Basin, India Geocarto
International, doi: 10.1080/10106049.2016.1178818.
Singh A, Imtiyaz Mohd, Isaac RK and Denis DM (2014)
Assessing the performance and uncertainty analysis of
the SWAT and RBNN models for simulation of sediment
yield in the Nagwa watershed, India Hydrological Sciences
Journal 59 351-364, DOI:10.1080/02626667.
2013.872787
Singh RD (2008) Real-time flood forecasting: Indian experience,
Chapter 10, Hydrological modelling in arid and semi-arid
areas, Edited by WheaterH, Sorooshian S and Sharma KD,
Cambridge University Press, Cambridge, United Kingdom
and New York, NY, USA
Sonali P and Nagesh Kumar D (2016) Detection and attribution
of seasonal temperature changes in India with climate
models in the CMIP5 archive J Water and Climate Change
7 83-102, http://jwcc.iwaponline.com/content/early/2015/
10/24/wcc.2015.072
Srivastav RK, Sudheer KP and Chaubey I (2007) A simplified
approach to quantifying predictive and parametric
uncertainty in artificial neural network hydrologic models
Water Resour Res 43 W10407, doi: 10.1029/2006WR00
5352
Suriya S and Mudgal BV (2012) Impact of urbanization on
flooding: The Thirusoolam sub watershed – A case study
J Hydrology 412-413 210-219 https://www. sciencedirect.
com/science/article/pii/S002216941100 3180
Thirumalaiah K and Deo MC (2000) Hydrological forecasting
using neural networks, ASCE J hydrological Engineering 5
April 2000, 180-189 https://ascelibrary.org/doi/abs/
10 .1061/%28ASCE%291084-0699%282000%
295%3A2%28180%29
722 R Chandra Rupa and PP Mujumdar
Warmink JJ, H Van der Klis, Booij MJ and Hulscher SJMH
(2011) Identification and quantification of uncertaintiesin
a hydrodynamic river model using expert opinions Water
Resour Manage 25 601-622, doi: 10.1007/s11269-010-
9716-7
Wilby RL and Keenan R (2012) Adapting to flood risk under
climate change, Progress in Physical Geography: Earth
and Environment 36 348-378, http://journals.sagepub.
com/doi/abs/10.1177/030913331243 8908
WMO (2008) Urban Flood Risk Management. APFM Technical
Document No. 6, Flood Management Tools Series,
Associated Programme on Flood Management (WMO),
Geneva. Available at: http://www.apfm.info/pdf/ifm_tools/
Tools_Urban_Flood_Risk_Management.pdf
WMO (2009a) Flood Management in a Changing Climate. APFM
Technical Document No. 9, Flood Management Tools
Series, Associated Programme on Flood Management
(WMO), Geneva. Available at: http://www.apfm.info/pdf/
ifm_tools/Tools_FM_in_a_changing_climate.pdf
WMO (2009b) Integrated Flood Management - Concept Paper.
APFM Document No. 1047. Associated Programme on
Flood Management (APFM). Available at: http://
www.apfm.info/pdf/concept_paper_e.pdf
WMO (2011a) IFM as an Adaptation Tool for Climate Change:
Case Studies. APFM Technical Document No. 14, Flood
Management Tools Series, Associated Programme on Flood
Management (WMO), Geneva. Available at: http://
www.apfm.info/pdf/ifm_tools/Case_studies_CCA.pdf
WMO (2012) Urban Flood Management in a Changing Climate.
Flood Management Tools Series No. 14, Associated
Programme on Flood Management (WMO), Geneva.
Available at: http://www.floodmanagement.info/
publications/tools/APFM_Tool_14.pdf
WMO (2017) Selecting Measures and Designing Strategies for
Integrated Flood Risk Management. Policy and Tools
Documents Series No.1 version 1.0, Associated
Programme on Flood Management (WMO), Geneva
Xie P and Arkin PA (1996) Analyses of Global Monthly
Precipitation Using Gauge observations, Satellite
Estimates, and Numerical Model Predictions Journal of
Climate 9 840-858, doi: 10.1175/1520-0442(1996)009
<0840:AOGMPU>2.0.CO;2
Yaduvanshi A, Srivastava P, Worqlul AW and Sinha AK (2018)
Uncertainty in a Lumped and a Semi-Distributed Model
for Discharge Prediction in Ghatshila Catchment Water 10
381; doi:10.3390/w10040381
Yaduvanshi A, Sharma RK, Kar SC and Sinha AK (2018) Rainfall-
runoff simulations of extreme monsoon rainfall events in a
tropical river basin of India Nat Hazards 90 843-861
https://doi.org/10.1007/s11069-017-3075-0
Zope PE, Eldho TI and Jothiprakash V (2016) Impacts of land
use–land cover change and urbanization on flooding: Acase
study of Oshiwara River Basin in Mumbai, India, Catena
145 142-154, http://dx.doi.org/10.1016/j.catena.2016.
06.009.