International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 709

ISSN 2229-5518

FLOOD MITIGATION STUDY IN

ADAYAR RIVER USING
MIKE-FLOOD
Vidyapriya.Va and Ramalingam. Mb
a
Research scholar, bPrincipal,Jerusalem College of Engineering,Velachery main
road,Pallikkaranai,Narayanpuram,Chennai-600 100, Tamil Nadu, India.
Mobile No-9952940236,Email id: priyavidya22@gmail.com,

ABSTRACT: The present paper aims to model the hydrodynamic transport of flood runoff located in Adayar
river in Chennai, Tamil Nadu from the Nandambakkam bridge up to the sea mouth of Adayar river through 1D
– 2D coupled MIKE FLOOD hydrodynamic models. Secondly, the study aims at finding the specific reason for
the occurrence of floods and the promising measure such as to study the impact of bed geometry of Adayar
river to be taken for different scenarios and concluded in suggesting site specific preventive measures for the
annual floods so that there are no severe flood events occurs in the coming future. The study compares the
result obtained from the model with the help of observed flood level taken during 2005 flood event. According
to the results from the model, the model was found satisfactory after validated.

IJSER
Keywords: Hydrodynamic models, MIKE-11,MIKE-21, MIKE- FLOOD,flood inundation simulations and DEM.

————————————————————

1. INTRODUCTION

Accurate representation of elevation data for river channels and floodplain are of great
importance for topographic mapping and disaster mitigation. High resolution digital terrain and
elevation models (DTM/DEMs) are an important component of this intelligence providing data on
surface and feature heights, landform variation, flooding extents and, indirectly flood behavior as per
Cobby et al. (2001).Airborne Laser Scanners (ALS) emit millions of high frequency laser beams that
interact with surface structures and penetrate vegetation to varying degree. The transmit and reflection
times of the laser pulses are recorded and subsequently used to derive the height of ground and above
ground features.The costlier LiDAR based DEMs are the best for its unique capability of accurate
topographical representation at 30 GSD accuracy. It is best suitable for accurate flood modeling,
channel network extraction and channel bed morphology analysis (Cook and Merwade 2009).

The historical records of several catastrophic flooding in Chennai since 1943,

1976,1985,2002,2005 and 2015 are due to heavy rain associated with cyclonic activity. The reasons for
the flooding are two- fold. Most of the existing waterways are silted and their flow channels and banks
are obstructed with encroachments and structures. Secondly several of the areas under tanks and their
anicuts have been developed as residential neighbourhoods over the years which indicate the growth of
urban population. Therefore it is high time to review and redesign drainage system to find out the best
solutions. Their was a study by Chennai Municipal corporation on Pre-Feasibility Study support for
Waterways Rehabilitation and Solid Waste Management in Chennai in association with DHV B.V.
Netherlands and with DHV India Pvt. Ltd. After analysis, they concluded with some rehabilitation
measures for Adayar river such as 1. Earth work excavation/ dredging by machineries for 2m depth
below mean sea level from 0.0 km to 0.5 km; 2. Earth work excavation by machineries to provide
gradient to the river bed and bed levels -2m below MSL at 4.2km to 0.5m above MSL at 7km;3. Earth

IJSER © 2016
http://www.ijser.org
International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 710
ISSN 2229-5518

work excavation by machinery to provide gradient to the river bed and bed levels -0.5m below MSL at
7km to 2m above MSL at 12.2km.This study has taken the suggested measures given for Adayar river
and incorporated through the hydrodynamic model.

2. STUDY AREA AND DATA USED

The study area covers Adayar watershed of 42.84 sq.km. It lies between the North Latitudes
13°1’8.513’’ N and 13°3’29.645’’N and East Longitudes 80° 11’9.106’’E and 80°15’54.819’’E. Figure
3.2 depicts the index map of the study area. The Chennai city is bounded by Thiruvallur district in the
north and west, Kancheepuram district in the south and Bay of Bengal in the east. The total length of
the Adayar river is 43 km.It originates at the confluence of Thiruneermalai that drains from the
upstream area of Chembarambakkam tank and crosses various bridges such as Nandambakkam,
Maraimalai Adigal, Kotturpuram and Thiru vi ka across the river and then ultimately reaches the sea.
The length of the Adayar river within the city limits is 12.2 km is considered for this study. The
maximum elevation is around 320 m above the mean sea level at Thiruneermalai, Pallavaram and
Kadapperi Hills and minimum elevation is around 5 to 15 m above the mean sea level. The study area
is a low-lying flat, slightly undulating terrain with a gentle slope of 3°-5° toward the North East
direction .The city enjoys the tropical climate with the annual rainfall of 1206.5 mm and the north east
monsoon contributes almost 60% of total rainfall in this semi-arid region. Most of the rainfall is
associated with clear synoptic systems of depressions and cyclones. It is characterized with seasonal
variations of winter season from November to February, summer season from March to mid-June and
rainy season from mid-June to October. The figure 1 shows the study area of the Adayar river.

IJSER
Figure 1: Study area of the Adayar river on the background of DEM

3. RAINFALL-RUNOFF MODEL USINGMIKE URBAN TIME-AREA METHOD

The Time-area (TA) rainfall-runoff model is used to derive the runoff hydrograph based on a
given excess rainfall hyetograph. In this method, the watershed is divided into a number of subareas
separated by isochrones; i.e. the isolines of equal travel time to the outlet. This process is known as
time area histogram. After plotting the time –area histogram the runoff hydrograph may be determined

IJSER © 2016
http://www.ijser.org
International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 711
ISSN 2229-5518

by the formula:

Where j = Time step number

Q = Runoff discharge
E = Excess rainfall intensity
A = Area bounded by the isochrones

The parameters required for estimation of runoff and flood depth are catchment data which comprises
of imperviousness, initial loss, reduction factor, time-area curve type , time of concentration and the
head loss, boundary data and the simulation data.

3.1 Runoff hydrograph for selected location across Adayar river

Hourly runoff at various locations on the rivers Adayar was simulated for a rainfall period
from 2nd December, 2005 to 4th December, 2005. Figure 2 (a) ,(b) and (c) present a plot of the derived
runoff simulations for the Adayar river. These simulated flows have been used as an input in the
hydrodynamic study for which results are presented in the text that follows.

IJSER

(a)

IJSER © 2016
http://www.ijser.org
International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 712
ISSN 2229-5518

IJSER
(b)

(c)

Figure no:2 Runoff hydrograph at (a) Nandambakkam bridge across Adayar river (b)
Maraimagal Adigal bridge across Adayar river ( c ) Thiru-vi-ka bridge across Adayar river

IJSER © 2016
http://www.ijser.org
International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 713
ISSN 2229-5518

4. HYDRODYNAMIC MODEL USING MIKEFLOOD

4.1 One dimensional hydrodynamic modeling using MIKE 11

MIKE-11 is a versatile engineering tool for modeling 1D hydrodynamic condition in rivers

(DHI, 2009). MIKE-11 HD solves the Saint-Venant equations to obtain the hydrodynamic state of the
river networks The hydrodynamic (HD) model is the nucleus of the MIKE-11 modeling system and
forms the basis for simulation of flood inundation. The HD model is capable of simulating 1D
unsteady flow in a network of rivers using hydrodynamic wave approach. The five basic input
parameters to be specified for the MIKE-11 HD setup are the river network layout, cross sections,
boundaries conditions, hydrodynamic parameters and simulation parameters. Before running the model
simulation, control parameters such as simulation period, simulation time step, data to be stored and
storage time were specified. There exists a versatile relationship between the time step and the
computational distance to define the Courant number which is widely considered to choose the time
step for the model simulation. The river network of the model is shown in the figure 3.

IJSER
Figure 3: River network of Adayar river inside the MIKE 11model

In order to achieve stable flood simulation the computational time step was brought to a value
of 3 seconds. The result of HD simulation consists of a time series of water level and discharges at
various points along the river system which can be viewed through MIKEVIEW through graphical
and animated interfaces.

4.2 Two-dimensional MIKE-21 HD model setup

MIKE-21 solves the full, time-dependent, non-linear equations of continuity and conservation
of momentum. The MIKE-21 resolves the solution using an implicit finite difference scheme of
second-order accuracy.

IJSER © 2016
http://www.ijser.org
International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 714
ISSN 2229-5518

MIKE-21 model input parameter is bathymetry or terrain elevation is the most important
input parameter, which contains the information regarding the elevations of the flood plain. The
Digital Elevation Model, obtained from LiDAR-DEM for the study area was processed to obtain
bathymetry as input for MIKE-21. The resolution of the input bathymetry was 10m x 10 m, so the
computational distance was 10m and the time step adopted was 3 seconds for different simulations.
The remaiming input parameters such as the flood plain roughness coefficient (Manning’s n) is kept as
0.038. Figure 4 shows the bathymetry used in the study area.

IJSER
Figure 4: Shows the bathymetry used in the study area

4.3 Hydrodynamic model simulation using MIKEFLOOD

The MIKE-11 river network was connected to the MIKE-21 bathymetry using the lateral link
option available in the MIKE-FLOOD (Kadem et al 2012). The river bank was dynamically linked
with the MIKE-21 grids using a cell-by-cell approach. Whenever the overtopping takes place from the
MIKE-11, the MIKE-21 calculates the discharge over each cell using weir formula. The inundation
extent and depth in the flood plain for the overtopping water is calculated by MIKE-21.Other
parameters for the left and right bank lateral links, such as momentum factor, weir coefficient and
depth tolerance factor are kept at their default values. The Manning’s n at all the link points was also
kept as 0.038 value. MIKE-FLOOD setup along with the river and both the lateral links is shown in
figure 5.Of the three curves in the figure, the central represents the Adayar River and the other two
above and below are the left and right links respectively, which connect banks of the river cells of
MIKE-11 with the bathymetry in MIKE-21. The simulation period for both MIKE-11 and MIKE-21
was kept the same and model computational time step was brought to a low value of 3 seconds in order
to keep the Courant Number (CR) less than or equal to 1 so as to achieve stable MIKE-FLOOD
simulation run without any errors.

5. RESULTS AND DISCUSSIONS

The flood inundation map by MIKE-FLOOD for two scenarios has been simulated and
validated. From the figure 5 (a) and (b) the maximum flood level reached is 6.5m whereas after
modification of the river bed level upto 1m there is a reduction of flood level up to 4.4m.This clearly
states the flood mitigation through altering the river bed level has gained beneficial to the society in
order to avoid flooding across the river. Since there is lack of observed data for calibration the model
couldn’t performe calibration but validation of the model have been done with the help of observed
water level during the flood event 2005( Ramalingam, M 2007). According to the model results, the
model validation was found satisfactory using R2 goodness of fit of 0.93 at Nandambakkam
bridge,0.96 for Maraimagal adigal bridge and 0.93 for Thiru-vi-ka bridge. .

IJSER © 2016
http://www.ijser.org
International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 715
ISSN 2229-5518

( a)

IJSER (b)
Figure: 5 Flood inundation map of Adayar river for two scenarios (a) Surveyed cross-section (b)
Modified cross-section up to 1.0 m
6. CONCLUSION

A 1D -2D coupled hydrodynamic model was developed for the Adayar river with the help of
the state-of-the-art MIKE FLOOD software package. The model was validated for the year 2005 with
little amount of available input data. The overall study helped in easily finding a easy solution to the
flood problem. Such modeling study can prove extremely beneficial for many more similar cases that
would help in mitigating the flood risk level.

ACKNOWLEDGEMENTS

We are grateful to thank Dr. M. Ramalingam former Director of IRS, Anna University for providing
LiDAR DEM of Adayar watershed and team members from DHI India for their valuable contribution
and guidance.

REFERENCES

Cobby, DM, Mason, DC & Davenport, IJ 2001,‘Image processing of airborne scanning laser altimetry
for improved river flood modeling’,ISPRS Journal of Photogrammetry and Remote
Sensing,vol.56,pp.121-138.

IJSER © 2016
http://www.ijser.org
International Journal Of Scientific & Engineering Research, Volume 7, Issue 7, July-2016 716
ISSN 2229-5518

Cook, A., Merwade, V., 2009, ‘Effects of topographic data, geometric configuration and modeling
approach on flood inundation mapping’, Journal of hydrology, vol 377(1-2),pp 131-142..

DHI, M., 2009. A Modelling System for Rivers and Channels. Reference Manual. DHI Software.

Kadam, P., Sen, D., 2012. Flood inundation simulation in Ajoy River using MIKE-FLOOD. ISH
Journal of Hydraulic Engineering, 18(2): 129- 141.

Ramalingam, M 2007, ‘Flood risk mapping of Chennai city and its suburbs using LIDAR data’,
Institute of Remote Sensing, State Remote Sensing Application Centre, Anna University, Chennai,
pp.(Report unpublished).

IJSER

IJSER © 2016
http://www.ijser.org