Flood drivers of the change in precipitation and extreme

Flood is a natural disaster which is
able to damage human beings, animals and other properties than other natural
disasters such as landslides, droughts, wind storms etc. So, the analysis of
risk of flood is very important to get many decisions as to mitigate above
problems. Suitable modelling approach can use as a solution of this risks(Heimhuber, Hannemann, & Rieger,
2015). Changes
in climatic variables, especially temperature and humidity, are probably the
main drivers of the change in precipitation and extreme hydrological risk.
Climate change can affect the intensity, frequency, duration and scope of risks
and the vulnerability of communities to disasters.(Ntajal, Lamptey, Mahamadou, &
Nyarko, 2017).There
is the flood which can occur in river and surrounding of the river basins when
the flowrate exceeds the capacity of the river, particularly at bends or
meanders on the path of the river. Occurrence of flood in Sri Lanka is mainly
due to excessive of rainfall during monsoon period(Consultancy, EngineeringCentral,
2014)   Therefore, this study is based on the
available data of the study area.

Recently, usage of HEC-HMS and HEC-RAS software
has been increased for flood analysis. Those can use to manage spatially
distributed data and the distributed basin model (Hashemyan, Khaleghi, & Kamyar,
2015). The
use of these hydrological and hydrological parameters of the river and river
basin can analyse the risk of flooding in selected river basin in Sri Lanka.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

The HEC model is used to simulate the
surface runoff of the catchment due to precipitation by interconnecting
hydrologic and hydraulic components. It is primarily appropriate to flood
simulations. In HEC-HMS, the river basin model includes different functions of
the precipitation-runoff process of the basin(Jamrussri & Toda, 2017).

An element of this application is
depicted a surface runoff. Each of the elements is assigned a variable in
HEC-HMS which defines the attribute of the element and mathematical
relationship of the actual processes and also the result of the modelling
process is the computation of stream flow hydrographs at the catchment outlet(Oleyiblo & Li, 2010).

 

 

1.     
Modelling of flood event

According
to past studies, the simulation of runoff events with
high hydraulic risk has posed many challenges for policymakers,
environmentalists and engineers around the world. Using 1-D modelling to
predict flood risk from different return period events or multiple land use and
climate change scenarios are common(Horritt & Bates, 2002).

It is
noticeable that the use of the Digital Elevation Model (DEM) in the creation of
flood models have reached an important role of the topographic and hydrological
analysis of basin data, since it represents a series of elevations in the basin
at regularly spaced intervals. This removes the assumption that the basin or
area is a flat surface without contours(Heimhuber et
al., 2015).

In case study on flood risk and flood prediction using GIS and
the model of hydrodynamic presented the possibility of using DEM controlled in
a GIS and translated into MIKE21. In the study, different scenarios were
checked out, and results were translated into the GIS environment for flood
visualization and analysis during a 100-year flood return period(Ntajal et al.,
2017). However, Jagadish Prasad Patraa, Rakesh Kumara
and Pankaj Manib pointed out that there was no real way to calibrate the
simulations from the modelling output, as flood and stage data for the floods
were rarely recorded and compared between the MIKE21 and MIKE1 results, the
first being an improvement of the last one(Prasad, Kumar, &
Mani, 2016).

In a research conducted by Sarawut
Jamrussri and Yuji Toda on the hydraulic models and GIS for the study of the
Mae Klong River in Thailand. Flow frequency analysis was used in the creation
of a flood risk map. The study also showed that the simulation results were
correctly presented in GIS and DTM format, using contour and height data from
the river point. Sarawut Jamrussri and Yuji Toda conclude their study by
suggesting that more studies be done in large basins, dividing them into
sub-basins and introducing the network link to integrate them to have a general
view of the basin. Runoff from floodplains, fluvial canals and artificial
structures are important factors in the study of the prediction of runoff flow patterns,
the researchers added. rainwater in upstream areas and not stable(Jamrussri & Toda, 2017).

 

2.1 Analysing methods

HMS uses
a project name as the identifier for a hydrological model. A HMS project must
have the following components before it can be executed: a basin model, a
climate model and control specifications. The characteristics of the basin model
and basin were created as a lower map file imported into the HMS from the data
derived from HEC-HMS for model simulation(Oleyiblo &
Li, 2010). The observed rainfall and discharge data were
used to create the climate model using the User Indicator Weight Method (UIWM),
and then the control specification template was created(Moya Quiroga,
Kure, Udo, & Mano, 2016). The control specifications determine the time
model for the simulation; its characteristics are: a start date and time, a
date and time of completion, and a calculation time step(Kawasaki et al.,
2017). To operate the system, the basin model, the
climate model and the control specifications were combined. The historical data
observed from precipitation stations representing each sub-basin and one
measuring station in the river basin and precipitation stations representing
each sub-basin and one measuring station in the basin were used to calibrate
the model. check. One-time step per hour was used for the simulation according
to the time interval of the observed data(Hashemyan et
al., 2015).

Figure 1. Model representation of
the Ravine Lan Couline watershed in HEC-HMS.

 

The HEC-HMS model for flood simulation uses a graphical interface to
build the semi-distributed basin model. For each sub-basin of the main basin,
the hydrological model is forced using a unit hydrograph(Bates & Roo, 2000). First, using the kriging
method, spatial rainfall distributions were generated from the time values
recorded in the rain gauge station located at the top of the river. Then, for
each sub-basin, rainfall series per hour were calculated. The Soil Conservation
Service curve number (SCS-CN)  method was
used to calculate the runoff volumes in the precipitation-runoff model(Hashemyan et al., 2015). Calculation of weighted
curve number (WCN) is shown by

Where, WCN is weighted curve number, Ai is area for ith land use
type and CNi is curve number for ith land use type(Plate, 2002).

The
HEC-RAS model was implemented using the cross sections to provide channel width
and bed elevations. These sections were extended on both sides of the channel
using DEM derived from LiDAR to provide a floodplain topography(Mancusi &
Abbate, 2017). The section was then described by 5-10 points
on each side of the canal coinciding with significant topographic features such
as slope failures. The elevation profile of the bed and examples of cross
sections used in the HEC-RAS model are given in following figure(Heimhuber et
al., 2015).

Figure 2.
Cross-sectional profiles of the river and its flood plains

The
boundary conditions of the model are a dynamic discharge imposed at the
upstream end of the section and an elevation of the surface of the water
imposed downstream. end, both provided by scene recorders and a nominal section
in the case of imposed discharge(Plate, 2002). Although the use of free surface elevation
measured at the downstream end means that boundary conditions and validation data
(travel time) are not completely independent, the effect of this situation has
been judged low(Horritt &
Bates, 2002). Flood wave propagation time remains a good
source of calibration data, as they are highly dependent on the calibration of
the model and are not significantly affected by the downstream boundary
conditions. The expected flood range was then calculated by re-projection of
the water levels of the cross sections in the high-resolution DEM(Logah, Amisigo,
Obuobie, & Kankam-yeboah, 2017).

Figure 3.
Floodplain zone map with different return
periods

In a research
conducted by Fowze the HEC-RAS model when presented with the
appropriate hydraulic and geometric data, calculates water-surface profiles.
The original reference for the method to determining the roughness coefficient
in reaches is Cowan method(Jamrussri & Toda, 2017) because it includes
several factors control the roughness coefficient. Then HEC-Geo RAS extension
was used to preparing and inputting geometric information about the reach that
these data are include flow path, left and right bank and cross sections that
in the form of new data layers was entered to HEC-RAS model. Then while
importing of the output hydrographs resulted from HEC-HMS and introducing the
roughness, channel convergence and divergence coefficients, HEC-RAS model was
run and the results of the hydraulic analysis and extracting of flood zones and
flood depth was done in ARCVIEW software and the floodplains was determined for
return periods of 10, 20 and 50 years(Consultancy, EngineeringCentral, 2014).

Figure
4. Flood hydrograph for various return periods

2.     
Conclusion

The HEC-HMS and
HEC-RAS model performs as well in terms of predicting inundated area when
calibrated on the inundated area for the other event as when calibrated on
flood risk of the river basin.

Acknowledgements

The author is grateful to the research
supervisor Dr. Kasun De Silva for the guidance and to the collegues for the
help.