ENGINEER - Vol. XXXXIII, No. 02, pp.

[28-37], 2010
© The Institution of Engineers, Sri Lanka

Event Based Modeling of a Watershed Using

HEC-HMS
H . K . N a n d a l a l a n d U . R . Ratnayake

Abstract: Modeling a watershed can be very much useful to quantify water resources for effective system
management. Rainfall runoff models are very much useful in forecasting floods, which will enable
taking mitigation measures. Kalu-Ganga River basin in Sri Lanka is subjected to frequent floods. This
paper presents a rainfall-runoff model developed for this basin using HEC-HMS lumped conceptual
hydrologic model. Two different models, one having four sub basins and the other having ten sub basins
were formulated. They were calibrated and verified using four historical flood events. Streamflow data
at three gauging stations along the river were used in the calibration and verification. The resulting
hydrographs at these three gauging stations were compared with the observed hydrographs based on
flood peaks, time to peak and the Nash Sutcliffe coefficient The results show the suitability of the HEC-
HMS software in the modeling of the Kalu-Ganga River basin. Further the results of the two models
indicate that there is no impact of the number of sub basins considered in the modeling of the basin on
the prediction of floods due to rainfalls.

Key Words: Watershed Modeling, Scale effect

1. Introduction

Hydrology as defined by Penman, 1961 is type of soil, rainfall and flow monitoring data).
the science that answers the question, "What In practice the availability and quality of these
happens to the rain?" Modeling a watershed data are often an issue one needs to cope with.
to answer this question is a very complex task Sometimes, one has to compromise the overall
involving collection of necessary data, selection modeling quality because of insufficient high-
of methods to analyze, availability of affordable resolution data for developing, calibrating, and
software and the knowledge of the watershed validating a model (Chu, 2009). Therefore, it is a
concerned. Modeling a natural event like run-off crucial task to model a river basin under limited
from a rainfall will help water resources managers data availability.
to mitigate the consequences of floods in
advance. Modern day technology and tools help Hydrological models are being developed and
engineers to model natural environment. Tools applied in increasing number and variety.
like Geographical Information Systems (GIS) All hydrological models are simplified
and watershed models contribute a lot for the representations of the real world. Models can
modeling of natural events like floods. Advances be either physical (e.g. laboratory scale models),
in technology of computers, availability of data electrical analogue or mathematical. The physical
and necessary software also make modeling an and analogue models have been very important
easy task. Singh and Woolhiser (2002) provided in the past. However, the mathematical group of
a historical perspective of watershed modeling, models is by far the most easily and universally
a short outline of currently used models, and applicable, the most widespreaded and the one
reflects on new developments. However, with the most rapicl development with regard
modeling a watershed not only needs necessary to scientific bases and application (Refsgaard,
software, data and models but also it needs a 1996).
sound knowledge of the particular watershed. It
is important to analyze watershed and the data
available in detail before applying a model.
Eng. (Mrs.) H.K.Nandalal, BSc Eng, MSc, MlE(Sri Lanka),
CEng, Lecturer, Department of Civil Engineering, University
Watershed hydrologic modeling and the of Peradeniya,
associated model calibration and verification Eng.(Dr.) U. R. Ratmayake, BSC, MSc, PhD, MIE(Sri
require a large amount of spatial and temporal Lanka), CEng, Senior Lecturer, Department of Civil
Engineering, University of Peradeniya,
data (e.g., topography, land use, land covers,

ENGINEER 28
T h e t w o classical types of hydrological m o d e l s experiences an a v e r a g e a n n u a l rainfall of 4000
are the deterministic a n d the stochastic. T h e m m . It varies from 2800 m m in lower reaches to
deterministic m o d e l s use data available as it is 5300 m m in higher elevations. Geographically it
while the stochastic m o d e l s use the statistical lies b e t w e e n the 6.32°N a n d 6.90°N, a n d 79.90°E
n a t u r e of available d a t a to predict rainfall runoff. a n d 80.75°E as p e r WGS84 coordinate s y s t e m
T h e deterministic m o d e l s can b e classified a n d flows from a h e i g h t of a b o u t 2,250 m MSL.
according to w h e t h e r the m o d e l gives a l u m p e d
or distributed description of the considered Tributaries of the K a l u - G a n g a River collect w a t e r
area, a n d w h e t h e r the description of the from Sri-Pada a n d Sinharaja virgin rainforests
hydrological processes is empirical, conceptual, a n d quickly d r a i n from its u p p e r catchments.
or m o r e physically-based. Therefore, m o d e r n Rainfalls o n the Sri-Pada rainforest flow t h r o u g h
d a y hydrological m o d e l s can be classified into a length of a b o u t 36 k m a n d from an elevation of
empirical m o d e l s , l u m p e d conceptual m o d e l s a b o u t 2,250 m MSL to 14 m MSL at R a t n a p u r a
a n d distributed physically-based m o d e l s . O u t t o w n . There o n w a r d s the Kalu-Ganga River
of these m o d e l s empirical m o d e l s n e e d accurate travels t h r o u g h a c o m p a r a t i v e l y flat terrain
rainfall a n d runoff data for calibration w h i l e of 14 m MSL to sea level t h r o u g h a distance of
distributed physically based m o d e l s n e e d a large a b o u t 70 k m collecting a large a m o u n t of w a t e r
n u m b e r of spatial characteristics of the area a n d flowing from the other tributaries o n the w a y
meteorological d a t a to calculate the runoff for a m a k i n g the lower reaches of the river p r o n e to
given rainfall. frequent floods.

L u m p e d conceptual m o d e l s need m o d e r a t e l y
accurate rainfall a n d runoff d a t a a n d a v e r a g e
physical characteristics of the area concerned.
P a r a m e t e r s of these m o d e l s can b e calibrated
a n d verified w i t h historical data available. Most
of the hydrological m o d e l s available today c o m e
u n d e r this category. T h e H y d r o l o g i c Engineering
C e n t e r ' s Hydrologic M o d e l i n g System, H E C -
H M S (USACE-HEC 2006) also falls into this
category.

M o d e l i n g also can b e event based or c o n t i n u o u s .

Event b a s e d calculations n e e d initial conditions
while c o n t i n u o u s m o d e l s need data of soil t y p e to Figure 1. K a l u - G a n g a River n e t w o r k o n D E M
calculate soil moisture content a n d a t m o s p h e r i c
d a t a as well to calculate e v a p o r a t i o n losses.
2. Data
T h e goal of the c u r r e n t s t u d y is to verify the Terrain data
applicability of the H y d r o l o g i c E n g i n e e r i n g
Digital Elevation Model (DEM) d a t a of the river
Center's H y d r o l o g i c M o d e l i n g System, H E C -
basin is a r e q u i r e m e n t to apply H E C - G e o H M S
H M S , o n e v e n t based m o d e l i n g u n d e r t w o
(USACE-HEC-GeoHMS, 2006) m o d e l to d e v e l o p
scenarios of s u b basin scale for Kalu-Ganga
the basin m o d e l . D E M d a t a can b e obtained from
River basin located in the S o u t h West slopes
on-line sources a n d also can b e d e v e l o p e d u s i n g
of Sri Lanka. Data for four rainfall events w e r e
digital c o n t o u r m a p s . H o w e v e r , d e v e l o p i n g
used to calibrate a n d validate the m o d e l .
a D E M u s i n g c o n t o u r m a p s is a n e x p e n s i v e
p r o c e d u r e since it n e e d s fine c o n t o u r m a p s in
Study Area the scale of at least 1:10,000. Availability a n d
T h e Kalu-Ganga River w a t e r s h e d covers 2658 applicability of freely available on-line d a t a for
k m 2 a n d is d o m i n a n t by forest, residential a n d the Kalu-Ganga River basin has been discussed
agricultural c r o p l a n d land use types. Kalu-Ganga in N a n d a l a l a n d R a t n a y a k e (2008). For this
River is still an u n t a m e d river c o m p a r e d to the s t u d y the Shuttle Radar T o p o g r a p h y Mission
other rivers in the c o u n t r y t h o u g h it c a p t u r e s (SRTM) DEM (Jarvis et al., 2008) d a t a w i t h 90m
a large a m o u n t of rainfall a n d d r a i n s freely to grid resolution, w h i c h is freely available on-line
the O c e a n at Kalutara (Figure 1). T h e c a t c h m e n t is used.

29 ENGINEER H
Rainfall data
Daily rainfall records from 01st January 1991 to
31st December 1994 of thirteen rainfall gauging
stations lying in and around Kalu-Ganga River
basin were taken from the Department of
Meteorology, Sri Lanka and used in this model.
Aggregated rainfall for the basin is calculated
using the Thiessen polygon method as shown
in Figure 2. Four high rainfall events shown in
Table 1 were selected out of the rainfall data series
to calibrate and validate the model for the basin.

The event in November 1992 was used to Figure 3. River Basin model with 4 sub catchments
calibrate the model while the others were used
to validate the model. model was developed by HEC-GeoHMS using
the DEM model of the study area clipped from
the SRTM DEM (Jarvis et al., 2008) data with 90
m grid resolution.

3.1 HEC-GeoHMS 1.1

HEC-GeoHMS is a geospatial hydrology toolkit
for engineers with limited GIS experience. It is
an extension package used in ArcView software
and had been used extensively in modeling
watersheds; two of recent applications are Bakir
and Xingnan (2008) and Khakbaz et.al (2009). In
this study, HEC-GeoHMS is used to derive river
network of the basin and to delineate sub basins
and the basin from the DEM of the Kalu-Ganga
Figure 2. Rainfall gauging stations and the developed River basin.
Thiessen polygons
When the number of sub basins increases, the
Table 1. Rainfall events selected spatial variation of the characteristics of the
basin can be included efficiently in the model.
Event Time period To study that effect two types of basin models
1992 November 13 days were developed, one with 4 sub basins and the
1993 May 26 days other with 10 sub basins as shown in Figures 3
1993 October 17 days and 4, respectively.
1994 May 34 days
These basin models were imported to the HEC-
Discharge data HMS where there are several options available
Discharge data are very much important in the to derive rainfall runoff accounting for loss,
calibration and validation process of the model. transformation, and base flow calculation from
Daily discharge data at three main gauging a watershed. Simplest models with least number
stations Ratnapura, Ellagawa and Putupaula of parameters were selected because the detailed
along Kalu-Ganga River were obtained from data such as soil type, land cover, land use and
the Department of Irrigation of Sri Lanka for rainfall runoff data taken at smaller time intervals
the same period as for rainfall data to use in this are not available. Calibrating many parameters
study. for increased accuracy cannot be justified under
this condition and same accuracy in calibrating
is achieved with least number of parameters.
3. Model Development

In the process of model development, at first 3.2 HEC-HMS

the basin model has to be developed either by HEC-HMS (version 3.01) is hydrologic modeling
HEC-GeoHMS with DEM or by feeding data software that includes many well-known and
into HEC-HMS itself. In this study the basin well-applicable hydrologic methods to simulate

ENGINEER 30
rainfall-runoff process in river basins (USACE- The underlying concept of the initial and
HEC 2006). It consists of several models for constant-rate loss model is that the maximum
potential rate of precipitation loss is constant
throughout an event after the initial loss.
These two parameters can be calibrated
with the available gauged runoff data.

3.2.2 Modeling Direct Runoff

This process refers to the transformation of
precipitation excess into point runoff. HEC-
HMS allows calibrating a few different models
including, at the event based level:

- Clark's model - Snyder's model

- SCS-UH model - Kinematic wave model
Figure 4. River Basin model with 10 sub catchments
Clark's model is a synthetic unit hydrograph
calculation of losses and runoff due to a single that utilizes an instantaneous unit hydrograph
rainfall event or aconrinuous rainfall. Application (Clark, 1945). This widely used method is often
of HEC-HMS for modeling watersheds is called the time-area method and has appeared
extensive and some of the recent applications in several computer programs for hydrograph
are Knebla et.al. (2003), Al-Sabhan et.al. (2003), analysis. Clark's model is used in this study
Neary et.al. (2004), Pingel et.al. (2005) and Chu since it needs a small number of parameters.
and Steirunan (2009).
Application of the Clark model requires
3.2.1 Modeling Losses calibration of only two parameters:
During a given rainfall, water is continually being
abstracted to moisten the upper levels of the soil - Time of concentration of watershed tc.
surface; however, this infiltration is only one of - Storage coefficient, R.
many continuous abstractions. Rainfall is also
intercepted by trees, plants, and roof surfaces, Both of these parameters can be estimated via
and at the same time is evaporated (Viessman calibration if gauged precipitation and stream
and Lewis, 2003). Therefore, losses can be mainly flow data are available. Though R has units of
described as initial and continuous. HEC-HMS time, there is only a qualitative meaning for it
program considers that all land and water in a in the physical sense. Clark (1945) indicated that
watershed can be categorized as either directly- R can be computed as the flow at the inflection
connected impervious surface or pervious point on the falling limb of the hydrograph
surface. Directly-connected impervious surface divided by the time derivative of flow.
in a watershed is that portion of the watershed
for which all contributing precipitation runs off, 3.2.3 Modeling Baseflow
with no infiltration, evaporation, or other volume
Baseflow is the sustained or "fair-weather"
losses. Precipitation on the pervious surfaces is
runoff of prior precipitation that was stored
subjected to losses. The models available for the
temporarily in the watershed, plus the delayed
calculation of cumulative losses are:
subsurface runoff from the current storm. The
program includes three alternative models of
- the initial and constant-rate model
baseflow:
- the deficit and constant-rate model
- the SCS curve number loss model and - Constant, monthly-varying value
- the Green and Ampt loss model - Exponential recession model
Out of these models the initial and constant - Linear-reservoir volume accounting model
model was selected for this study which
Exponential recession model was selected for
needs only two parameters. The initial and
this study because it was found to be the one
constant-rate model includes one parameter
fitted well for the Kalu-Ganga River basin. The
(the constant rate) and one initial condition
parameters of this model include the initial flow,
(the initial loss). These represent physical
the recession ratio, and the threshold flow. The
properties of the watershed soils, land use and
initial flow is the likely average flow that would
the antecedent condition (USAE-HEC 2006).

31 ENGINEER I
occur at the start of the storm runoff. For frequent Univariate Gradient search algorithm was
events, the initial flow might be the average selected for this study. A variety of objective
annual flow in the channel (USAE-HEC, 2006). functions are provided to measure the goodness
The recession constant, k, depends upon the of fit between the simulated and observed runoff
source of baseflow. The recession constant can in different ways such as peak weighted RMS
be estimated if gauged flow data are available. error, percent error peak, percent error volume,
sum absolute residuals, sum squared residuals,
3.2.4 Modeling Channel flow and time-weighted error (USACE-HEC, 2006).
Channel flow model or routing model is used to This study used the peak weighted RMS error as
model flow from upper catchment to the outlet the objective function.
of downstream catchment. The routing models
available in HEC-HMS include: 3.4 Comparison of the two basin models
Numerous goodness of fit statistical criteria
- Lag - Muskingum - Modified Puis, are proposed in the literature for evaluating
- Kinematic-wave - Muskingum Cunge hydrological modeling results. In this study the
Nash-Sutcliffe index (Nash and Sutcliffe, 1970)
Reaches of the Kalu-Ganga River behave is used as the criteria to compare the HEC-
differently; therefore both lag model and HMS model performance under two scenarios
Muskingum model are used; lag model is used of basin models mentioned in section 3.1. The
for steeper reaches and Muskingum model is ASCE Watershed Management Committee
used for reaches of gradual slope near coast. The (ASCE 1993) recommends the Nash-Sutcliffe
lag model is the simplest routing model and it index for evaluation of rainfall runoff models.
Merz and Bloschl (2004) used the index in the
has only one parameter. Outflow hydrograph
calibration and verification of catchment model
in it is simply the inflow hydrograph, but with
parameters. Kalin et al. (2003) used the index
all ordinates translated (lagged in time) by a
as a goodness-of-fit indicator for a storm event
specified duration. The flows are not attenuated, model. It is also widely used with hydrological
so the shape is not changed. This model is widely models (Roux and Dartus, 2006; Norbiato et al.,
used, specially in urban drainage channels 2008; Ao et al, 2006; Moreda et al, 2006; Hogue et
(Pilgrim and Cordery, 1993). Muskingum model al, 2006). The use of the index for a wide variety
needs calibration of two parameters; K, travel of model types indicates its flexibility as a
time of the flood wave through routing reach; goodness-of-fit statistic. When the Nash Sutcliffe
and X, dimensionless weight which corresponds model efficiency coefficient is between 0 and 1,
to the attenuation of the flood wave as it moves the model does better than simply forecasting.
through the reach. However, it was subjected The closer the Nash Sutcliffe model efficiency
to discussion upon applicability of this model coefficient to one, the better the performance of
for flood wave propagation where there are the model.
large flat flood plains (Dooge et al., 1982, Tung,
1985 and Yoon and Padmanabhan, 1993).
Nash Sutcliffe model efficiency coefficient is
defined as: '
3.2.5 Model Calibrations
Parameters for each model should be entered
as input to the model to obtain the simulated
runoff hydrographs. Some of the parameter may
be estimated by observation and measurements
H («-&)'
of stream and basin characteristics, but some of Where, Q' 0 is observed discharge,
them can only be estimated through trial and error Q'm is modeled discharge, and Q is Q

calibration for best fit. The model parameters are average observed discharge at time t.
usually calibrated, i.e. in the presence of rainfall
and runoff data the optimum parameters are McCuen etal (2006) indicated that the Nash
found as a result of a systematic search process Sutcliffe model efficiency coefficient is a good
that yields the best fit between the observed performance criterion indicator as long as the
runoff and the computed runoff. In HEC-HMS sample size is moderate (i.e., number of ordinates
two different search algorithms (Nelder and are more than 12). In this study the sample size
Mead search algorithm and Univariate Gradient (i.e. the number of days) was always more
search algorithm) are provided that moves from than twelve and the time interval was one day
the initial estimates to the final best estimates. making it very much applicable for this study.

ENGINEER 32
Table 2 Parameters used in the two basin models shown in Figures 3 and 4

Basin Subbasin Loss -initial and constant

Initial Constant Impervious Transform - Clarks UH Baseflow -Recession
Time of Storage Initial Initial Recession Threshold Ratio
model loss
mm
rate m m / h % concentration
h
Coefficient
h
type discharge
3
m /s
Constant type to
peak
Ellegawa 1 0.00 1 72 48 Discharge 65.00 0.90 Ratio to Peak 0.05
4
Horana 2 0.30 1 72 48 Discharge 25.00 0.80 Ratio to Peak 0.20
basin
Sinharaja 5 3.00 1 96 96 Discharge 30.00 0.80 Ratio to Peak 0.20
model
Sri Pa da 1 0.20 2 24 48 Discharge 15.00 0.80 Ratio to Peak 0.40
Agalawaththa 3 0.90 1 24 48 Discharge 10.00 0.60 Ratio to Peak 0.20
Ayagama 2 0.05 2 96 72 Discharge 15.00 1.00 Ratio to Peak 0.10
Balangoda 3 0.30 10 48 48 Discharge 14.00 0.80 Ratio to Peak 0.20
Horana 3 0.20 1 48 24 Discharge 4.00 1.00 Ratio to Peak 0.20
10
Kalawana 2 0.90 2 36 72 Discharge 4.00 0.60 Ratio to Peak 0.20
basin
Kuru wita 2 0.05 1 24 24 Discharge 20.00 1.00 Ratio to Peak 0.10
model
Malwala 5 0.20 2 24 36 Discharge 14.00 0.80 Ratio to Peak 0.20
Rakwana 2 0.05 2 24 36 Discharge 12.00 1.00 Ratio to Peak 0.20
Sinharaja 1 2 2.50 2 36 72 Discharge 10.00 0.60 Ratio to Peak 0.20
Sinharaja 2 2 2.50 1 36 72 Discharge 10.00 0.60 Ratio to Peak 0.20

Table 3 Parameters used in for channel flow (reaches) of the two basin models shown in Figures 3 and 4

Basin Model
4 basin model Reach
ElIagawaToPutupaula
RatnapuraToEllagawa
Model
Muskingum
Muskingum
Muskingum K h Muskingum X
12
48
0.4
0.5
Lag min
SinharajaToPutupaula Muskingum 12 0.1
10 basin model MReach3 Muskingum 36 0.1
MReach4 Muskingum 24 0.1
MReach5 Muskingum 12 0.1
MReachl Lag 240
MReach2 Lag 720
MsReachl Lag 2160
Ms Reach 2 Lag 720
4. A n a l y s i s a n d Results basin. T h e n these p a r a m e t e r s w e r e o p t i m i z e d
using t h e calibration process in the H E C - H M S
4.1 Calibration and Verification m o d e l . O p t i m i z a t i o n of p a r a m e t e r s of the s u b
F o u r selected s t o r m e v e n t s are u s e d for calibration basins a b o v e the R a t n a p u r a g a u g i n g station w e r e
a n d verification of b o t h basin m o d e l s . For the c o n d u c t e d first u s i n g the d i s c h a r g e d a t a of the
calibration a n d o p t i m i z a t i o n of p a r a m e t e r R a t n a p u r a g a u g i n g station. N e x t o p t i m i z a t i o n
v a l u e s the rainfall e v e n t occurred in N o v e m b e r of the p a r a m e t e r s of the s u b basins a n d
1992 (Table N o . l ) w a s u s e d w h i l e the other three reaches a b o v e Ellagawa a n d b e l o w R a t n a p u r a
e v e n t s w e r e u s e d for verification. S u b s e q u e n t l y , g a u g i n g stations w e r e c o n d u c t e d . Finally, the
for a c o m p a r i s o n , calibration of the m o d e l s w a s p a r a m e t e r s of the s u b basins a n d reaches a b o v e
c a r r i e d o u t based o n t h e rainfall e v e n t occurred t h e last g a u g i n g station at P u t u p a u l a a n d below
in M a y 1993 a n d the resulting m o d e l s w e r e Ellagawa g a u g i n g station w e r e p e r f o r m e d .
verified u s i n g the other three rainfall e v e n t s Tables 2 a n d 3 p r e s e n t the p a r a m e t e r s resulted
g i v e n in Table 1. for s u b basins a n d reaches for 4 basin m o d e l a n d
10 basin m o d e l after o p t i m i z a t i o n respectively.
T h e p a r a m e t e r s of the m o d e l s w e r e first a s s u m e d
w i t h the k n o w l e d g e of the Kalu-Ganga River

15-W«Y-W 12 .MAY .84 M-MAYAI 5JUN*« 1I«JUIVM IS+MY-W K-WIY-W £HI»Y-W i-Jm*i
Figure 5a. Rainfall runoff at Ratnapura for May 1994 Figure 6a. Rainfall runoff at Ratnapura for May 1994
rainfall event for 4 basin model rainfall event for 10 basin model

L5-MAY-34 Z2-MAY-94 • 29-MAY-S4 5-JUII-94 12-JUN-M

KMif^* 22-MAY-M 29-MIY«4 S-AJN-W 12*JUN-M
Figure 5b. Rainfall runoff at Ellagawa for May 1994 Figure 6b. Rainfall runoff at Ellagawa for May 1994
rainfall event for 4 basin model rainfall event for 10 basin model

1S4J»Y-W 22-U»Y-84 29JJ«Y-9J S-JUN-M IJJIIN*! TS-WIY-94 22-MTY-M 29-MTY<04 SUUN«4 UOUN-W

Figure 5c. Rainfall runoff at Putupaula for May 1994 Figure 6c. Rainfall runoff at Putupaulaa for May 1994
rainfall event for 4 basin model rainfall event for 10 basin model

B ENGINEER 34
4.2 Comparison of the two basin models 5. Discussion and Conclusions
Hydrographs for the three gauging stations,
Ratnapura, Ellagawa and Putupaula (Figure 5.1 Discussion
1) for 1994 May rainfall event for the two Figures 5a to 5c for the 4 basin model and
sub basin models are given in Figures 5a to Figures 6a to 6c for the 10 basin model indicate
5c for 4 basin model and 6a to 6c for 10 basin that the overall performance of these models as
model for graphical evaluation of the model very satisfactory since the model resulted peak
performance for both scenarios of sub basin level. runoffs coincide with the observed peak runoffs
for all three gauging stations. There were no
From thegraphsitisevidentthatthepeak discharge much differences between the two scenarios of
agrees with observed value. The time to peak also the basin models used at three gauging stations.
agrees well with the observed value in both basin
models and for all the three gauging stations. The model performance for both the 4 basin
model and the 10 basin model can be accepted
Table 4 Nash Sutcliffe model efficiency coefficient well as observed in Tables 4 and 5. In the 4 basin
for observed gauging points for the 4 basin model model the model performance was very good at
Putupaula and the 10 basin model performed
well at Ratnapura and Ellagawa since Nash
Event -Ratnapura Ellagawa Putupaula Sutcliffe coefficient is close to one in these cases.
1993 May 0.900334 0.949387 0.858501
1993 Oct 0.887190 0.961047 0.771844
Table 6 Nash Sutcliffe model efficiency coefficient
1994 May 0.835777 0.964079 0.940262 for observed gauging points for the 4 basin model
0.836231 0.896588 0.934740 with parameters optimized using 1993 May rainfall
event
Table 5 Nash Sutcliffe model efficiency coefficient
for observed gauging points for the 10 basin model Event Ratnapura Ellagawa Putupaula
1992 Nov 0.880166 0.732225 0.845788
Event Ratnapura Ellagawa Putupaula 1993 Oct 0.916842 0.865624 0.816133
1993 May 0.920920 0.942308 0.700043 1994 May 0.930893 0.907112 0.933217
1993 Oct 0.932305 0.942336 0.803982 0.9093 0.834987 0.865046
1994 Mav 0.912886 0.959965 0.957709
0.924131 0.922629 0.860085 Table 7 Nash Sutcliffe model efficiency coefficient
for observed gauging points for the 10 basin model
with parameters optimized using 1 9 9 3 May rainfall
The two basin models were compared using event
Nash Sutcliffe model efficiency coefficient
calculated for computed discharge and the
Event Ratnapura Ellagawa Putupaula
observed discharge at the three gauging stations,
Ratnapura, Ellagawa and Putupaula. Table 1992 Nov 0.915299 0.773556 0.888731
4 shows the Nash Sutcliffe model efficiency 1993 Oct 0.933920 0.894022 0.813787
coefficient yielded for basin model with 4 sub 1994 May 0.894805 0.905383 0.945018
basins while Table 5 presents them for the basin 0.914674 0.857654 0.882512
model with 10 sub basins for the three rainfall
events considered. Last rows in both tables The Nash Sutcliffe model efficiency coefficient
represent the average value for all three rainfall values yielded for the two models with the
events- parameters optimized using 1993 May rainfall
Table 6 and Table 7 give the Nash Sutcliffe model event show that there is no much difference
efficiency coefficient for observed gauging points between the outputs obtained from the two
for the 4 basin model and 10 basin model with calibrations.
parameters optimized using 1993 May rainfall
event. With respect to runoff simulation at Putupaula
gauging station the 4 basin model performs
better than the other as indicated by the Nash
Sutcliffe model efficiency coefficient of 0.934740.
However, when the other two gauging stations

35 ENGINEER
are considered the 10 basin model performs Acknowledgements
a little better. Though in general rainfall
representation and the other sub basin scale Authors would like to acknowledge the
parameters should improve when the model has University Grant Commission for providing
more sub basins, it does not happen in case of the necessary funds for carrying out this research.
Putupaula gauging station. This may be due to
the different spatial distribution of rainfall in the
Reference
Sinharaja watershed which is very large. If the
gauging station was situated before Sinharaja 1. Al-Sabhan, W., Mulligan, M., Blackburn, G.A.,
runoff enters the Kalu-Ganga River a better (2003). A real-time hydrological model for
rainfall runoff simulation would have resulted flood prediction using GIS and the WWW,
at Putupaula gauging station. Computers, Environment and Urban Systems,
27 (2003) 9-32.

Conclusions 2. Ao, T., Ishidaira, H., Takeuchi, K., Kiem, A.S.,

Yoshitari, J., Fukami, K., Magome, J. (2006).
The results based on the Nash Sutcliffe model "Relating BTOPMC model parameters to
efficiency coefficient and the graphical evaluation physical features of MOPEX basins", journal
show that the HEC-HMS model is well suited of Hydrology, 320, 84-102.
for the simulation of rainfall runoff of the Kalu-
Ganga River basin, regardless of the sub-basin's 3. ASCE Task Committee on Definition of Criteria
scale. Clark unit hydrograph method available in for Evaluation of Watershed
HEC-HMS model to transform rainfall into runoff
is suitable to model the Kalu-Ganga River basin. 4. Models of the Watershed Management,
However, for channel routing both lag time and Irrigation, and Drainage Division (ASCE).
Muskingum channel flow routing methods have (1993). "Criteria for evaluation of watershed
to be used depending on the characteristics ofthe models." Journal of Irrigation and Drainage
channel. Birkhead and James (2002) mentioned Engineering, 119(3), 429-442.
that Muskingum procedures account explicitly
5. Bakir, M., Xingnan, Z., (2008) " GIS-Based
for channel storage only, and not total storage
Hydrological Modeling: A Comparative
along a river reach which may include lateral
Study of HEC-HMS and The XINANJ1ANG
inflows (tributary or diffuse inputs) or outflows,
Model", Proceedings of Twelfth International
losses (recharge of groundwater aquifers or
Water Technology Conference, IWTC12 2008
evapotranspiration), and temporal changes in
Alexandria, Egypt, 855-865,
bank storage. Therefore, the model sometimes
generates unrealistic values for Muskingum 6. Birkhead, A.L., and James CS., (2002)
parameter during the calibration. However, Muskingum river routing with dynamic bank
though the lower reaches of the Kalu-Ganga storage, Journal of Hydrologic Engineering,
River bear such characteristics the Muskingum 264,113-132.
method was able to successfully model the
lower reaches of the Kalu-Ganga River. The 7. Chu, X., and Steinman, A. (2009). "Event and
Lag time channel flow routing method which is Continuous Modeling with HEC„HMS."
suitable to steeper channel lengths was found to Journal of Irrigationand Drainage Engineering,
be applicable for the upper reaches of the river. 135(1), 119-124.

8. Clark, C O . (1945). "Storage and the unit

Since the difference between the results obtained hydrograph." Transactions, ASCE, 110, 1419-
from the 4 basin model and the 10 basin model 1446.
is negligible it can be concluded that there is no
scale effect (that is the number of sub basins) 9. Dooge, J.C.I., Strupczewski, W.B.,
in modeling the Kalu-Ganga River basin using Napiorkowski, ].]., 1982. Hydrodynamic
HEC-HMS for flood prediction. The model derivation of storage parameters of the
parameters resulted from the calibrations carried Muskingum model. Journal of Hydrology 54,
out using two rainfall events were observed to 371- 387.
be very similar.

ENGINEER
36
10. Jarvis A., Reuter, H.I., Nelson, A. and E. 20. Penman, H.L. (1961). "Weather, plant and soil
Guevara (2008) Hole-filled seamless SRTM factors in hydrology" Weather, 16, 207-219.
data V4, International Centre for Tropical
21. Near>', V. S., Habib, E., Fleming, M., (2004),
Agriculture (CIAT), available at http://srtm.
Hydrologic Modeling with NEXRAD
csi.cgiar.org.
Precipitation in Middle Tennessee, Journal of
11. Horgue, T.S., Gupta, H., Sorooshian, S. (2006). Hydrologic Engineering, 9(5), 339-349.
"A 'User-Friendly' approach to parameter
22. Norbiato D., Borga, M., Esposti, S. D., Gaume,
estimation in hydrologic models" Journal of
E., Anquetin, S., (2008). "Flash flood warning
Hydrology, 320, 202-217.
based on rainfall thresholds and soil moisture
12. Kalin, L., Govindaraju, R. S., and Hantush, conditions: an assessment for gauged and
M. M., (2003) "Effect of geomorp ho logical ungauged basins", Journal of Hydrology, 362,
resolution on modeling of runoff hydrograph 274-290.
and sedimentograph over small watersheds."
23. Pingel, N.P.E., Jones, C , Ford, D., (2005),
Journal of Hydrology, 276, 89-111.
Estimating Forecast Lead Time, Natural
13. Khakbaz, B., Imam, B., Hsu, K., Sorooshian, S., Hazards Review, 6(2), 60-66.
(2009). From lumped to distributed via semi-
24. RefsgaardJ.C, (1996) Terminology, Modelling,
distributed: Calibration strategies for semi-
Protocol and Clasification of Hydrological
distributed hydrologic models, Journal of
Model Codes, in Abbott, M.B. and Refsgaard,
Hydrology'., doi:10.1016/j.jhydrol.2009.02.021.
J.C. (eds.) Distributed hydrological Modelling
14. Knebla, M.R., Yanga, Z.L., Hutchisonb, K., (Kluwer Academic Publishers), The
Maidment D.R., (2005), Regional scale flood Netherlands, 17-39.
modeling using NEXRAD rainfall, GIS, and
25. Roux, R., and Dartus, D (2006). "Use of
HEC-HMS/RAS: a case study for the San
parameter optimization to estimate a flood
Antonio River Basin Summer 2002 storm
wave: Potential application to remote sensing
event, Journal of Environmental Management,
of rivers" Journal of Hydrology, 328, 258-266.
75 (2005) 325-336.
26. Singh VP, Woolhiser AD. (2002). Mathematical
15. McCuen, R.H., Knight, Z., Cutter, A.G. (2006).
modeling of watershed hydrology. Journal of
"Evaluation of the Nash-Sutcliffe Efficiency
Hydrologic Engineering 7(4): 270-292.
Index", Journal of Hydrologic Engineering,
11(6), 597-602. 27. Tung, Y.K., 1985. River flood routing by
nonlinear Muskingum method. J. Hydraul.
16. Merz,R.,andBloschl,G.,(2004)"Regionalization
Engng. I l l (12), 1447-1460.
of catchment model parameters." Journal of
Hydrology, 287,95-123. 28. United States Army Corps of Engineers,
Hydrologic Engineering Center (USACE_
17. Moreda, F., Koren, V., Reed, S., Smith, M. (2006)
HEC). (2006). Hydrologic modeling system
"Parameterization of distributed hydrological
HEC-HMS technical user's manual, Davis,
models: learning from the experience of
California.
lumped modeling" Journal of Hydrology, 320,
218-237. 29. United States Army Corps of Engineers,
Hydrologic Engineering Center (USACE_
18. Nandalal, H.K. and Ratnayake, U.R. (2008),
HEC-GeoHMS). (2006). Geospatial Hydrologic
"Comparison of a river network delineated
Modeling Extension, User's Manual, Davis,
from different digital elevation models
California.
available in public domain", Proceedings, 29th
Asian Conference on Remote Sensing, CD_ 30. Yoon, J., Padmanabhan, G., 1993. Parameter
ROM, Colombo, Sri Lanka. estimation of linear and nonlinear Muskingum
models. J. Water Resour. Planning Mgmt
19. Nash, J.E., and Sutcliffe, J.V. (1970) River
119(5), 600-610.
flow forecasting through conceptual models,
part-1- A discussion of principles. Journal of
Hydrology, 10(3), 282-290.

37
ENGINEER