Environmental Challenges 17 (2024) 101017

Contents lists available at ScienceDirect

Environmental Challenges
journal homepage: www.elsevier.com/locate/envc

Hydrological modeling using HEC-HMS model, case of Tikur Wuha River

Basin, Rift Valley River Basin, Ethiopia
Jerjera Ulu Guduru a,* , Ayatullah Shis Mohammed b
a
Department of Hydraulics and Water Resources Engineering, Haramaya Institute of Technology (HiT), Haramaya University, Dire Dawa, P.O. Box 138, Ethiopia
b
Department of Hydraulics and Water Resources Engineering, Engineering Unit, Haramaya University, Dire Dawa, P.O. Box 138, Ethiopia

A R T I C L E I N F O A B S T R A C T

Keywords: Modeling rainfall-runoff is widely recognized as one of the most complex types of hydrological modeling, pri­
HEC-HMS marily because it involves the integration of a diverse array of watershed characteristics. Due to its ability to
Runoff simulation Tikur Wuha watershed emulate the hydrological behavior of a watershed, the modeling of rainfall-runoff plays a crucial role in pre­
dicting the runoff generated at the watershed’s outlet. The present study aimed to simulate runoff by utilizing
HEC–HMS in the Tikur Wuha River watershed situated in the Rift Valley Basin of Ethiopia. To achieve this goal,
tools such as HEC-GeoHMS and ArcGIS were employed to establish the necessary input parameters for
HEC–HMS. Various methods were implemented at different stages of the modeling process, including SCS-CN
for estimating precipitation loss, SCS-UH for transforming excess rainfall, Muskingum for flood routing, and
the monthly constant method for modeling base flow. The process of calibration and validation entailed the use
of daily observed flow data from the periods (1990 to 2009) and (2010 to 2015) correspondingly. Nash Sutcliff
Efficiency (NSE) and coefficient of determination (R2) were employed as metrics to evaluate the model’s per­
formance. The findings showed that the model exhibited high performance in both calibration and validation
stages, producing values of (NSE = 0.83, R2 = 0.91) and (NSE = 0.84, R2 = 0.86) respectively. Furthermore, the
Percent Bias (PBIAS) values in calibration and validation remained within acceptable ranges, registering at 2.69
% and 4.67 % respectively. After the calibration and validation of the model, the estimated peak flood discharge
simulated by the model (206.3m3/s) was compared with the observed stream flow (197.1m3/s), indicating a
significant similarity between the model’s output and the observed data. Consequently, it can be inferred that the
model exhibits a high capability in replicating hydrological parameters effectively for the Tikur Wuha watershed
and other watersheds sharing similar hydrological characteristics.

1. Introduction employed method to assess the hydrological reaction of a basin to pre­

cipitation. The model assists in the management of watershed practices
Hydrological modeling involves the analysis of how the hydrology of by considering the hydrological response associated with the projected
a watershed responds to diverse physical characteristics of the basin. It peak flood, to combine data for enhancing comprehension and execu­
has been employed in numerous river basins globally to enhance un­ tion of these practices (Kadam, 2010). Rainfall-runoff modeling is
derstanding of the availability of water resources (Sintayehu, 2015). crucial hydrological modeling that is commonly utilized to examine the
Currently, the use of hydrological models plays a vital role in assessing relationship between rainfall and runoff, considering factors unique to
the water availability in river basins and devising effective approaches the watershed (Salwa and Wardah, 2015). Moreover, it plays a vital role
to manage environmental changes. Estimating the volumes of runoff and in various activities such as flood simulations, monitoring water levels,
peak floods can be facilitated by employing a modeling framework and a and forecasting floods (Wang and Liu, 2023). Furthermore, it provides
comprehensive understanding of the factors that initiate runoff (Zhang valuable insights for effective planning and management of water re­
et al., 2010). The simulation of Stream flow data from rainfall events has sources. Additionally, it is important for catchments with limited data,
been advanced over numerous decades (Todini, 2007; Obasi et al., 2020; as it provides accurate predictions. Hence, adequate knowledge of hy­
Gholami and Sahour, 2021). Hydrological modeling is frequently drological modeling is very important to predict runoff produced from

* Corresponding author.
E-mail address: [email protected] (J.U. Guduru).

https://doi.org/10.1016/j.envc.2024.101017
Received 12 June 2024; Received in revised form 8 September 2024; Accepted 19 September 2024
Available online 21 September 2024
2667-0100/© 2024 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-
nc/4.0/).
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

watersheds. The fluctuation in stream flow within the Tikur Wuha river basin
Despite the data scarcity, different researchers conducted rainfall- varies seasonally, leading to flooding issues for surrounding agricultural
runoff modeling in various river basins. For instance, Sardoii et al. areas, particularly between June and September (Mekin et al., 2020). It
(2012) simulated the rainfall-runoff process in the Amirkabir watershed. is challenging to estimate flood depth without using hydrological
Kimhuy et al. (2016) developed rainfall-runoff modeling to assess stream models. To address these issues, the development and application of
flow and water resource accessibility in the Stung Sangker watershed of rainfall-runoff modeling is crucial. Therefore, this study aims to estab­
the Mekong’ Tonle Sap Lake basin in Cambodia. Bitew et al. (2019) lish rainfall-runoff modeling specifically for this area. The outcomes of
created a precipitation spillover model for stream simulation in the Lake this research are significant for the region as the model was meticulously
Tana Basin for the case of Gilgel Abay catchment, Upper Blue Nile Basin, prepared, incorporating various elements such as DEM maps, Isohyetal
Ethiopia. maps, land use/land cover data, soil types, curve numbers, rainfall in­
Although numerous options of rainfall-runoff models are available, formation, and simulated discharge values. Additionally, the charac­
selecting an appropriate model is crucial for effective watershed plan­ teristics of the Tikur Wuha river watershed were delineated using
ning and management. HEC-Geo HMS and transferred to HEC–HMS. The calibrated hydro­
Currently, numerous models are available for hydrological logical parameters will be used for upcoming hydrological in­
modeling. The choice of model is contingent upon the specific charac­ vestigations in this watershed and others alike, helping in the efficient
teristics of the basin as well as the intended purpose of the hydrological management of water resources.
forecasting within that particular basin (Ocio et al., 2019). This research
utilized a semi-distributed hydrologic model, the Hydrologic Engineer­ 2. Materials and methods
ing Center-Hydrologic Modelling System (HEC–HMS), to study the
rainfall-runoff process in the Tkur Wuha River watershed. The Hydro­ 2.1. Description of study area
logic Engineering Center’s Hydrologic Modeling System (HEC–HMS)
model was created by the US Army Corps of Engineers (Chu and Stein­ The Tikur Wuha river basin is situated within the geographical co­
man, 2009) and is suitable for various hydrological simulations. Hy­ ordinates of 6◦ 46′ to 7◦ 10′ N and 78◦ 32′ 23′’ to 78◦ 39′ 23′’ E in the
drologic Engineering Center (HEC–HMS) is a hydrologic modeling that southern Rift Valley basin of Ethiopia. It encompasses an area of 623
contains an integrated tool for modeling hydrologic processes of den­ km2 at the gauging station with elevations ranging from 1645 m to 2986
dritic watershed systems (Chu and Steinman, 2009). Moreover, it has m above mean sea level (Fig. 1). Originating from the Cheleleka Wet­
been widely used in many hydrological studies due to its simplicity and lands, the Tikur Wuha River is the primary watercourse that flows into
capability (Halwatura and Najim, 2013; Jayanti, 2020; Sahu et al., Lake Awasa (Mekin et al., 2020). The predominant soils and land use in
2023). Furthermore, HEC–HMS is a comprehensive tool that encom­ the watershed are Chromic Luvisols and Cultivated land respectively, as
passes all hydrologic processes of dendritic watershed systems. The illustrated in Table 1. The watershed displays significant topographic
widespread adoption of HEC–HMS in hydrology is attributed to its diversity, with the most prevalent slope classes being 8–15 % and 15–30
capacity to simulate runoff in various event durations, ease of use, and %, which together cover approximately 54 % of the total area. The slope
utilization of standard methods (Oleyiblo et al., 2010). class of 3–8 % encompasses 21.6 %, while slopes exceeding 30 %
The Geospatial Hydrologic Modeling Extension (HEC-GeoHMS) is a constitute 19 %, primarily found in the Northern, Eastern, and South
public-domain software package for use with Geographical Information Eastern escarpments of the sub-watershed area. The remaining 5 % of
Systems (GIS), GeoHMS ArcView, and Spatial Analysis to develop the total area is categorized under the 0–3 % slope class according to the
several hydrological modeling inputs. Upon examination of the data FAO slope classification for soil and water conservation (FAO, 2006). In
from the Digital Elevation Model (DEM), HEC-GeoHMS converts the terms of climate, the watershed experiences a semi-humid climate, with
drainage routes and watershed limits into a hydrological data frame­ average annual rainfall ranging from 210 mm to 250 mm based on data
work that illustrates the watershed’s reaction to precipitation from 1990 to 2017, with over 80 % of precipitation occurring during the
(Hoogestraat, 2011). Several researchers employed the HEC–HMS summer season (June to August).
model to depict flow through simulated rainfall-runoff processes. The
study by Tahmasbinejad et al. (2012), HEC–HMS and GIS were effec­ 2.2. Data set, and sources
tively utilized to replicate the rainfall-runoff process in the Karun River
basin in Iran. Similarly, Sampath et al. (2015) carried out runoff simu­ This study used various datasets including daily precipitation, stream
lation in the Tropical Region of Deduru Oya River Basin in Sri Lanka flow, soil types, Land use/cover (LULC), and Digital Elevation Model
using the HEC–HMS model, demonstrating its effectiveness in simu­ (DEM) 12.5 × 12.5 m data. The precipitation data was collected from
lating runoff. Abdessamed et al. (2018) formulated a rainfall-runoff four meteorological stations (Hawasa, Hasayita, Tula, and Waterersa)
model in a semi-arid region of the Ain Sefra watershed in Algeria by available in the study area spanning from 1990 to 2017 sourced from the
implementing an HEC–HMS model. Zelelew and Melesse (2018) Ethiopian Meteorological Agency, and stream flow data from 1990 to
emphasized the location-specific nature of model simulation results, 2015 at the gauging station (indicated on Fig.1) from the Ethiopian
where different combinations of model components responded variably. Hydrological Agency. Similarly, the Soil and DEM data were acquired
Tassew et al. (2019) performed a rainfall-runoff simulation with from the Ethiopian Ministry of Water Resource, Irrigation, and Elec­
HEC–HMS for the Lake Tana Basin of the Gilgel Abay catchment in the tricity, whereas the LULC data was obtained from the Ethiopian Map­
upper Blue Nile basin in Ethiopia, using six extreme daily time-series ping Agency. The accuracy of the gathered original meteorological and
events. The findings suggested the model’s suitability for hydrological hydrological data has a notable impact on the accuracy of the data used
simulations. Saeedrashed et al. (2020) utilized computational hydro­ in the model and, by extension, the model’s simulations. Initially, the
logical and hydraulic modeling systems that integrate GIS with the daily rainfall and stream flow data underwent visual inspection for
modeling systems to predict floodplains for the Greater Zab River using quality assurance (Kim et al., 2023). Subsequently, missing data points
HEC–HMS. Wana et al. (2020) applied HEC–HMS for rainfall-runoff were filled in and a thorough evaluation was conducted using the
modeling in the Awash Bello sub-catchment, concluding its efficient normal ratio and double mass curve analysis techniques (Gao et al.,
predictive capabilities. Furthermore, Hamdan et al. (2021) conducted 2017). Data consistency was assessed using the double mass curve
Rainfall-Runoff Modeling with the HEC–HMS Model for the Al-Adhaim method, and additional data quality tests like the homogeneity test were
River Catchment, Northern Iraq, affirming its suitability for hydrological carried out using Excel Stat statistical software. The land use/land cover
simulation. The Tikur Wuha River watershed is situated in the southern (LULC) and soil data of the Tikur Wuha river watershed were extracted
part of the Rift Valley River basin. from the LULC and soil map of the Rift Valley basin. Furthermore, given

2
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

Fig. 1. Location of Tikur Wuha River Watershed.

Note: M.station: meteorological station, Strof: streamflow line.

2.3. Methods
Table 1
Spatial areal coverage and percentage of LULC and soil type of study area.
2.3.1. Software used
No LULC Area Area Soil type Area Area ArcGIS, HEC-GeoHMS, and HEC–HMS were the software employed
(km2) (%) (km2) (%)
in this work. ArcGIS was used to prepare curve numbers grid by merging
1 Bush Land 83.86 12.9 Chromic 258.01 40 soil and land use data, which was subsequently utilized as an input in
Luvisols HEC-GeoHMS to produce curve numbers. The ArcGIS HEC-GeoHMS
2 Cultivated 224.9 34.85 Haplic 245.56 38
land Luvisols
extension toolkit was also used for watershed delineation and the cre­
3 Forest land 332.13 51.5 Eutric 30.4 4.7 ation of parameters such as curve numbers, basin model files, gage
Cambisol model files, and met model file (precipitation data) to construct a model
4 Residential 1.173 0.18 Vitric 29.08 4.4 in HEC–HMS. HEC-GeoHMS acts as an intermediary between ArcGIS
area Andosols
and HEC–HMS, facilitating the smooth transfer of data from ArcGIS to
5 wetland 2.1 0.36 Vitric 78.7 12.2
Luvisol HEC–HMS. HEC–HMS requires watershed background shape files for
6 Water body 0.132 0.025 Leptosol 3.66 0.56 precise watershed configuration.
​ total 645.36 100 total 645.36 100
2.3.2. HEC–HMS model
The Hydrologic Engineering Centre-Hydrological Modeling System
the variability in precipitation levels across different locations, accu­
(HEC–HMS), is a highly adaptable and effective hydrological model
rately estimating the average precipitation for the basin is crucial rather
used to analyze the rainfall-runoff process within a watershed. This
than relying solely on single-point rainfall data. Hence, the Isohyetal
model has been widely embraced in numerous hydrological in­
method was selected for calculating average precipitation due to its
vestigations due to its capacity to replicate runoff occurrences in both
ability to provide the most precise estimation (Subramanya, 2008).
short-term and long-term scenarios, as well as its user-friendly interface
Accordingly, the average areal precipitation data was computed for each
(Najim, 2013). It is specifically created to mimic the precipitation-runoff
station (W1130, W680, W690, and W800) as shown in Fig.2. Subse­
mechanisms of complex watershed systems (Scharffenberger, and
quently, the hydrological and meteorological data in the form of spatial
Fleming, 2016). The HEC–HMS model can be employed to simulate
and time series were organized in a way that was appropriate for the
various scenarios such as single storm events lasting from hours to days,
application of the HEC–HMS hydrological model in the context of
or extended periods of stream flow (daily, monthly, and seasonal)
rainfall-runoff modeling.
(Fleming, and Doan, 2013). The setup of the HEC–HMS model com­
Table 2
prises four key components: basin model, meteorological model, control
specifications, and input data (time series, paired data, and gridded

3
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

Fig. 2. Isohyetal map study area.

(Im et al., 2020). The values of the Curve Number for a specific water­
Table 2
shed is determined upon various factors such as land use, soil compo­
Average precipitation computed by Isohyetal method for this study area.
sition, and Antecedent Soil Moisture (AMC), with values ranging from
No Isohyetal range Average Area sq.Km Average rainfall 30 soils with high infiltration rates) to 100 (water bodies)
(mm) (mm) (mm)
(Scharffenberg, Fleming, 2016). The SCS CN model is represented by Eq.
1 <990 985 32 49 (1) (Uwizeyimana et al., 2019).
2 990–1020 1005 69 107.45
3 1020–1050 1035 104 167 (P − Ia )2
4 1050–1080 1065 134 221 Q= (1)
(P − Ia ) + S
5 1080–1110 1095 207 352
6 1110–1140 1125 83 145 Where: Q = runoff value (mm); P = precipitation (mm); Ia = initial
7 1155 16.36 29.3
abstraction (mm); S = potential maximum retention is given by Eq. (2):
>1140
∑ ∑
A = Ave = 1071
The potential maximum retention (S) is a function of Curve Number
645.36
(CN) and is inversely proportion to CN. The potential maximum reten­
tion is given by Eq. (2):
data). There are multiple techniques available within the model to
25400
simulate processes like infiltration losses, direct runoff estimation, S= − 254 (2)
CN
channel routing, base flow modeling, and meteorological modeling. The
selection of the appropriate model component depends on factors such
as data availability, modeling objectives, and global standards. For the 2.4. Model calibration and validation
present study, methods such as Soil Conservation Service Curve Number
(SCS-CN), Soil Conservation Service Unit Hydrograph (SCS-UH), Calibration of a model involves the meticulous adjustment of specific
Muskingum and Gage weights methods, and constant monthly base flow parameter values within the model to ensure that the simulated out­
were chosen to calculate runoff volume, transform excess precipitation comes align closely with the actual observations. The effectiveness of a
to surface runoff, manage channel routing, and simulate the overall hydrologic model is contingent upon the accuracy of its calibration
process, respectively. Additionally, the watershed was subdivided into process, as highlighted by Vaze et al. (2011) in their research. Following
four sub-basins to enhance the model’s performance. The SCS-CN the initial simulation of a model using certain parameters, the subse­
methods employed by the Soil Conservation Service (SCS) uses the quent step involved the calibration of the model against observed
Curve Number method for calculating runoff and is applied to ascertain streamflow data. This calibration process entailed the fine-tuning of
the overall volume of infiltration that occurs during a rain event (Lee parameters such as the curve number, lag time, and initial abstraction
et al., 2023). Curve Number (CN) is an index developed by the Soil values until a satisfactory alignment between the simulated and
Conservation Service (SCS) to estimate rainfall loss into soil and the observed results was achieved. Both automated and manual methods
subsequent surface runoff, which is essential for the modeling of were employed to complete this calibration process, reflecting the
rainfall-runoff in hydrological simulations (Jayanti, 2020). It is complexity and precision required in model calibration procedures. To
commonly employed for the estimation of surface runoff in hydrology carry out the calibration, 18 years of daily observed streamflow data
spanning from 1990 to 2010 were utilized, ensuring a robust and

4
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

comprehensive evaluation of the model’s performance. The iterative for effective disaster management and mitigation strategies (Keast &
nature of model calibration underscores the significance of this process Ellison, 2013;Strupczewski et al., 2017). In this study, the quantity of the
in enhancing the reliability and accuracy of hydrologic models for flood was determined utilizing the HEC–HMS model in conjunction
various applications in water resource management and environmental with the statistical approach known as the Gumbel distribution method,
studies. Through the systematic adjustment of parameters based on which is widely recognized for its efficacy in extreme value analysis and
observed data, model calibration serves as a critical step in validating has been extensively utilized in hydrological studies to predict the
the predictive capabilities of hydrologic models and improving their probability of occurrence of rare flood events with a high degree of
overall performance in simulating real-world scenarios. The utilization reliability and precision (Farooq et al., 2018; Parchure and Gedam,
of extensive observed data over a prolonged period further enhances the 2019; El Mehdi Saidi et al., 2020). To comprehensively assess the fre­
robustness and credibility of the calibration outcomes, leading to more quency of flooding events various return periods (2, 10, 25, 50, and 100
informed decision-making in water-related projects and policy devel­ years) were considered. This rainfall depths data corresponding to these
opment. Model validation involves confirming the model’s predictive return periods, was calculated using Intensity-Duration-Frequency (IDF)
ability for timeframes beyond the calibration period while keeping the curve. The IDF results for different storm durations and average recur­
input parameters unchanged from the calibration process (Vaze et al., rence intervals (ARI) of the region are detailed in Table 3. The greatest
2011). This process guarantees the effectiveness of the calibrated pa­ rainfall depth was observed over a 24-hour period, and these values
rameters when applied to a separate dataset. The validation of the model were employed in the model simulations.
was carried out by analyzing daily streamflow data from 5 years (2011 Following model calibration and validation, flood frequency analysis
to 2015). was carried out for return periods. For this analysis, the depth of rainfall
recorded over a 24-hour period was employed as the critical input
2.4.1. Model performance evaluation parameter for the HEC–HMS, while simultaneously, the maximum
The assessment of the HEC–HMS model’s performance included annual rainfall data pertinent to the geographic study area was used in
analyzing the accuracy of the observed and simulated stream flow. A the Gumbel method to examine and quantify the magnitude of potential
statistical error test was utilized to assess the quality and dependability flood events.
of the simulated values produced by the HEC–HMS model. In this study, This dual approach not only facilitates a robust evaluation of flood
the evaluation of the model’s effectiveness considered the Nash-Sutcliffe frequency but also enhances predictive accuracy regarding the antici­
Efficiency (NSE) and coefficient of determination (R2) as widely used pated severity of flooding scenarios within the designated area of study,
metrics in hydrological modeling. ultimately contributing valuable insights for effective flood manage­
ment and mitigation strategies.
A. Nash-Sutcliffe Efficiency (NSE)
3. Results and discussion
The Nash-Sutcliffe Efficiency (NSE) is a normalized statistic that
determines the relative magnitude of the residual variance compared to 3.1. Hydrologic parameters
the measured data variance (Moriasi et al., 2007). Mathematically, it is
calculated as Eq. (3). As deliberated in the section on methodologies, crucial watershed
∑ attributes such as Curve number, basin lag time, watershed area, basin
(Qob(t) − Qsim(t) )2 slope, potential maximum retention (S), and the initial abstraction from
NSE = 1 − ∑ (3)
(Qob(t) − Qob )2 the watershed were established. The spatial variation of the CN value
within the study area was determined. The study area exhibited a range
Where, NSE: Nash and Sutcliffe Efficiency, Qob: observed value at
of CN values, with a minimum of 30 for forested areas and a maximum of
the ith time interval, Qsim: simulated value at the ith time interval, Qob :
100 for water bodies. However, for rainfall-runoff simulation, each sub-
mean of the observed discharges
basin necessitates a singular CN value. Consequently, HEC-GeoHMS had
already calculated the weighted CN values for each sub-basin. The
A. Coefficient of determination (R2)
minimum and maximum weighted curve number values were identified
as 74.15 and 72.13, respectively. These extremities were observed in
The Coefficient of determination (R2) is another commonly utilized
sub-basins W690 and W680, respectively. The Curve Number value
statistical measure that signifies the level of co-linearity existing be­
directly influences runoff generation. Sub-basins with lower CN values
tween the simulated and observed data. It reflects the accuracy of a
result in minimal runoff and higher infiltration rates. Conversely, sub-
model, with a range from zero to one. A value of one indicates flawless
basins with higher CN values exhibit greater runoff potential. Hence,
prediction, while a value of zero indicates inadequate prediction
sub-basins W690 and W680 generate high and low runoff, respectively.
(Moriasi et al., 2007). Its mathematical calculation is performed using
The spatial distribution of CN is illustrated in Fig.3. Similarly, the
Eq. (4).
minimum and maximum initial abstractions were recorded as 17.71 mm
∑ ∑
(Qobs(t) − Qobs ) (Qsim(t) − Qsim(t) )2 in W690 and 19.68 mm in W680 sub-basin. This discrepancy signifies
R2= ∑ ( ∑ (4) the variance in runoff values between W680 and W690. The steepest
(Qobs(t) − Qobs )2 (Qsim(t) − Qsim(t) )2

Where, R2 = coefficient of determination, Qobs = observed value at

the ith time interval, Qsim(t) = simulated value at the ith time inter­ Table 3
Design storm of rainfall intensity for Tikur Wuha river watershed.
val, Qobs = mean of observed discharges, Qsim(t) = Mean of simulated
Rainfall intensity duration(hr.) Rainfall depth(mm) versus return period (yrs.)
discharges value at ith time interval.
10 25 50 100

1 43.2 50.22 55.125 60.03

2.5. Flood prediction 2 51.78 59.60 65.42 71.25
3 56.125 64.6 70.9 77.23
Comprehensive flood frequency analysis is critical, as it serves as a 6 62.73 72.2 79.25 86.31
critical tool for elucidating the intricate characteristics and potential 12 68.7 79.07 86.8 94.53
24 74.45 85.70 94.07 102.45
severity of both present and forthcoming flood events, thereby equip­
ping researchers and policymakers with invaluable insights necessary Source: (ERA, 2013).

5
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

Fig. 3. Mean Curve Number map of the Meki River watershed.

basin slope, measuring 44.93 %, was observed in sub-basin W800. This

Table 5
indicates that W800 is the steepest sub-basin among all. The basin lag
Simulated results at each sub-basin and river reaches.
time ranged between 20.72 and 36.32 min for the study area. A lower
basin lag time implies quicker surface runoff reaching the outlet point. Sub- Daily peak River Routed River Routed
basin discharge (m3/ channel flood (m3/ channel flood (m3/
The detailed watershed characteristics are presented in Table 4.
s) s) s)

W680 152.5 R480 33.9 R400 226.0

3.2. Model simulation result W690 153.6 R470 34.0 R430 226.6
W800 31.8 R500 33.8 R350 258.5
W1130 64.9 R540 33.7 R240 258.8
The maximum daily flow rate at the watershed outlet has been Outlet 286.8 R330 156.6 R420 278.1
calculated to be 286.8m3/s, as indicated by the findings. This particular R320 157.8 R410 278.7
value carries significant implications for the overall hydrological pro­
cesses within the watershed. Furthermore, a comprehensive breakdown
of the daily estimated flow rates for each sub-basin and river can be characteristics along the river network, potentially influencing down­
found in the detailed information provided in Table 5, enhancing the stream water flow and quality.
understanding of the flow dynamics within the watershed. The analysis
highlights the W690 sub-basin as contributing the most significant 3.3. Model calibration and validation
volume compared to other sub-basins. On the contrary, the W800 sub-
basin is identified as the least significant in terms of the amount of 3.3.1. Model calibration
runoff it generates. This information underscores the variations in hy­ The calibration of the HEC–HMS model involved the adjustment of
drological contributions from different areas within the watershed. model parameters to ensure a close match between simulated and
Additionally, the noticeable increase in water depth as the river ad­ observed flow data within an acceptable range of deviation. This process
vances towards the outlet location signifies a shift in the hydrological aimed to enhance the agreement between the two datasets by refining
the model’s representation of the hydrological processes. Notably, the
Table 4 time to peak of both simulated and observed events coincided, further
Watershed parameters generated by HEC-GeoHMS for Tikur Wuha River validating the model’s performance. Through iterative adjustments
watershed. during the calibration process, the peak stream flow values were refined
Sub-basin Area Slope Ia (mm) S CN∗ Lag time to 286.8 m3/s for simulated data and 208.6 m3/s for observed data,
Code km2 % (mm) (min) indicating a significant improvement in model accuracy. To achieve this
W680 112.87 26.3 19.68 98 72.13 27.11 calibration, various watershed parameters such as lag time, curve
W690 58.2 43.94 17.71 88.5 74.15 20.72 number, initial abstraction, flood wave traveling time (Muskingum-k),
W800 281.05 44.93 18.76 93.8 73.03 26.45 and weighted coefficient of discharge (Muskingum-x) were carefully
W1130 179.42 39.68 19.44 97.2 72.32 36.32
selected. The sensitivity analysis revealed that all these parameters

6
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

played crucial roles in optimizing the HEC–HMS model, with particular of the simulated and observed stream flow hydrograph at the outlet of
emphasis on curve number and lag time. Detailed results presented in the watershed during the validation period show a similar pattern as the
Table 6 showcased the initial and optimized values of these parameters model calibration period.
along with their corresponding objective function sensitivity values. For
instance, the range of initially computed values for Muskingum-k 3.4. Model performance evaluation
spanned from 1:23 to 3:00 h, which were subsequently adjusted to
1:18 and 2:77 h post-calibration. The performance of the model was evaluated using Nash-Sutcliffe
Furthermore, the calibration process involved computing and Efficiency (NSE), Root Mean Squared Error (RMSE), and Coefficient of
adjusting the Muskingum-k values for each reach, as outlined in the Determination (R2). Moriasi et al. (2007) studied model evaluation
comprehensive Table 6. This meticulous approach ensured that the guidelines for systematic quantification of accuracy in Watershed sim­
model accurately captured the flow dynamics at different sections of the ulations and stated that the model to be very good, the value of NSE and
watershed. Similarly, the determination of Muskingum-x values for each R2 should be between 0.75 and 1.0, whereas, the value of RMSE should
reach was achieved through a methodical trial and error process, ulti­ be 0 to 0.5. Similarly, Schaefli and Gupta (2007), Kashid (2010), and
mately resulting in optimized values of 0.24. These detailed adjustments Vaze (2011) stated that, if the value of NSE and R2 during calibration
and optimizations underscore the thoroughness and rigor applied to the and validation are between 0.75 and 1 the model performance rating is
calibration of the HEC–HMS model, leading to a more reliable repre­ classified as a very good model. For the current research, the NSE and R2
sentation of the hydrological system. values exceeded 0.75 in both the calibration and validation periods,
Fig. 5 illustrates the comparison between the simulated and observed while RMSE values were 0.5 and 0.4 during calibration and validation,
flow hydrographs of the Tikur Wuha River watershed at the outlet respectively. The results from calibration and validation phases
following the calibration period. The agreement between the low flow demonstrated a robust correlation between the modeled and observed
and peak flow of both the simulated and observed hydrographs was streamflow data. Consequently, the HEC–HMS model achieved a high
notably strong, displaying a consistent pattern throughout the calibra­ performance rating based on these statistical metrics. The model’s
tion period. Additionally, the scatter plot depicting the relationship ability to predict accurately during calibration and validation stages
between the measured and simulated stream flow for the same cali­ suggests its efficacy in efficiently simulating daily streamflow from
bration period reveals a correlation coefficient (R2) of 0.91, signifying a rainfall data in the study area. Therefore, the model’s performance was
robust correlation between the simulated data and the actual observed deemed satisfactory and deemed suitable for predicting peak floods
stream data as depicted in Fig.4. under various management scenarios in the future. Additionally, the
simulated streamflow data effectively represented the observed
3.3.2. Model validation streamflow data in the study area.
All statistical error tests were observed to fall within an acceptable Table 7
range (0.75–1) throughout the validation process, specifically during the
calibration period, indicating that the predicted calibration outcome has 3.5. Flood prediction
been duly confirmed. The graphical representation in Fig.6 illustrates a
close alignment between the low flows and peak flow of both the 3.5.1. Flood prediction by HEC-HMS
simulated and observed streamflow hydrographs, demonstrating a Flood frequency analysis was carried out for return periods of 2, 10,
consistent pattern during the validation phase. This empirical evidence 25, 50, and 100 years utilizing the HEC–HMS model within the Tikur
strongly suggests that the HEC–HMS model exhibits a high level of Wuha River watershed. The analysis took into account a 24-hour rainfall
performance in replicating stream flow data within the designated study depth, resulting in peak floods of varying magnitudes. The investigation
area. Hence, it can be inferred that the model holds promise for accu­ revealed that the range of peak flood levels at the outlet of the Tikur
rately representing the hydrological processes in the region. Wuha River watershed spanned from 133.2 m3/s to 346.19 m3/s. This
The scatter plot of measured and simulated flow during the valida­ disparity signifies that the lowest peak flood within the Tikur Wuha
tion period shows a fair linear correlation between the simulated and River watershed occurred during a 2-year return period with a 24-hour
observed data during the calibration period Fig. 7. The shape and scatter storm duration, while the highest flood was observed for the 100-year
return period under similar storm conditions. Assuming a constant
basin lag time across all return periods (2, 10, 25, 50, and 100 years), the
Table 6 study successfully predicted the peak discharge and shape of the
HEC–HMS optimized parameters of Tikur Wuha watershed. hydrograph for each scenario. These predictions were visually observed
Element Parameter Unit Initial Optimized Objective in Table 8, providing a comprehensive overview of the hydrological
value value function dynamics in the Tikur Wuha River watershed.
sensitivity

W680 Lag time HR 75.95 77.23 − 0.37 3.5.2. Comparison of HEC-HMS and gumbel distribution result
Curve 65.7 56.21 − 0.17
​
The distribution that exhibited the highest level of suitability to the
Number
W690 Lag time HR 53.6 54.3 − 0.46 stream flow data provided was found to be the Gumbel distribution
Initial MM 22.3 22.45 − 0.23 through rigorous statistical analysis. The anticipated maximum flow
abstraction rate derived from the application of this specific probability distribution
W800 Initial MM 10.3 15.15 0.00 function has been delineated and presented thoroughly within Table 9,
abstraction
W1130 Initial MM 8.7 9.13 0.00
offering valuable insights into the potential peak flow scenarios in the
abstraction context of the studied data set. This statistical modeling approach not
Curve ​ 85.4 84.34 − 0.32 only aids in understanding the underlying patterns and trends within the
Number stream flow data but also facilitates the projection of peak flow values
R480 Muskingum-k HR 3.00 2.77 0.00
with a certain degree of confidence, thereby enhancing the predictive
Muskingum-x ​ 0.25 0.24 0.00
R330 Muskingum-k HR 2:00 1.88 0.00 capabilities of the analysis.
R320 Muskingum-k HR 1:89 1.78 0.00 The outcome derived from the Gumbel method was observed to
Muskingum-x ​ 0.25 0.24 0.00 exhibit a high degree of proximity to the simulated outcome generated
R400 Muskingum-k HR 1:75 1.65 0.00 by HEC–HMS model as illustrated in Fig.8. Nevertheless, it was noted
R350 Muskingum-k HR 1:23 1.18 − 0.01
that the maximum flood level forecasted by the HEC–HMS model

7
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

Fig. 4. Simulated and observed flow hydrographs after calibration.

Fig. 5. Scatter plot of observed and simulated flow after calibration.

Fig. 6. Simulated and observed stream flow hydrographs after validation.

8
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

Fig. 7. Scatter plot of observed and simulated flow after validation.

strategic planning of water resources management. The utilization of the

Table 7
HEC–HMS model in the current research endeavor facilitated the
Summary of model performance evaluation.
modeling of stream flow within the Tikur Wuha watershed, showcasing
Performance rating After Allowable Remark the efficacy and reliability of this computational tool. Various essential
range
Calibration Validation datasets including hydro-meteorological information, soil characteris­
Nash-Sutcliffe Efficiency 0.83 0.80 0.75–1.0 accepted tics, land use/land cover data, and Digital Elevation Model (DEM) data
(NE) were meticulously incorporated into the framework of this research to
Coefficient of 0.91 0.86 accepted ensure a comprehensive and robust analysis. The integration of
determination (R2) advanced software tools such as ArcGIS and HEC-GeoHMS enabled the
creation of basin models, alongside the determination of crucial input
parameters like Curve Number, lag time, and initial abstractions, which
Table 8 are pivotal in accurately representing the hydrological processes within
Simulated peak flood of different return period by HE-HMS. the watershed. Moreover, the excluded value was computed using both
Return period (year) 24 hour storm (mm) peak flood (m3/s) by HEC–HMS simple and standard ratio techniques, while areal precipitation was
assessed through Isohyetal methods. Additionally, the missed stream
2 47.54 133.2
10 67.66 178.1 flow value was estimated using linear regression methods. The research
25 77.92 239.7 employed techniques such as the Soil Conservation Service Curve
50 85.62 313.2 Number, Soil Conservation Service Unit Hydrograph, constant base
100 93.34 346.19 flow, and Muskingum method to evaluate different aspects of the hy­
drological cycle, encompassing rainfall loss, runoff, base flow modeling,
and channel routing. The model was subjected to calibration and vali­
Table 9 dation processes using 18 years (1987–2004) and 6 years (2005–2010)
Peak discharge found from flood frequency analysis. of daily observed streamflow data, respectively. The study site displays
No. Return period (year) Peak flood (m3/s) curve numbers ranging from 30 to 100. Key factors influencing output
Simulated Computed
include lag time, curve number, initial abstraction, flood travel time
HEC–HMS Gumbel (Muskingum-k), and discharge weighting factor (Muskingum-x). The
model’s efficacy was assessed using the Root Mean Squared Error, Nash-
1 2 133.2 126.7
2 10 178.1 167.8 Sutcliffe Efficiency (NSE), and Coefficient of Determination (R2),
3 25 239.7 223.5 achieving values of 0.5, 0.832, and 0.91 respectively during calibration,
4 50 313.2 287.9 and 0.4, 0.804, and 0.86 during validation. These findings demonstrate
5 100 346.19 331.87
the robust performance of the model, indicating that the HEC–HMS
model is well-suited for simulating streamflow data based on rainfall
surpassed the maximum flood level calculated through the Gumbel data within the study area. Based on the results of the goodness of fit
method. This observation serves to imply that the simulated peak analysis carried out using the easy fit software, the Gumbel method
discharge values produced by the HEC–HMS model hold potential for emerged as the most appropriate probability distribution function for
further application in the realms of flood mapping and the imple­ fitting the observed stream flow time series data, ranking first in both the
mentation of strategies aimed at mitigating the impact of floods. Kolmogorov-Smirnov and chi-squared tests. Subsequent modifications
to the model configuration led to the performance of flood frequency
4. Conclusion analysis for various return periods, ranging from 2 to 100 years, utilizing
rainfall data from a 24-hour storm in the Tikur Wuha river watershed
Rainfall-runoff modeling holds significant importance in the realm of obtained from ERA, 2013. Consequently, the projected maximum flood
hydrology as it plays a crucial role in simulating the intricate response discharge estimated using HEC–HMS and the Gumbel method for re­
patterns of watersheds to varying intensities of rainfall events, ulti­ turn periods of 2, 10, 25, 50, and 100 years were found to be 133.2,
mately leading to the generation of flow hydrographs. These hydro­ 178.1, 239.7, 313.2, and 346.19 m3/s, and 126.7, 167.8, 233.5, 287.9,
graphs are widely utilized in the realm of flood forecasting and the and 331.87 m3/s respectively. The peak flood anticipated by the
HEC–HMS model surpasses that predicted by the Gumbel distribution.

9
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

Fig. 8. Graphical Comparison of HEC–HMS result with the Gumbel method.

In general, study verified the practical applicability of the proposed Farooq, M., Farooq, M., Shafique, M., & Khattak, M.S. (2018). Flood frequency analysis of
river swat using Log Pearson type 3, Generalized Extreme Value, Normal, and Gumbel
model in conjunction with the quantification of flood events across the
Max distribution methods. 11(9). https://doi.org/10.1007/S12517-018-3553-Z.
study area, thereby establishing its potential utility as a base input for Fleming, M.J. and Doan, J.H. (2013) HEC-GeoHMS geospatial hydrologic modeling
subsequent researchers who aim to delineate the areal extent of flooding extention manual vesion 10.1.
as well as the identification of areas that are particularly susceptible to Fleming and Scharffenberg, (2016) ‘Hydrologic Modeling System (Hec-Hms): new
features for urban’.
flooding and implementing suitable strategies to mitigate its effects in Gao, P., Gao, P., Li, P., Li, P., Zhao, B., Zhao, B., Xu, R., Xu, R., Zhao, G., Zhao, G.,
the specified research area within the watershed. Sun, W., Sun, W., Mu, X., Mu, X., 2017. Use of double mass curves in hydrologic
benefit evaluations. Hydrol. Process. 31 (26), 4639–4646. https://doi.org/10.1002/
HYP.11377.
CRediT authorship contribution statement Gholami, V., Sahour, H., 2021. Simulation of rainfall-runoff process using an artificial
neural network (ANN) and field plots data. Theor. Appl. Climatol. 1–12. https://doi.
Jerjera Ulu Guduru: Writing – review & editing, Writing – original org/10.1007/S00704-021-03817-4.
Halwatura, D., Najim, M., 2013. Application of the HEC-HMS model for runoff
draft, Resources, Methodology, Investigation, Conceptualization. Aya­ simulation in a tropical catchment. Environ. Modell. Softw. 46, 155–162.
tullah Shis Mohammed: Writing – original draft, Visualization. Hamdan, A.N.A., Almuktar, S.A.A.A.N., Almuktar, S.A.A.A.N., Scholz, M., Scholz, M.,
Scholz, M., 2021. Rainfall-runoff modeling using the hec-hms model for the al-
adhaim river catchment, northern iraq. Hydrology 8 (2), 58. https://doi.org/
Declaration of competing interest 10.3390/HYDROLOGY8020058.
Hoogestraat, G.K., 2011. Flood hydrology and Dam-Breach Hydraulic analyses of four
The author declared no conflict of interest. reservoirs in the Black Hills, South Dakota. U.S. Geological Survey Scientific
Investigations Report 2011–5011. U.S. Geological Survey:, Reston, VA, USA, p. 37.
Im, S., Lee, J., Kuraji, K., Lai, Y.-J., Tuankrua, V., Tanaka, N., Gomyo, M., Inoue, H., &
Data availability Tseng, C.-W. (2020). Soil conservation service curve number determination for forest
cover using rainfall and runoff data in experimental forests. 25(4). https://doi.org/
Data will be made available on request. 10.1080/13416979.2020.1785072.
Jayanti, M. (2020). Climate change impacts on hydrology regime and water resources
sustainability in cimanuk watershed, west java, indonesia. 19(71). https://doi.org/10
.21660/2020.71.9215.
Acknowledgments Kadam, A.S., 2010. Event based rainfall-runoff simulation using HEC-HMS model.
J. CAET, Dr. PDKV, Akola.
Kim, H.-Y., Frommknecht, A., Bieberstein, B., Stahl, J., Huber, M., 2023. Automated end-
The authors express their gratitude to the Ethiopian Ministry of of-line quality assurance with visual inspection and convolutional neural networks.
Water, Irrigation, and Electricity for graciously providing essential data Tm-Tech. Mess. 90 (3), 196–204. https://doi.org/10.1515/teme-2022-0092.
Kimhuy, S. and Chantha, O., 2016. Application of HEC-HMS Model to assess streamflow
sets including soil characteristics, stream flow information, and land and water resources availability in stung sangker catchment of mekong’ tonle sap
use/cover statistics. Furthermore, they extend their appreciation to the Lake Basin in Cambodia DOI: https://doi.org/10.20944/preprints201612.0136.vl.
Ethiopian National Meteorological Agency for supplying crucial rainfall Lee, K.K.F., Ling, L., & Yusop, Z. (2023). The revised curve number rainfall–runoff
methodology for an improved runoff prediction. 15(3). https://doi.org/10.3390/
data, which significantly contributed to the comprehensive analysis w15030491.
conducted in this study. Mekin, M., Birhanu, B., Mulugeta, D.B., 2020. Hydrological impacts of climate change in
Tikur Wuha River watershed. Ethiopian Rift Valley Basin (10), 28–49. Vol.
Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Bingner, R.L., Harmel, R.D., Veith, T.L.,
Funding
2007. Model evaluation guidelines for systematic quantification of accuracy in
Watershed simulations. ASABE 50, 885–900.
This research did not receive any specific grant from funding Najim, 2013. Application of the HEC-HMS model for runoff simulation in a tropical
agencies. catchment. Environ. Model (46), 155–162.
Obasi, A.A., Ogbu, K.N., Orakwe, C.L., Ahaneku, I.E., 2020. Rainfall-river discharge
modelling for flood forecasting using Artificial Neural Network (ANN). J. Water
References Land Dev. 44. https://doi.org/10.24425/JWLD.2019.127050.
Ocio, D., Beskeen, T., Smart, K., 2019. Fully distributed hydrological modelling for
Abdessamed, D., Abderrazak, B., Kamila, B., 2018. Modelling rainfall-runoff relations catchment-wide hydrological data verification. Hydrol. Res. 50 (6), 1520–1534.
using HEC-HMS in a semi-arid region: case study in Ain Sefra watershed, Ksour https://doi.org/10.2166/NH.2019.006.
Mountains (SW Algeria). J. Water Land Develop., Polish Acad. 36 (I–III), 45–55. Oleyiblo, J.O., Li, Z.-j., 2010. Application of HEC-HMS for flood forecasting in Misai and
Bitew, G.T., Mulugeta, A.B., Miegel, K., 2019. Application of HEC-HMS model for flow Wanan catchments in China. Water Sci. Eng. 3, 14–22.
simulation in the Lake Tana Basin: the case of gilgel abay catchment, upper blue Nile Parchure, A.S., & Gedam, S.K. (2019). Probability distribution analysis of extreme rainfall
Basin, Ethiopia. Hydrology 6 (21). https://doi.org/10.3390/hydrology6010021. events in a flood-prone region of Mumbai, India. 12(11). https://doi.org/10.100
Chu, X., Steinman, A.D., 2009. Event and continuous hydrologic modeling with HEC- 7/S12517-019-4473-2.
HMS. J. Irrig. Drain. Eng.-Asce 135 (1), 119–124. https://doi.org/10.1061/(ASCE) Saeedrashed, Y.S., 2020. Hydrologic and hydraulic modelling of the greater zab river-
0733-9437(2009)135:1(119. basin for an effective management of water resources in the Kurdistan Region of Iraq
El Mehdi Saidi, M., Saouabe, T., El Alaoui El Fels, A., El Khalki, E.M., & Hadri, A. (2020). using DEM and raster images. Environmental Remote Sensing and GIS in Iraq.
Hydro-meteorological characteristics and occurrence probability of extreme flood events in Springer, Cham, Switzerland, pp. 415–446.
Moroccan High Atlas. 11. https://doi.org/10.2166/WCC.2020.069. Sahu, M.K., Shwetha, H.R., Dwarakish, G.S., 2023. State-of-the-art hydrological models
FAO, 2006. World Reference Base For Soil resources. World Soil Resources Report 103. and application of the HEC-HMS model: a review. Model. Earth Syst. Environ. 1–23.
Food and Agriculture Organization of the United Nations, Rome, p. 132. https://doi.org/10.1007/s40808-023-01704-7.

10
J.U. Guduru and A.S. Mohammed Environmental Challenges 17 (2024) 101017

Salwa, R., Wardah, T., 2015. Application of HEC-Geo HMS and HEC-HMS as Tassew, B.G., Belete, M.A., Miegel, K., 2019. Application of HEC-HMS model for flow
rainfall–runoff model for flood simulation. ISFRAM 181–192. simulation in the Lake Tana basin: the case of Gilgel Abay catchment, upper Blue
Sampath, D.S., Weerakoon, S.B., Herath, S., 2015. HEC-HMS model for runoff simulation Nile basin. Ethiopia. Hydrol. 6, 21.
in a tropical catchment with intra-basin diversions – case study of the Deduru Oya Todini, E., 2007. Hydrological catchment modelling: past, present and future. J. Hydrol.
river basin, Sri Lanka. Eng.: J. Inst. Eng. 48 (1), 1. https://doi.org/10.4038/ Earth Syst. Sci. 11, 468–482.
ENGINEER.V48I1.6843. Uwizeyimana, D., Mureithi, S.M., Mvuyekure, S.M., Karuku, G.N., Kironchi, G., 2019.
Sardoii, et al., 2012. Calibration of loss estimation methods in HEC-HMS for simulation Modelling surface runoff using the soil conservation service-curve number method in
of surface runoff (case study: amirkabir dam watershed, Iran). Adv. Environm. Biol. a drought prone agro-ecological zone in Rwanda. Int. Soil Water Conserv. Res. 7 (1),
6 (1), 343–348. 9–17. https://doi.org/10.1016/J.ISWCR.2018.12.001.
Schaefli, B., Gupta, V., 2007. Do Nash values have value? Hydrol. Process 21 (ISS 15), Vaze J., Jordan P., Beecham R., Frost A., Summerell G. (2011) Guidelines for Rainfall-
2075–2080. Runoff Modelling.
Sintayehu, 2015. Application of the HEC-HMS model for runoff simulation of Upper Blue Wana G.N.,Tamene A.,Fayera G.T.(2020) Rainfall runoff modeling using HEC-HMS: the
Nile River Basin. J. Hydrol. Current Res 6 (2), 1–8. https://doi.org/10.4172/2157- case of awash Bello Sub-Catchment, Upper Awash Basin, Ethiopia, Vol.9. pp. 68–86.
7487.1000199. DOI: Wang, X., Liu, C., 2023. Flash Floods: Forecasting, Monitoring and Mitigation Strategies.
Strupczewski, W.G., Kochanek, K., & Bogdanowicz, E.. (2017). Historical floods in flood Water 15 (9), 1700. https://doi.org/10.3390/w15091700.
frequency analysis: is this game worth the candle?. 554. https://doi.org/10.1016/J. Zelelew, D., Melesse, A., 2018. Applicability of a spatially semi-distributed hydrological
JHYDROL.2017.09.034. model for watershed scale runoff estimation in Northwest Ethiopia. Water 10, 923.
Subramanya, 2008. Engineering Hydrology, 3rd edn. Tata McGray-Hill Publishing Zhang, J., Li, Q., Gong, H., Li, X., Song, L., Huang, J., 2010. Hydrologic Information
Company, New Delhi. Extraction Based on Arc Hydro Tool and DEM. In: Proceedings of the 2010
Tahmasbinejad, H., Feyzolahpour, M., Mumipour, M., Zakerhoseini, F., 2012. Rainfall- International Conference on Challenges in Environmental Science and Computer
runoff simulation and modeling of Karon River using HEC-RAS and HEC-HMS Engineering, Wuhan, China, 6–7 March.
models, Izeh District, Iran. J. Appl. Sci. 12, 1900–1908.

11