Untitled

TABLE OF CONTENTS
List of Table.............................................................................................................................. iv
List of Figure............................................................................................................................. v
ACRONYMS AND ABBREVIATIONS......................................................................................vi
EXECUTIVE SUMMARY.........................................................................................................vii
1. INTRODUCTION............................................................................................................... 1
1.1. Background................................................................................................................ 1
1.2. Objectives...................................................................................................................1
1.3. Scopes of the study....................................................................................................2
1.4. Review of the previous study......................................................................................3
1.4.1. Omo Gibe integrated Resources Development Master Plan Study Project
(Richard Woodroofe & Associate,1995).............................................................................3
2. CHARACTERISTICS OF BONGA CATCHMENT..............................................................5
2.1. Catchment characteristics..........................................................................................5
2.1.1. Slope of the study catchment..............................................................................5
2.1.2. Land use/cover....................................................................................................6
2.2.3. Soil....................................................................................................................... 6
2.2.4. Rainfall and Temperature....................................................................................7
3. DATA COLLECTION, PROCESSING, AND ANALYSIS....................................................8
3.1. Data Collection........................................................................................................... 8
3.1.1. Availability of Hydro-meteorological data................................................................8
3.1.2. Biophysical data....................................................................................................10
3.2. Data Processing and Quality Control............................................................................10
3.2.1. Imputing missing data...........................................................................................10
3.2.3. Hydrological data processing................................................................................15
3.3 Analysis of climatic variables.........................................................................................16
3.3.1. Rainfall data processing.........................................................................................16
3.3.2. Rainfall Frequency Analysis (RFA)........................................................................19
3.3.4. Other climatic parameters......................................................................................24
4. Estimation of Flow and Sediment....................................................................................25
4.1. Estimation of flow at different water abstractions points...............................................25
4.1.1. Description of SWAT model...............................................................................25
4.1.2. SWAT Input.......................................................................................................26
4.1.3. Model calibration schemes................................................................................27
4.2. SWAT model results: Flow and Sediment.........................................................28
4.3. Estimating Dependable Flow.............................................................................32
5. FLOOD and LOW-FLOW ESTIMATION.........................................................................34
5.1. Flood estimation using statistical flood frequency analysis.......................................34
5.1.1. Selection of the best-fit regional frequency distribution......................................34
5.1.2. Prediction of flood at the weir site (Ungauged site)............................................34
5.1.3. Food Frequency Analysis Results.....................................................................35
5.2. Low-flow and baseflow analysis................................................................................37
6. DESIGN FLOOD ESTIMATION FOR CROSS-DRAINAGE.............................................40
6.1. Design Flood Estimation for Cross-Drainage Structures...........................................40
6.1.1. SCS Method......................................................................................................41
6.2. Flood estimation for the cross drainage structures...................................................44
7. CONCLUSION AND RECOMMENDATION.....................................................................46
REFERENCE.......................................................................................................................... 47
ANNEX................................................................................................................................... 48
List of Table
Table 3. 1: List of meteorological stations representing the study catchment...........................8

Table 3. 2: Hydrological station relevant to the Project.............................................................9
Table 3. 3: Locations of the proposed headwork site................................................................9
Table 3. 4: Statistical test results in time series data of Bonga Station...................................14
Table 3. 5: Statistical test results in time series data of Shadatura Station.............................14
Table 3. 6: Statistical test results in time series data of Guma Nr. Andarach. NS means not
significant................................................................................................................................ 15
Table 3. 8: The descriptive statistics of the daily rainfall amount in the selected rain gauge
stations for the period 1990-2018...........................................................................................17
Table 3. 9: Long-term mean monthly, annual and dependable rainfall (mm) for each stations
............................................................................................................................................... 18
Table 3. 10: Annual maximum daily rainfall depth in the selected rain gauge stations for the
period 1990 to 2018................................................................................................................ 20
Table 3. 11: List of candidate distributions used to model extreme rainfall retrieved from
Bonga station. The best fit distribution is in Bold.....................................................................21
Table 3. 12: Estimation of high-quantile extremes for different return periods for the best 13
distributions, in order............................................................................................................... 22
Table 3. 13: Long-term mean monthly climate data for the study catchment .The relative
humidity, wind speed and solar data were obtained from remotely sensed datasets (CFSR). 24
Table 4. 1: Optimized SWAT model Parameters for the study catchment...............................29
Table 4. 2: Model performance using various performance metrics during the calibration
period 1990-1995.................................................................................................................... 29
Table 4. 3 : Long-term mean monthly flow and 90 % dependability for each water abstraction
points, options. Option 1 was selected for further analysis......................................................31
Table 4. 4: The long –term mean monthly sediment estimating using SWAT model (m3)......31
Table 4. 5: 90 % dependable flow (MCM) estimated at monthly timescale............................32
Table 5. 1: Set of candidate distributions for modelling floods at Dincha Gauging station.......35
Table 5. 2: Annual Maximum series flow at Dincha Gauging station.......................................35
Table 5. 3: High-quantile flood for different return period at gauging station and weir site......36
Table 5. 4: Low flow analysis..................................................................................................38
Table 5. 5: Baseflow index (BFI) at the selected option site, weir..........................................39
Table 6. 1: Location of cross-drainage points and their corresponding catchment area. The
projection is WGS-UTM-37N...................................................................................................41
Table 6. 2: Estimated high quantile extreme rainfall for different return periods at Bonga
station, which was used as input for estimating peak discharges...........................................42
Table 6. 3: Flood estimation for the cross drainage structures located at different points.......44
List of Figure
Figure 1. 1: Location map of Ethiopian major river basins (A), Omo-Gibe River Basin (B) and
the study catchment (C)........................................................................................................... 2
Figure 2. 1: Spatial distribution of terrain slope of Bonga catchment.........................................5
Figure 2. 2: Land use/cover map of Bonga Catchment.............................................................6
Figure 3. 1: Location of the hydro-meteorological stations used in the study along with the
water abstraction options.......................................................................................................... 9
Figure 3. 2: Estimation of missing daily rainfall data for all the stations during 1990-2018......11
Figure 3. 3: Estimation of missing daily rainfall data for the selected six stations during 1990-
2018........................................................................................................................................ 11
Figure 3. 4: Correlation matrix in precipitation data for the selected stations..........................12
Figure 3. 5: The Double Mass Curve for checking the consistency of the selected rain gauge
stations................................................................................................................................... 13
Figure 3. 6: Long-term mean monthly rainfall and streamflow of the study catchment............16
Figure 3. 7: Long-term Annual Rainfall for the selected rain gauge stations along with their
trends...................................................................................................................................... 18
Figure 3. 8: Autocorrelation of the annual maximum daily rainfall events for the time lag
between 1and 9. Black dots represent the values for each station, while red-filled circles their
average................................................................................................................................... 19
Figure 3. 9: High-quantile extreme rainfall estimated from Bonga station for different return
periods with uncertainty bounds at 95 % confidence level (purple color)...............................22
Figure 3. 10: CDF of Bonga Station as determined from 17 candidate distributions...............23
Figure 3. 11: PDF of Bonga Station.......................................................................................23
Figure 4. 1: SWAT model setup and sub watershed discretization.........................................27
Figure 4. 2: Global Sensitivity analysis....................................................................................28
Figure 4. 3: Observed Vs Simulated discharge at monthly timescale for the period 1990-1995.
The green color shows the 95 PPU........................................................................................30
Figure 4. 4: Observed vs simulated flow for the whole period 1990-2003...............................30
Figure 4. 5: Flow Duration curve for the selected option site, weir, at daily timescale.............32
Figure 4. 6: Flow Duration curve for the selected option site, weir, at monthly timescale........33
Figure 5. 1: High-quantile flood estimated from Dincha gauging station for different return
periods with uncertainty bounds at 95 % confidence level (purple color) 36
Figure 5. 2: CDF for Dincha gauging station with uncertainty bounds at 95 % confidence level
(purple color) 37
Figure 5. 3: 1-day Low-flow frequency analysis for different return levels 38
Figure 5. 4: Streamflow and baseflow hydrograph for the period 1990-2003 39
Figure 6. 1: Location of cross drainage points and curve number (CN) of the command area 40
Figure 6. 2: IDF curve for Bonga Station 45
ACRONYMS AND ABBREVIATIONS
AFSIS Africa Soil information service
ArcGIS Arc Geographic Information System (Software package)
CFSR Climate Forecast System Reanalysis
CN Curve Number
DEM Digital Elevation Model
ECDSWC Ethiopian Construction Design and Supervision Works Corporation
ESGI Ethiopian space science and geospatial institute
EMA Ethiopian Mapping Agency
EMI Ethiopian Meteorological Institute
FAO Food and Agricultural Organization
ISRIC International Soil Reference and Information Centre
ITCZ Inter-Tropical Convergence Zone
Km Kilometre
Km2 Square kilometre
L Length
m meter
m3/sec Meter cube per second (cms)
masl Mean above sea level
MICE Multiple Imputation by Chained Equation
MCM Million cubic meter
MoWE Ministry of Water and Energy
NASA National Aeronotic and Space Adminstration
RF Rainfall
RFFA Regional Flood Frequency Analysis
RT Return Time (return period)
S Slope steepness
SCS Soil Conservation Service
SRTM Shuttle Radar Topographic Mission
ToR Terms of Reference
UTM Universal Transverse Mercator
WEDSWS Water and Energy Design and Supervision Works Sector
WMO World Meteorological organization
EXECUTIVE SUMMARY
This climatic and hydrological study aims to provide reliable hydro-meteorological data of the
study catchment that could be used as input for planning and designing water supply scheme,
and flood protection works. We used the state-of-the- art hydrologic model and rainfall
frequency analysis to determine flow at the selected water abstraction point and to quantify
floods for different return periods.
Based on the findings of this study, the following conclusion and recommendation are drawn;
 The mean annual rainfall of the study catchment is estimated to be 2217 mm.
 The long-term mean annual flow at the weir site is 209 MCM with the highest flow
recorded in the August.
 The 90 % long-term mean annual dependable flow at the weir site is 69.4 MCM with the
highest dependable flow was observed in August and lowest in February.
 Compared to the maximum demand (around 1.8 MCM), water deficit is noted in the
month January, February and March but not for other months. It is thus important to deal
with this water scarcity issues in those months by using the ground water as a source
 The estimated flood quantiles at 25, 50 and 100 years return period are 67, 78 and 91
cms, respectively.
 The 1-day and 7-day low flow quantiles at 10 years return period is 0.27 and 0.54 cms,
respectively
 The baseflow index is estimated to be 0.58, indicating that a large fraction of streams
obtain the majority of their streamflow from baseflow, and demonstrating the importance
of the baselow component of streamflow.
 The design discharge for the cross drainage structures for 25 years return period ranges
from 5.4 to 44.8 cms.
 The long-term mean annual sediment yield at the weir site is estimated to be 86,000 m3
 Overall, the present study provides flow and sediment rates required for designing weir
and its appurtenant structures for water supply purpose. However, these requirements
may change as additional information about local meteorology becomes available, thus
changing the properties of the likely extreme events. Further, as the watershed changes
due to development or natural shifts, the volume of runoff and the extreme event may
change, which may change the results of this study. Therefore, it is very important to
collect 5-years’ time-series flow and sediment data for the specific project site before
constructing the weir. Moreover, various watershed management intervention measures
shall be implemented at the catchment scale to reduce sediment inflow into the
reservoir.
1. INTRODUCTION
1.1. Background
Ethiopia has12 river basins with an annual runoff volume of 122 billion m3 of water and an
estimated 2.6-6.5 billion m3 of groundwater potential. This corresponds to an average of 1,575
m3 of physically available water per person per year, a relatively large volume. However, due to
large spatial and temporal variations in rainfall and lack of storage, water is often not available
where and when needed (IWMI 2007). Only about 3% of water resources are used, of which
only about 11% (0.3% of the total) is used for domestic water supply.
Access to water supply and sanitation in Ethiopia is amongst the lowest in Sub-Saharan Africa
and the entire world. While access has increased substantially with funding from foreign aid,
much still remains to be done. Therefore, ensuring access to clean water and sanitation is
essential. Taking this background, the Bonga water supply project is designed to improve
access to safe, reliable water supplies and sanitation services, recognizing that lack of access
to these resources are a barrier to growth and health. This project is supplied with water that is
directly diverted from the river using a diversion structure. The Bonga water supply project is
located in the south west region Bonga town. The weir site is located in Dincha river, one of the
tributaries of Guma River and found in Dakiti kebele under Gimbo wereda. It is geographically
located at 7.23 0 N and 36.32 0E.
This report generally highlights the estimation of meteorological and hydrological design
parameters that could be used as input for the planning, operation and management of the
water supply scheme. It also gives insight on the delicate balance between water supply and
increasing demand of water owing to rapid population growth, urbanization, and increasing
water needs from agriculture, industry, and energy. The following issues are also covered in this
sectorial report: a review of previous documents, data collection and processing, estimation of
surface water availability and flood quantile estimation for different hydraulic structures.
1.2. Objectives
The main objective of this study is to derive meteorological and hydrological design
parameters that could be used as input for the planning, operation and management of the
water supply scheme
The specific objectives are to;
 Revise the meteorological and hydrological design parameters and variables determined
by the previous studies
 Estimate dependable rainfall of the project site using long-term time series rainfall data
 Developing Intensity-Duration-Frequency (IDF) for the selected rainfall stations
 Estimate high-quantile extreme rainfall for different return periods using the state-of-the-
art rainfall frequency analysis
 Estimate monthly and annual flow at the weir site and other options using the state-of
the-art hydrologic model
 Estimate sediment inflow to the weir site using an appropriate method
 Estimate flood quantiles at the weir site and cross drainage structures for different return
levels
 Estimate 1-day and 7-day low flow quantiles at the weir site for different return periods
Figure 1. 1: Location map of Ethiopian major river basins (A), Omo-Gibe River Basin (B) and the
study catchment (C).
1.3. Scopes of the study

Based on the request of the client and identified gaps of the previous study, the scope of this
study includes;
 Review of previous hydrological studies undertaken in the project area and identify gaps
 Assess historical records of streamflow and precipitation data available within and
nearby project area
 Bridge streamflow data gaps using the state-of-the-art hydrologic model
 Quantify long-term mean monthly flow at different options for water abstraction
 Analyze supply and demand for each optional site and select the better option for further
analysis.
 Estimate rainfall and flood quantiles for different return periods using the state-of-the-art
frequency analysis
 Quantify long-term mean monthly sediment inflow to the weir site
 Quantify low flow and baseflow index of the study catchment
1.4. Review of the previous study

1.4.1. Omo Gibe integrated Resources Development Master Plan Study Project
(Richard Woodroofe & Associate,1995)
The Gibe river rises just north of latitude 9 0N and longitude 370E on the Ethiopian plateau. The
Gibe head water are at an elevation of about 2200m. Although there are some important
tributaries from different directions, the general direction of Gibe river flow is southwards,
towards the Omo river. The Gibe river is known as the Omo river in its lower reaches,
southwards from the confluence with the Gojeb river. From the confluence with the Danchiya
river the Omo river changes its character in traversing the flood peak plain leading to Lake
Turkana as well defined meandering river channel. The Gojeb river is an important tributary of
the Omo-Gibe from the west. It drains upland that have been less intensively used for
agriculture than the area around Jima to the north-east. The prevailing vegetation of the Gojeb
headwaters is wooded shrub-land. To the south of the Gojeb catchment around Bonga, are the
headwaters of Sherma, Guma and Danchiya rivers.
The climate of the Omo-Gibe river basin varies from a hot arid climate in the southernmost part
of the flood plain to a tropical humid one in the highlands that includes the extreme north near
Bako, the areas surrounding Jima and around the headwaters of the Gojeb river. The main
sources of moist air is from the Atlantic Ocean, to the south-west, the eastern parts of the
highlands are more or less rain shadowed. The north-east of the basin, where the rainfall is
heavy and the seasonal nature of the rainfall is very marked with July and August having
particularly high amounts. The north-west represented by Bonga station, where the project area
located, the seasonal effect is not as extreme and the rainfall is more-or-less constant from April
to September. The mean annual temperature in the basin varies from less than 18 0C in the
highlands to the west of it to over 22 0C to the south. Monthly PET was calculated by generally
accepted Penman method. The whole basin divided accordingly in to four zones, the project
area located in zone 2 with annual excess greater than 200mm for more than 4 months.
The simulated flow records in the Omo-Gibe river basin produced for 25 tributary sub-
catchments and 14 riparian runoff areas. The simulation of flows was based on 22 gauging sites
for which the records available. It is believed that flows upstream of Abelti gauging station,
located in the upper basin, reasonably well simulated. Accordingly, the mean annual runoff of
Gojeb river at Shebe and at Gibe confluence was estimated and is in the order of 50 and 86
MMC respectively.
Flood magnitudes were estimated based on regression and rainfall-runoff models. Flood growth
curves were derived for developing regional regression curves. Accordingly, a regression
formula developed, which broadly agrees with annual flood series of gauged catchments. While
for rainfall-runoff model SCS curve number method applied to estimate execs rainfall. With the
excess rainfall a flood hydrograph have been generated. The reconnaissance report of this
study included maps of sediment risks and of forested areas. The information from these maps
has been showing the areas with high erosion potential in the east of the basin, with high
sediment yields and forested areas to the west having low sediment loads. On the basis of
these data the basin subdivided in to four categories: high potential, medium potential, low
potential and low potential (forested). As per this category the rate of sedimentation were taken
as 500t/km2/yr, 100 t/km2/yr, 50 t/km2/yr and 20 t/km2/yr for high, medium, low and low (forested)
sediment potentials.
2. CHARACTERISTICS OF BONGA CATCHMENT

2.1. Catchment characteristics
Bonga catchment is geographically bounded by latitude N 70 5’and 7028’ and by longitude E 360
1’ and 360 50’ with a mean elevation of 2560.8 m (Figure 1.1). Considering the outlet of the
catchment, the upper part of the Bonga catchment is characterized by high altitude whereas the
lower and middle parts are relatively dominated by low altitude (Figure 1.1). The area of the
catchment at the weir site is 155.8Km2.
2.1.1. Slope of the study catchment
Bonga catchment is characterized by different landforms which are ranged from flat plains (0-
3% slope), undulating plains (3-8%), rolling land (8-15%), hilly (15-35%), steep hilly plains (35-
50%) and mountainous (>50%) as shown in Figure 2.1.
Figure 2. 1: Spatial distribution of terrain slope of Bonga catchment

The slope ranges and area coverage of each landform were estimated from the digital elevation
model (DEM). The average slope of the Bonga catchment is estimated to be 21.4 % and falls
under the hilly plains landform.
2.1.2. Land use/cover

The major land use /land cover of the Bonga catchment along with their spatial distribution is
presented in Figure 2.2. In Bonga catchment, seven major land use/ covers are identified
namely Annual Cropland, Dense Forest, Moderate Forest, Sparse Forest, Woodland, Open
Grassland, and Closed Grassland. Moderate Forest (42.5%) is the dominant land cover of the
study area followed by Annual Cropland (39%). Overall, Forest covered 55% of the study
catchment
Figure 2. 2: Land use/cover map of Bonga Catchment
2.2.3. Soil
Based on the FAO soil classification system, about 3 major soil types are identified in the Bonga
catchment: Dystric Fluvisols, Dystric Nitisols, Orthic Acrisols. Dystric Fluvisol (59 %) is the
dominant soil type followed by Dystric Nitisols (29 %). Wide variations of soil types are observed
in the study catchment spatially (Figure 2.3). The definition & characterization of these FAO soil
groups is well documented in the FAO database.
Figure 2.1: Soil Map of the study catchment
2.2.4. Rainfall and Temperature
The long-term mean annual rainfall of the study area is 2217 mm with an average temperature
of 19 OC. Based on the long-term mean monthly rainfall estimated from the selected rain gauge
stations, the region’s rainfall pattern is bimodal that is two rainfall peaks occurred in the primary
rainy season (from March to April) and the secondary rainy season (from October to
November). These rainy seasons are usually controlled by the movement of the Inter-Tropical
Convergence Zone (Seleshi and Zanke, 2004).
3. DATA COLLECTION, PROCESSING, AND ANALYSIS
3.1. Data Collection
Surface water resources assessment is fundamentally a data/information collection exercise
followed by the presentation of this data/information to provide value-added products on the
spatial and temporal characteristics of the available water supplies and their usage. Therefore,
data collection across the two areas (hydro-meteorological and biophysical) is the fundamental
basis for surface water resources assessment.
3.1.1. Availability of Hydro-meteorological data
The required climatic and hydrometric data were collected from the Ethiopian Meteorological
Institute (EMI) and the Ministry of Water and Energy (MoWE), respectively. About eight rain
gauge stations were found within and outside the catchment area. Based on their long-term
records, out of eight stations, 6 stations were selected for studying the hydrology and
climatology of the study catchment. Table 3.1 summarizes the characteristics of the stations
located within and outside the study catchment. The stations categorized under class III provide
three meteorological elements, i.e. maximum air temperature of the day, minimum air
temperature of the day, and total rainfall amount in 24 hours. The fourth class stations record
only one climatic variable, i.e. the total rainfall in 24 hours. Regarding hydrological data, there
are 4 gauging stations available in the project area (Table 3.2). The location of the hydro-
meteorological stations used for this study is shown in Figure 3.1. In order to quantify flow at
different proposed water abstraction points (Figure 3.1 and Table 3.3), the state-of-the-art
hydrological model was calibrated at Guma near the Anderacha gauging station. It is important
to note that we selected option 1 (weir site) based on supply and demand analysis but
quantifying flow at the remaining option site would support our client in planning water resource
development projects in the foreseeable future.
Table 3. 1: List of meteorological stations representing the study catchment
Station Class Elevatio Location (Degree) Input
n (m) Lat Long Station Hydrologic Rainfall Frequency
representing al modeling analysis
Bonga III Opt 4 and 5
1599 7.28 36.24 catchments  
Chiri* IV 1459 7.14 36.18
Goga Kemise IV Opt 2
1388 7.43 36.38 catchment 
Kemise IV 1450 7.33 36.13 
Keresi IV 1898 7.23 36.23 
Modyo Gombora* IV 7.20 36.29
Shedatura IV Opt 1and 6
2478 7.28 36.38 catchments 
Wushwush IV Opt 3
1975 7.31 36.13 catchment 
*excluded in the analysis because of short record length
Table 3. 2: Hydrological station relevant to the Project
S. Catchmen
Sub-Catchment River Station North East Km2
No t
1 80535 19303 242.
GIBE-OMO Middle Ghibe Gecha Gecha Nr. Bonga
8 9 4
2 80624 19458 129.
GIBE-OMO Middle Ghibe Sheta Sheta @ Bonga
0 0 6
3 Guma Nr. 79116 19623 772.
GIBE-OMO Middle Ghibe Guma
Andaracha 7 8 2
4 Dinch 79654 19986 218.
GIBE-OMO Middle Ghibe Dincha @ Bonga
a 2 8 5
Table 3. 3: Locations of the proposed headwork site

River UTM Coordinates Headwork Remark
Headwork Types
S. No Name
Options Easting Northing
1 Option -1 Guma R. 204200.00 800136.00 Weir Selected

2 Option -2 Gecha R. 192117.48 809131.35 Dam Not Selected
3 Option -3 Gecha R. 195693.17 807272.90 Dam Not Selected
4 Option -4 Berta R. 199125.705 804102.36 Dam Not Selected
5 Option -5 Kasha R. 191934.95 806037.95 Dam Not Selected
6 Option -6 Guma R. 200148.45 797178.27 Pump Not Selected
Figure 3. 1: Location of the hydro-meteorological stations used in the study along with the water
abstraction options
3.1.2. Biophysical data
The biophysical data that are relevant for studying catchment hydrology were collected from
different sources. These spatial data include land use (section 2.12), Digital Elevation Model
(DEM), and soil (section 2.2.3). The 30 m Shuttle Radar Topography Mission (SRTM) DEM was
used to create the stream network of the study catchment. This data was collected from the
NASA website. To represent the land cover of the study catchment, the 2016 land cover map of
the study catchment, which was collected from the Ethiopian space science and geospatial
institute, was used. Regarding soil data, the Africa Soil information service (AFSIS) soil map
with 250 m spatial resolution (Leenaars et al., 2014) was used to represent the physical and
chemical soil properties of the study catchment. This data was collected from the International
Soil Reference and Information Centre (ISRIC) soil database. These datasets were used to
describe the physical catchment characteristics of the study area (see section 2.1)
3.2. Data Processing and Quality Control

Checking the quality of the hydro-meteorological data is a prerequisite before applying them for
operational applications. In this section, we present the climatic and hydrologic data processing
and quality control.
3.2.1. Imputing missing data
The outcome of data analysis depends on the quality and completeness of data. The accurate
planning and management of water resources depend on the presence of consistent and
reliable climatic (rainfall) and hydrometric (streamflow) data. In cases where it has not been
possible to accurately and consistently record these datasets, it is necessary to impute these
datasets before applying them in the hydrological models. The estimation of missing data in
hydrological studies is necessary for the timely implementation any water resource development
projects.
Several techniques are available for imputing missing data. In this study, the imputation of
missing data was performed using Multiple Imputation by Chained Equation (MICE; van Buuren
and Groothuis-Oudshoorn (2011). MICE is a robust, informative method of dealing with missing
data in datasets. The procedure ‘fills in’ (imputes) missing data in a dataset through an iterative
series of predictive models. In each iteration, each specified variable in the dataset is imputed
using the other variables in the dataset. These iterations should be run until it appears that
convergence has been met. The accuracy of the imputations depends on the information
density in the dataset. A dataset of completely independent variables with no correlation did not
yield accurate imputation. Therefore, it is very important to determine the correlation between
the selected stations. For this purpose, the Person correlation coefficient (R) was determined at
monthly timescale and this was performed in R programming. Moreover, efforts were made to
visualize the number of missing data observed in each selected meteorological stations. Figures
3.2 and 3.3 show the missing data percentage estimated from the stations located in the study
catchment. Results show that Chiri and Modyo Gamboro stations showed the highest
percentage of missing records for the period 1990-2018 (Figure 3.2). These stations were thus
excluded from the analysis. In contrast, Bonga and Wushiwush stations produced the lowest
missing data percentage. The correlation between the selected stations is strong as
demonstrated by the highest R values (Figure 3.4).
Figure 3. 2: Estimation of missing daily rainfall data for all the stations during 1990-2018
Figure 3. 3: Estimation of missing daily rainfall data for the selected six stations during 1990-2018
..
Figure 3. 4: Correlation matrix in precipitation data for the selected stations.
3.2.2. Data Quality Control

Most of the hydraulic structures have been designed and operated under the assumption of
stationarity. Stationarity means that the hydrological and climate variables fluctuate within an
unchanging envelope of variability. However, in reality, anthropological interventions and climate
change would cause significant impacts in changing these variables. Even though considering
the hydrologic and climatic nonstationary assumption is central for designing hydraulic
structures, the uncertainty associated with the quantification of human actions on the climate
system and the physical environment poses a major challenge to account these factors for
planning and designing various water infrastructures. For this reason, this study used the
assumptions of stationarity to quantify climatic and the hydrologic variables of the study
catchment. Therefore, it is very important to check the stationarity of time-series hydro-climatic
variables before applying them for different operational applications.
Several methods have been widely applied to check stationarity in time series data: time series
plot (visual inspection), statistical test (parametric and non-parametric test), and double mass
curve. The first and second methods were applied to check the temporal consistency and
randomness in time series data and the latter was used for checking the spatial homogeneity.
The present study used the second and third methods to analyze the quality of the data
quantitatively. The Double Mass Curve was implemented using the DMC package in R
programming whereas the statistical tests were employed by using TREND statistical software,
which was developed by CRC for catchment hydrology in Australia.
The double mass curve was used to check the consistency of many kinds of climatic and
hydrologic data by comparing data for a single station with that of a pattern composed of the
data from several other stations in the area. The double-mass curve is used to adjust
inconsistent precipitation data as well. To perform a stationarity test using DMC, we performed
DMC for the selected six rain gauge stations. The spatial homogeneity test (DMC, Figure 3.5)
results showed that there is no significant break observed in the time series of the selected rain
gauge stations as explained by the highest R2 value. Therefore, all the selected rain gauge
stations are consistent.
Figure 3. 5: The Double Mass Curve for checking the consistency of the selected rain gauge stations
Moreover, TREND software was applied to test trend, change, and randomness in hydrological
and other time-series data. The present study used 12 statistical tests, based on the
WMO/UNESCO Expert Workshop on Trend/Change Detection, to detect changes in time series
data on rainfall at 1%, 5%, and 10% significant levels. The 5 % significant level was selected to
accept or reject the null hypothesis, i.e., “the time series data are inconsistent”. The following
statistical tests were used in this study;
 Mann-Kendall (non-parametric test for trend)
 Spearman's Rho (non-parametric test for trend)
 Linear Regression (parametric test for trend)
 Distribution-Free CUSUM (non-parametric test for step jump in mean)
 Cumulative Deviation (parametric test for step jump in mean)
 Worsley Likelihood Ratio (parametric test for step jump in mean)
 Rank-Sum (non-parametric test for difference in median from two data periods)
 Student's t (parametric test for difference in mean from two data periods)
 Median Crossing (non-parametric test for randomness)
 Turning Points (non-parametric test for randomness)
 Rank Difference (non-parametric test for randomness)
 Autocorrelation (parametric test for randomness).
Both the parametric and non-parametric statistical test results (Median crossing, turning points
and Rank difference) revealed that all the selected rain gauge stations are consistent and
random. (Tables 3.4 and 3.5).The Man-Kendall statistical test results showed that an increasing
trend was observed in the Annual rainfall data of Bonga station and this trend is statistically
significant at 5 %.In contrast, there is no trend observed in Shadatura Station for the period
1990-2018. See the results of other stations in annex 1-4.
Table 3. 4: Statistical test results in time series data of Bonga Station.

Test type Critical values (Statistical table)
Description Test statistic a=0.1 a=0.05 a=0.01 Result
Auto Correlation 1.697 1.645 1.96 2.576 S (0.1)
Linear regression 2.471 1.703 2.052 2.771 S (0.05)
Parametric Student's t -2.288 1.701 2.048 2.763 S (0.05)
Cumulative deviation 1.391 1.118 1.238 1.456 S (0.05)
Worsley likelihood 3.157 2.866 3.198 3.876 S (0.1)
Mann-Kendall 2.195 1.645 1.96 2.576 S (0.05)
Spearman's Rho 2.075 1.645 1.96 2.576 S (0.05)
Distribution-Free CUSUM 6 6.57 7.324 8.778 NS
Non Parametric Rank Sum -2.248 1.645 1.96 2.576 S (0.05)
Median Crossing 0.378 1.645 1.96 2.576 NS
Turning Point -1.447 1.645 1.96 2.576 NS
Rank Difference -0.189 1.645 1.96 2.576 NS
Table 3. 5: Statistical test results in time series data of Shadatura Station.

Test type Critical values
(Statistical table)
Description Test statistic a=0.1 a=0.05 a=0.01 Result
S
Auto Correlation
2.444 1.645 1.96 2.576 (0.05)
Linear regression 0.083 1.703 2.052 2.771 NS
Parametric
Student's t -0.721 1.701 2.048 2.763 NS
Cumulative deviation 0.795 1.118 1.238 1.456 NS
Worsley likelihood 1.821 2.866 3.198 3.876 NS
Mann-Kendall -0.882 1.645 1.96 2.576 NS
Spearman's Rho -1.071 1.645 1.96 2.576 NS
Distribution-Free CUSUM 7 6.57 7.324 8.778 S (0.1)
Non Rank Sum 0.415 1.645 1.96 2.576 NS
Parametric Median Crossing 1.134 1.645 1.96 2.576 NS
Turning Point 0.455 1.645 1.96 2.576 NS
S
Rank Difference
-2.395 1.645 1.96 2.576 (0.05)
NS: means not significant; S means significant
3.2.3. Hydrological data processing
Like the climatic data, the quality of the hydrological data was checked. The consistency check
was performed using different parametric and non-parametric statistical tests (Table 3.5). The
missed flow data was also imputed stochastically using the univariate time series imputation
method, I.e. ARIMA model. Moreover, the mean monthly rainfall of the study catchment was
plotted against the long-term monthly streamflow data to check the consistency of flow with
rainfall (Figure 3.6). The statistical results revealed that the flow data of Guma station are
random and consistent as demonstrated by turning point, rank difference and median crossing
test statistics. Moreover, the Mann-Kendall statistical test result show that there is no statistical
significant trend observed in the gauging station. Figure 3.6 shows the visual inspection of the
long-term mean monthly discharge data with respect to monthly rainfall for the period1990-
2003. Results revealed that a temporal consistency between discharge and runoff was
observed in most of the months but there is some uncertainties on the observational data in the
month of June.
Table 3. 6: Statistical test results in time series data of Guma Nr. Andarach. NS means not
significant.
Critical values
Test (Statistical table)
Statistics statistic a=0.1 a=0.05 a=0.01 Result
Mann-Kendall 0.985 1.069 1.274 1.674 NS
Spearman's Rho 1.117 1.069 1.274 1.674 S (0.1)
Cusum 4 4.565 5.089 6.099 NS
Rank Sum -1.533 1.069 1.274 1.674 S (0.05)
Student's t -1.461 1.771 2.16 3.012 NS
Turning Point 0.679 1.069 1.274 1.674 NS
Rank Difference 0.909 1.069 1.274 1.674 NS
Auto Correlation 0.308 1.069 1.274 1.674 NS
140 300
120
250
100
200
80
Rainfall (mm)
Runoff (mm)
150
60
100
40
50
20
0 0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Rainfall Runoff
Figure 3. 6: Long-term mean monthly rainfall and streamflow of the study catchment
3.3 Analysis of climatic variables
3.3.1. Rainfall data processing
After imputing missing data, the daily rainfall amount of 6 rain gauge stations is aggregated into
monthly and annual rainfall total. These data are very important to estimate dependable rainfall
and are mainly used as input to estimate effective rainfall. In contrast, the annual maximum
daily rainfall values are computed considering only those values different from NA (missing
data). The summary descriptive statistics of the daily rainfall total of each station are presented
in Table 3.8
The computed descriptive statistics are mean, median, variance, standard deviation, coefficient
of variation, skewness, kurtosis, and the maximum rainfall amount for each station. Results
showed that the mean of daily rainfall estimated for all the selected stations ranges from 3.69
mm to 6.06 mm with the highest mean value recorded from Shadatura station while the lowest
mean of 3.69 mm from Gonga Kemisse Station. Based on the coefficient of variation results, the
highest variability of daily rainfall was observed in Gonga Kemisse station while the lowest was
observed in Keresi stations. Overall, the results of the coefficient of variation show that the
values for each station are quite homogeneous where the range is between 2.1 and 3.0.
The coefficient of skewness is used to verify the degree of asymmetric distribution around the
mean. The values of skewness lie between 2.67 and 3.61 where these positive results indicate
that all the skewness are positively skewed for all stations. Gonga Kemisse station produces the
highest value of skewness that gives a clear indication that this station is very obviously skewed
and the asymmetric tail is extending to the right while Shadatura station with the smallest
skewness. The value of kurtosis can be used to determine the relative peakness or flatness of
the distribution. The values of kurtosis are all positive in the range of 20.1 to 6.34. The
maximum daily rainfall amount is ranged between 169 mm per day and 172 mm per day.
Table 3. 7: The descriptive statistics of the daily rainfall amount in the selected rain gauge stations for the
period 1990-2018.
Descriptive Stations
Statistics Bonga Gonga Kemisse Kemisse Keresi Shadatura Wushiwush
Median 0.00 0.00 0.00 0.90 1.20 0.60
Mean 4.80 3.69 5.86 6.02 6.06 4.81
3rd Qu. 7.30 4.60 8.40 9.00 8.50 6.30
Max. 80.10 102.00 146.60 169.20 73.40 72.70
IQR 7.30 4.60 8.40 9.00 8.50 6.30
sd 7.86 7.37 10.33 9.65 9.87 8.26
cv 1.64 2.00 1.76 1.60 1.63 1.72
Skewness 2.41 3.61 2.97 2.64 2.37 2.56
Kurtosis 7.75 20.10 15.82 13.89 6.34 7.82
The long-term mean monthly, and annual rainfall estimated from each rain gauge station are
also presented in Table 3.9. Results show that the long-term annual rainfall total estimated from
the selected station ranges from 1346 mm ( Gonga Kemisse station) to 2214 mm ( Shadatura
station). Figure 3.7 shows trends in annual rainfall in the selected stations for the period 1990 to
2018. Results revealed that an increasing trend in annual rainfall was observed in Gonga
Kemisse, Bonga, and Kemisse stations while a decreasing trend was observed for the
remaining rain gauge stations.
Table 3. 8: Long-term mean monthly, annual and dependable rainfall (mm) for each stations
Stations Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annu
Bonga 49.3 44.5 115.8 180.5 225.1 206.7 223.4 212.9 209.9 152.1 83.9 50.6 1754
Gonga
Kemisse 26.0 32.8 67.2 95.8 160.9 187.4 202.2 195.4 186.5 109.6 52.1 30.9 1346
Kemisse 39.1 38.0 78.2 142.0 250.1 320.8 326.1 347.3 312.9 184.0 57.1 46.3 2142
Keresi 45.3 38.9 91.1 140.9 256.7 315.1 312.9 339.2 340.3 189.8 89.3 40.3 2199
Shadatura 110.7 100.6 143.1 199.2 248.0 228.1 248.4 253.8 243.9 206.6 111.6 120.2 2214
Wushiwush 44.6 51.7 131.9 191.6 223.5 202.2 194.7 201.7 206.5 153.0 94.5 60.1 1756
Figure 3. 7: Long-term Annual Rainfall for the selected rain gauge stations along with their trends
3.3.2. Rainfall Frequency Analysis (RFA)
To estimate the rainfall depth for a given return period at a given site of interest, the present
study used rainfall frequency analysis (RFA). The RFA, extreme value theory, uses extreme
values extracted from either the annual maximum daily rainfall series (AMS) or peak over
thresholds series (PoT) to estimate high-quantile extremes. The basic assumption of this theory
is that the distribution of the event is identical and independent (Fisher and Tippett, 1928).
Therefore, it is a prerequisite to check the independence of the daily annual maximum rainfall in
each station before implementing RFA. For this purpose, the autocorrelation of the daily
maximum annual rainfall events between the time lag between 1 and 9 was used (Figure 3.8).
Results show that the average autocorrelation value estimated for the selected 6 rain gauge
stations is approximately less than 0.1 and that confirms the independence of the annual daily
maximum rainfall events since ACF value is less than 0.2.
Figure 3. 8: Autocorrelation of the annual maximum daily rainfall events for the time lag between 1and 9.
Black dots represent the values for each station, while red-filled circles their average.
The study used the AMS series for implementing RFA. However, we did not use the PoT series
because of the uncertainty associated with the selection of threshold in PoT series. Since the
effectiveness of AMS-based rainfall frequency analysis relies on the length of data records,
long-term daily rainfall data (29 years) were collected from the EMI, and then the extreme
values were extracted as indicated in Table 3.10 The best-fit distribution, among the set of
candidates, for modeling extreme rainfall are identified using the goodness of fit measure, Root
Mean Square Error (Table 3.11). The RMSE measures the line distance between the actual and
empirical cumulative distribution function. The distribution parameters were estimated by the
method of L- moment as this estimator produced more reliable results than the Maximum
Likelihood method in the presence of small samples (Hosking, 1990). Moreover, this parameter
estimator tends to address the impact of outliers as well. Based on the best-fitted distribution,
high quantiles of daily rainfall were estimated for 2, 5,10,25,50,100 return levels
For rainfall frequency analysis, we selected Bonag station as it represents the catchment of the
cross drainage structures (See section 6). Table 3.11 shows the RMSE value estimated from
the set of candidate distributions to model extreme rainfall for Bonga station. Results show that
the Wakeby distribution is the best-suited distribution to model the annual maximum daily
rainfall series of Bonga station. Using the best fit distribution, high-quantile extremes for different
return periods along with the uncertainty bound at a 95% significant level were computed.
Results show that the estimated rainfall depth for 2,5,10,25,50,and 100 return levels are
46.4mm, 55.9 mm, 62.9mm, 71.7mm , 78.0 mm and 84.1 mm, respectively (Table 3.12 and
Figure 3.9). Moreover, the CDF and PDF of Bonga station are indicated in Figures 3.10 and
3.11
Table 3. 9: Annual maximum daily rainfall depth in the selected rain gauge stations for the period 1990 to
2018
Bonga Gonga.Kemisse Kemisse Keresi Shadatur Wushiwush
year a
1990 61.1 49.7 50 63.1 52.8 56.7
1991 44.5 52.3 69.3 67.4 49 51.2
1992 43.3 53.5 71.3 53.4 72.2 58.3
1993 40.4 57 53 58.2 54 46.4
1994 50 31.4 76.3 72.9 72.2 49.2
1995 39.2 23.5 47.5 58 59 53.7
1996 35.2 60.1 61.8 54 25.5 46.6
1997 51 31.2 76.9 69 57.9 51.1
1998 48.6 36 65.2 69 56.4 54.7
1999 28 46.5 56.5 86 61.4 49
2000 30.1 24.5 46.2 67 46 36.7
2001 40.5 32.4 53.3 75 40.3 52.1
2002 42.8 20.4 62.1 57 46.6 54.1
2003 41.4 48 141.2 64.5 41.7 54.9
2004 48.5 59 39.1 54 20.2 54.6
2005 49.2 64 60.7 55.5 38.2 54.5
2006 63.3 78 72.5 86.5 38.1 72.7
2007 43 52.4 56.9 42.2 40.1 52.5
2008 52.1 69 81.8 52.7 73.4 55.9
2009 65.5 47.6 56.5 169.2 61.4 48.5
2010 53.7 46.5 75.9 70.4 73.4 42.5
2011 80.1 60.1 146.6 57.6 73.4 59.8
2012 48.9 78 60 64.4 62.6 55.9
2013 61.7 58.1 48.2 86 45.4 43.3
2014 46 51 64.1 49 50.6 54.1
2015 45.6 102 72.5 44.7 59 47.4
2016 41 33.4 68.6 44.1 48.6 47
2017 47 48.5 68.6 97 20.1 52.4
2018 60.1 60.1 141.2 51.7 48.6 55.9
Table 3. 10: List of candidate distributions used to model extreme rainfall retrieved from Bonga station.
The best fit distribution is in Bold.
Set of candidate distributions abbreviation RMSE
Wakeby wak 0.030
4-p Asymmetric Exponential Power aep4 0.031
Polynomial Density-Quantile3 pdq3 0.032
Laplace lap 0.033
Generalized Lambda gld 0.033
Generalized Logistic glo 0.037
Gumbel gum 0.039
Generalized Extreme Value gev 0.042
Generalized Normal gno 0.042
Log Normal ln3 0.042
Pearson Type III pe3 0.044
Weibull wei 0.048
Rayleigh ray 0.050
Gamma gam 0.051
Triangular tri 0.060
Generalized Pareto gpa 0.061
Rice rice 0.064
Polynomial Density-Quantile4 Pdq4 0.064
Student t (3-parameter) st3 0.064
Noraml nor 0.065
Exponential exp 0.066
Govindarajulu gov 0.096
Reverse Gumbel revgum 0.106
Table 3. 11: Estimation of high-quantile extremes for different return periods for the best 13 distributions,
in order
Distribution RP.2 RP.5 RP.10 RP.25 RP.50 RP.100
wak 46.37 55.99 62.93 71.68 78.00 84.07
lap 46.99 54.47 60.12 67.60 73.25 78.91
pdq3 46.84 56.01 62.48 71.04 77.56 84.10
glo 46.98 55.96 62.06 70.42 77.24 84.65
gum 46.49 56.42 62.99 71.29 77.44 83.56
gev 46.82 56.72 62.96 70.51 75.88 81.03
gno 46.84 56.72 62.90 70.40 75.80 81.06
ln3 46.84 56.72 62.90 70.40 75.80 81.06
pe3 46.82 56.87 63.04 70.35 75.49 80.39
wei 46.81 57.18 63.26 70.13 74.75 79.01
ray 47.08 57.28 63.09 69.56 73.86 77.79
gam 47.53 57.13 62.62 68.85 73.07 77.01
gpa 46.56 58.38 64.25 69.40 71.95 73.74
Figure 3. 9: High-quantile extreme rainfall estimated from Bonga station for different return periods with
uncertainty bounds at 95 % confidence level (purple color)
Figure 3. 10: CDF of Bonga Station as determined from 17 candidate distributions

Figure 3. 11: PDF of Bonga Station
3.3.4. Other climatic parameters
Other climatic data that are relevant for studying the hydrology and climatology of the study
catchment were collected from the Ethiopia Meteorological Institute. These data include
temperature, humidity, wind speed, and sunshine hour. Table 3.13 shows the monthly long-
term monthly climate data of the study catchment. The minimum temperature of the study
catchment varies from 10.23 to 12.88 oc whereas the maximum temperature lies between 28
and 24 oc. Overall, the average temperature, humidity and wind speed of the study catchment is
19 oc, 59 %, and 1.74 m/s respectively.
Table 3. 12: Long-term mean monthly climate data for the study catchment .The relative humidity, wind
speed and solar data were obtained from remotely sensed datasets (CFSR)
Wind speed
Min Temp Max Temp Relative @2m Solar
Month °C °C Humidity. m/s (MJ/M2)
January 27.43 10.43 0.46 1.79 23.45
February 28.09 11.09 0.41 1.83 24.95
March 28.03 12.14 0.49 1.82 25.62
April 27.41 12.74 0.64 1.76 25.65
May 26.46 12.88 0.69 1.74 24.91
June 26.02 12.58 0.68 1.63 23.51
July 24.66 12.76 0.68 1.58 21.73
August 24.95 12.49 0.67 1.60 22.35
September 25.90 12.54 0.64 1.63 23.92
October 27.18 11.86 0.64 1.81 25.12
November 27.77 11.00 0.59 1.81 23.93
December 28.05 10.23 0.51 1.83 23.18
Average 26.82 11.90 0.59 1.74 24.02
4. Estimation of Flow and Sediment

4.1. Estimation of flow at different water abstractions points
Precise information on streamflow is of great importance for planning and monitoring water
resources schemes related to hydropower, water supply, irrigation, flood control, and
maintaining the ecosystem. Hydrologists encounter challenges when streamflow data are either
unavailable or inadequate at target locations. In this particular study, our study catchment can
be considered as gauged catchment since four hydrological gauging stations are available
within the study area. These gauging stations provide daily flow data for the period 1990 to
2003, which is not enough for the required operational applications. Therefore, the state-of-the-
art the Soil and Water Assessment Model (SWAT) was applied to bridge streamflow data gaps
and to understand the present hydrological process of the study catchment. Using this model,
we estimated flow at different water abstraction points for the period 2004 to 2018 after
calibrating and validating the SWAT model by using observed streamflow data.
4.1.1. Description of SWAT model
The SWAT model is a continuous-time, semi-distributed hydrological and water quality model
that was developed by the USDA Agricultural Research Service in collaboration with Texas
A&M AgriLife Research (Arnold et al., 1998). The model simulates different components of the
hydrological cycle for each Hydrologic Response Unit (HRU) within every sub-watershed
(Arnold et al., 1998; Neitsch et al., 2011). The HRU is the smallest spatial unit of the model,
which consists of homogeneous land use, soil, and slope within each sub-watershed. The
present study used 5 % land use, 5% soil, and 5 % slope thresholds to define HRU and that
results in a total of 215 HRUs from the 12 sub-watersheds. Depending on the purpose, the
SWAT model simulates the hydrological processes (e.g., surface runoff, water yield, actual
evapotranspiration, and nutrient and sediment yields) of a catchment at a daily, monthly, and
annual timescale. These hydrological processes are simulated using the water balance
equation, Eq.4.1, (Arnold et al., 1998; Gassman et al., 2007);
t
SW t =SW O +∑ ( SW 0−R day−E a−W seep−Q gw ) Eq .4 .1
i=1
Where; SWt= the final water content (mm H2O), SWo= the initial soil water content on day i (mm
H2O), t = time, days, Rday= is the amount of precipitation on day i (mm H 2O), Qsurf = is the amount
of surface runoff on day i (mm H2O), Ea= is the amount of evapotranspiration on day i (mm H2O),
Wseep = is the amount of water entering the vadose zone from the Soil profile on day i (mm H 2O),
Qgw= is the amount of groundwater flow on day i (mm H2O)
In the SWAT model, different methods are available for estimating the mentioned hydrological
processes. For example, the model uses the Curve Number (CN) (SCS, 1972) or the Green
and Ampt infiltration (Green and Ampt, 1911) methods to estimate surface runoff. The CN
method required daily rainfall data while the green and ampt method needed intensity data at
daily/sub-daily timescale, which is not available in the study catchment. Therefore, based on the
availability of the input data, the CN method was used for this study. Moreover, the model uses
either Penman-Mointhes, Hargreves, or the Priestley and Taylor methods to determine the
Potential EvapoTranspiration (PET). The present study used the Pennman-monteith method to
estimate PET because of its robustness in estimating PET as compared to the other methods
(Herman et al., 2018). The detailed descriptions of the equation used for estimating PET and
surface runoff as well as other components of the water balance can be found in the SWAT
model theoretical documentation (Neitsch et al., 2011).
4.1.2. SWAT Input

A) Physiographic data
The Digital Elevation Model (DEM), land use, and soil are the three spatial input data required
for the SWAT model. These data were collected from different sources. The 30 m Shuttle Radar
Topography Mission (SRTM) DEM was used to create the stream network of the study
catchment. This data was collected from the NASA website. Following stream network creation,
a total of 12 sub watershed was demarcated (Figure 4.1). To represent the land cover of the
study catchment in the SWAT model, we used 2016 land cover map of the study catchment,
which was collected from the Ethiopian Space Science and Geospatial Institute (ESGI). A total
of 7 land cover classes are found in the study catchment based on ESGI land cover
classification scheme. However, the name of these classes is a bit different from the SWAT land
use database. Therefore, efforts were made to match the name of these land cover classes with
the SWAT land use database. Considering the description of SWAT land use classes. ESGI’s
land cover class (Figure 2.2) were aggregated into six major SWAT land-use types: Agricultural
Land-Generic (AGRL), Pasture (PAST), Forest-Mixed (FRST), Range-brush(RNGB), Water
(WATR), and Residential (URBN)
Regarding soil data, the Africa Soil information service (AFSIS) soil map with 250 m spatial
resolution (Leenaars et al., 2014) was used to represent the physical and chemical soil
properties of the study catchment in the SWAT model,. This data was collected from the
International Soil Reference and Information Centre (ISRIC) soil database. The AFSIS used a
Pedo-transfer function (Saxton and Rawls, 2006) to derive those soil properties for a soil layer
up to 6.
To simulate the hydrological processes of the study catchment, the SWAT model required daily
climatic data: rainfall, maximum and minimum temperature, humidity, wind speed, and solar
radiation. These data were collected from the Meteorological Institute of Ethiopia and Climate
Forcast System Reanalysis (CFSR) for the period 1990 to 2018. A total of 6 meteorological
stations are selected for the present analysis and their consistency was checked using a double
mass curve (Searcy and Hardison, 1960). The missing records of each gauging station were
filled using Multiple Imputation by Chained Equation (MICE; van Buuren and Groothuis-
Oudshoorn, 2011). Moreover, long-term daily streamflow data were collected from the
Ethiopian Ministry of Water and Energy for the same period. This data was used to calibrate the
SWAT model.
Figure 4. 1: SWAT model setup and sub watershed discretization
4.1.3. Model calibration schemes
In this study, the SWAT model was calibrated for each rainfall inputs independently at monthly
timescale using the Sequential Uncertainty Fitting version 2 (SUFI2) optimization algorithm
(Abbaspour et al., 2004). The SUFI2, one of the embedded algorithms in the SWAT calibration
and uncertainty program (SWAT-CUP), has been applied in different regions of the globe and
has provided a robust result as compared to the Generalized Likelihood Uncertainty Estimation
(GLUE), Parameter Solution (Para-Sol), Particle Swarm Optimization (PSO) (Nguyen and Thi,
2015; Shivhare et al., 2018). The Nash Sutcliff Efficiency (NSE) was used to optimize model
parameters since this metrics emphasis high flows. The SUFI2 algorithm uses Latin hypercube
sampling (LHS) to draw independent parameters set stochastically to calibrate model
parameters. The model was calibrated for the period 1990-1995.
In addition to model calibration and validation, a global sensitivity analysis was performed to
select the most sensitive parameters, considering the p and the t statistics. The p-value
determines the significance of the sensitivity while the t stat provides a measure of parameter
sensitivity. The p-value close to zero indicates the more significance of the parameter whereas
a larger absolute t value indicates the more sensitive parameters. This sensitivity analysis
allows to select the most sensitive parameters and facilitates to improve the efficacy of the
model.
4.1.4. Model performance Evaluation
The performance of the model was evaluated using four performance metrics: the Nash-Sutcliff
Efficiency (NSE), Kling Gupta Efficiency (KGE), Percent Bias (PBIAS), and Coefficient of
Determination (R2). The NSE was used to quantify how well a model simulation can predict the
observed flow. The range of NSE values between −∞ and 1.0 where 1 is the perfect fit between
the simulated and observed values. According to Moriasi et al. (2007), an NSE value greater
than 0.5 is considered as a good to the excellent performance of the model for predicting
streamflow. A negative value indicates the poor performance of the model. The KGE
decomposes the model errors into three different forms; the linear correlation, the bias ratio, and
the variability ratio. The optimal value of KGE is 1.0.
4.2. SWAT model results: Flow and Sediment
Port to modelling flow at different water abstraction points, sensitivity analysis was undertaken
to identify the most sensitive model parameters. Figure 4.2 shows the global sensitivity of model
parameters during the calibration period. .The four most sensitive model parameter in the study
catchment are Treshold depth of water in the shallow aquifer required for return flow to occur
(GWQMN), Soil evaporation compensation factor (ESCO), Threshold depth of water in the
shallow aquifer for "revap" to occur (REVAPMN) and curve number (CN). The optimized
parameter values are presented in Table 4.1.
Figure 4. 2: Global Sensitivity analysis
Table 4. 1: Optimized SWAT model Parameters for the study catchment

Parameters * Definition Range Optimized
Min max Parameter
Values
r__CN2.mgt SCS curve number for moisture -0.25 0.25 -0.003
condition II
v__ALPHA_BF.gw Base flow alpha factor (day) 0 1 0.068
v__GW_DELAY.gw Ground Water Delay (Days) 0 800 571
v__GWQMN.gw Threshold depth of water in shallow 0 5000 1732.04
aquifer for return flow to occur
v__GW_SPYLD.gw Groundwater Specific Yield 0 0.4 0.15
V_GWHT Initial groundwater height 0 25 1.07
v__REVAPMN.gw Threshold depth of water in shallow 0 500 332.62
aquifer for revap to occur
v__SHALLST.gw Initial depth of water in the shallow 0 50000 2801.65
aquifer (mm)
v__ESCO.hru Soil evaporation compensation factor 0 1 0.05
v__EPCO.hru Plant Uptake compensation factor 0 1 0.71
The performance of SWAT model in simulating monthly observed discharge was assessed
using various performance metrics as presented in Table 4.2. Results show that we found
reasonable model performance results during the calibration period as explained by the
highest NSE (0.76), KGE (0.79), PBIAS (2.2%) and R (0.72). As expected, the performance of
2
the model is relatively low at the validation period 1996-1999. During the whole period (1990-
2003), the performance of the model is reasonable in simulating monthly flow (Figure 4,4).As
exemplary hydrograph, simulated versus observed discharges at Guma gauging station is
shown in Figures 4.3 and 4.4, the model underestimate the peaks in all years of the calibration
period The main reason for the under estimation of the peak is mainly associated with the
structure of the model. On the other hand, SWAT predicted low flows comparted to peak flows
Table 4. 2: Model performance using various performance metrics during the calibration period 1990-
1995.
Figure 4. 3: Observed Vs Simulated discharge at monthly timescale for the period 1990-1995. The green
color shows the 95 PPU
Figure 4. 4: Observed vs simulated flow for the whole period 1990-2003
After calibrating and validating the SWAT model, efforts were made to model streamflow for the
present period (2004-2018) in order to bridge the data gaps observed in the gauging stations for
the same period .Table 4.3 shows the long-term mean monthly flow and 90 % dependability for
each water abstraction points for the period 1990-2018. The supply and demand for each
options were compared so as to select the best water abstraction point for planning and
operating the water supply scheme. Compared to the maximum demand, which is 1.87 MCM,
lower water deficit was observed in option 1 than the other options (Table 4.3). This option was
then selected to further analysis the required hydrological parameters that could be used as
input for designing the water supply scheme. Table 4.4 shows the monthly distribution of
sediment outflow estimated from the selected water abstraction point, weir. The estimated long-
term annual sediment yield of this option site using SWAT model is 86000 m 3 with the highest
recorded in the month August and lowest in December.
Table 4. 3 : Long-term mean monthly flow and 90 % dependability for each water abstraction points,
options. Option 1 was selected for further analysis.
Max. Demand
(0.7m3/s) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann
Demand(MCM) 1.87 1.69 1.87 1.81 1.87 1.81 1.87 1.87 1.81 1.87 1.81 1.87 22.08
Option 1
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann
Monthly (MCM) 5.75 4.79 5.73 7.53 17.46 27.4 31.67 35.77 31.79 26.72 9.19 5.24 209.10
7
90% DF (MCM) 1.49 1.26 1.49 1.98 3.88 6.19 14.01 13.21 13.04 7.18 3.20 2.42 69.35
Option 6
38.6
Monthly (MCM) 8.09 6.74 8.06 10.59 24.55 4 44.54 50.30 44.71 37.58 12.92 7.37 294.08
90% DF (MCM) 2.10 1.77 1.78 2.78 5.45 8.71 19.32 18.20 18.34 10.09 4.50 3.41 96.46
Option 2
13.8
Monthly (MCM) 2.54 2.53 2.98 3.61 7.95 9 18.08 20.43 18.25 14.33 4.52 2.74 111.84
90% DF (MCM) 0.52 0.32 0.43 0.69 1.34 2.11 4.61 6.59 4.76 2.34 1.06 0.81 25.57
Option 5
19.7
Monthly (MCM) 3.62 3.60 4.24 5.14 11.31 6 25.73 29.08 25.96 20.39 6.43 3.90 159.14
90% DF (MCM) 0.74 0.46 0.62 0.98 1.90 3.00 6.56 9.37 6.77 3.32 1.50 1.16 36.38
Option3
15.6
Monthly (MCM) 2.09 1.65 1.92 3.24 8.10 0 17.13 25.28 17.07 14.09 5.03 3.58 114.78
90% DF (MCM) 0.68 0.43 0.47 0.64 0.94 2.81 5.71 6.09 5.18 2.08 1.07 1.01 27.10
Option 4
Monthly (MCM) 0.61 0.48 0.56 0.95 2.36 4.55 4.99 7.37 4.97 4.11 1.47 1.04 33.46
90% DF (MCM) 0.20 0.12 0.14 0.19 0.27 0.82 1.66 1.78 1.51 0.61 0.31 0.29 7.90
Table 4. 4: The long –term mean monthly sediment estimating using SWAT model (m 3)
Annua
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec l
1354. 1729. 2490. 4549. 9645. 7864. 12104. 12576. 12922. 11475. 3628. 2299. 82640.
9 1 3 3 8 4 3 3 7 3 4 7 6
4.3. Estimating Dependable Flow
Dependable flows for option1 were quantified using flow duration analysis. To construct flow
duration curve at daily and monthly timescale, the following steps were followed. Steps include
(i) arranging the flow values (data points) in descending order of their magnitude, (ii) ranking the
data points, (iii) assigning plotting position (exceedance probability) to each data point using the
Weibull plotting position formula (Eq.4.2), and (iv) plotting data in a two-dimensional space of
flow magnitude versus exceedance probability and joining the resulting points to form a smooth
curve.
m
P ( X ≥ xm)= Eq 4.2
n+1
Where X is the random variable representing streamflow, xm is mth ranked data point, and n is
the total number of data points. Streamflow values corresponding to various exceedance
probabilities are referred to as flow quantiles.
Based on the flow duration analysis results, the 90% dependable flow based on the mean daily
flow was estimated to be 0.6m3/s (Figure 4.5 and Table 4.5). Meaning, this flow would be
available 90% of the year. Similarly, a flow duration curve for individual months is estimated
and results are presented in Table 4.5 and Figure 4.6.
Table 4. 5: 90 % dependable flow (MCM) estimated at monthly timescale
Flow Jan Feb Mar Apr Ma Jun Jul Aug Sep Oct No Dec Ann.
y v
90%
Dependable 1.4 1.2 1.9 3.8 6.1 14.0 13.2 13.0 7.1 3.2 2.4 69.3
flow (MCM) 9 6 1.49 8 8 9 1 1 4 8 0 2 5
80
70
60
Daily Flow (m3/s)
50
40
30
20
10
0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 96 99
Exceedance Probability
Figure 4. 5: Flow Duration curve for the selected option site, weir, at daily timescale
80
70
60
Jan Feb
50 Mar Apr
May Jun
Monthly flow (m3/s)
40
Jul Aug
Sep Oct
30
Nov Dec
20
10
0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 96 99
Exceedence Probablity
Figure 4. 6: Flow Duration curve for the selected option site, weir, at monthly timescale
5. FLOOD and LOW-FLOW ESTIMATION
Estimating flood quantiles for different return periods are required for the design of high-risk
structures such as dams, bridges, and levees. Methods for estimating these floods can be
generally classified into three groups: (1) statistical flood frequency analysis, (2) the single
design event simulation approach, and (3) continuous simulation approach. The present study
used the statistical flood frequency analysis to quantify high quantile flood for different return
periods.
5.1. Flood estimation using statistical flood frequency analysis

Since gauging station located at downstream of the weir site, the flood frequency analysis was
employed at the gauging station and then transferred to the weir site by considering the
catchment the gauging station and weir site.
5.1.1. Selection of the best-fit regional frequency distribution
The choice of frequency distributions is determined based on goodness-of-fit measures, which

indicate how much the considered distributions fit the available data. A shape parameter
describes the weight of the distribution tail of the random variable. Therefore, four different types
of three-parameter (location, scale, and shape) distributions, which have been commonly used
for flood frequency analysis, were used in this study. These distributions include the
Generalized Extreme Value (GEV), Generalized Pareto (GPa), General logistic (GLO) and
Pearson Type III (Pe3) were tested for the flood Frequency analysis. Hosking and Wallis, (1997)
suggested that the best choice of distribution is the one that is robust or capable of giving good
quantile estimates even though future values may come from a distribution somewhat different
from the fitted distribution when several distributions fit the data adequately. To select the best
fit distribution, the empirical and theoretical CDF plot of each distribution were compared using
RMSE.
5.1.2. Prediction of flood at the weir site (Ungauged site)
To predict flood at the weir site (ungauged site), the index flood method was used for estimating
flood quantiles corresponding to T-year return periods. The key assumption of an index flood
procedure is that the sites forming a homogeneous region have identical frequency distribution
called the regional growth curve (RGC) but a site-specific scaling factor could be needed. i.e the
index flood. The index flood can be determined as
Qe ( T ) =QI × q ( T ) Eq .5 .1
where Qe(T) is quantile estimate, QI is the so-called index flood and q(T) is the regional growth
curve which is a scaled quantile function assumed to be common to every site in a
homogeneous region, i = 1,2,. . .,N denotes the sites and N denotes the total numbers of sites.
In this study, the regression relationships between catchment area and index flood (Smith et al.,
2015) were used to derive the index flood at the ungauged site (weir site), which is 0.71.
5.1.3. Food Frequency Analysis Results
Among the set of distribution, we selected Generalized Extreme Value to model peak flows at
the gauging station, Dincha (Table 5.1 and 5.2) since it produced the lowest RMSE value (Table
5.1). The flood quantiles for different return periods was then estimated using this distribution
and results are presented in Table 5.3. The high quantile floods at the weir site for 25, 50 and
100 years return period are 66.7, 78. 2 and 91 cms, respectively (Table 5.3). High-quantile
flood estimated from Dincha gauging station for different return periods with uncertainty bounds
and CDF are presented in Figure 5.1 and 5.2.
Table 5. 1: Set of candidate distributions for modelling floods at Dincha Gauging station
Set of candidate distributions abbreviation RMSE
Generalized Extreme Value gev 0.048
Generalized Normal gno 0.051
Exponential exp 0.055
Pearson Type III pe3 0.056
Weibull wei 0.057
Generalized Pareto gpa 0.059
Gumbel gum 0.065
Gamma gam 0.076
Rayleigh ray 0.079
Rice rice 0.099
Normal nor 0.103
Reverse Gumbel revgum 0.148
Table 5. 2: Annual Maximum series flow at Dincha Gauging station

Year flow
1990 36.8
1991 43.7
1992 57.8
1993 43.9
1994 50.2
1995 54.5
1996 52.9
1997 104.1
1998 74.3
1999 35.2
2000 33.5
2001 53.6
2002 25.6
2003 40.9
Table 5. 3: High-quantile flood for different return period at gauging station and weir site
Flood Quantile at Dincha Gauging Station (FQDGS) Flood quantiles at the weir site (0.71* FQDGS)
Distribution RP.2 RP.5 RP.10 RP.25 RP.50 RP.100 RP.2 RP.5 RP.10 RP.25 RP.50 RP.100
gev 45.5 62.2 75.1 94.0 110.1 128.1 32.3 44.1 53.3 66.7 78.2 91.0
gno 45.4 62.8 76.0 94.4 109.4 125.3 32.2 44.6 53.9 67.0 77.6 89.0
exp 44.2 63.2 77.7 96.8 111.2 125.6 31.4 44.9 55.1 68.7 78.9 89.2
pe3 45.1 63.8 77.3 94.5 107.4 120.2 32.0 45.3 54.8 67.1 76.3 85.3
wei 45.1 64.1 77.5 94.5 106.9 119.1 32.0 45.5 55.0 67.1 75.9 84.6
gpa 44.9 64.6 78.2 94.6 105.9 116.4 31.9 45.9 55.5 67.2 75.2 82.6
gum 47.4 64.4 75.7 89.9 100.5 111.0 33.6 45.7 53.7 63.8 71.3 78.8
gam 48.2 65.3 75.6 87.7 96.1 104.1 34.3 46.4 53.7 62.3 68.2 73.9
ray 48.4 65.9 75.9 87.0 94.3 101.1 34.4 46.8 53.9 61.7 67.0 71.8
rice 50.2 66.1 74.5 83.5 89.3 94.5 35.7 46.9 52.9 59.3 63.4 67.1
nor 50.5 66.1 74.2 82.9 88.4 93.5 35.9 46.9 52.7 58.8 62.8 66.4
revgum 53.7 66.4 71.7 76.8 79.7 82.2 38.1 47.1 50.9 54.5 56.6 58.3
Figure 5. 1: High-quantile flood estimated from Dincha gauging station for different return periods with
uncertainty bounds at 95 % confidence level (purple color)
Figure 5. 2: CDF for Dincha gauging station with uncertainty bounds at 95 % confidence level (purple
color)
5.2. Low-flow and baseflow analysis
In this section, we present low-flow and base flow analysis for the selected weir site. This
analysis is useful for estimating the probability of water availability in streams during critical low
flow periods such as drought. Low flow statistics can be used in water supply planning to
determine allowable water transfers and withdrawals. The 1- and 7-day low flow with 10 years of
return period, is one of the most-used (design or reference) low flow indices. It is used for many
design practices such as the protection or regulation of water quality from waste water
discharges or waste load allocations. In this project, the low flow frequency analysis was
performed by modelling annual 1-day and 7-day minima flows at the weir site. Just like the
rainfall and flood frequency analysis, we selected the best fit distribution by comparing the
empirical and theoretical cumulative distribution function (CDF) for the set of distributions. We
found that the Weibul distribution is the best fit distribution to model low flows and its parameter
was determined using L moment. Table 5.4 and figure 5.3 show low the 1-day and 7-day flow
quantiles for different return periods. Results revealed that the 1-day and 7-day low flow quantile
at 10 years return period is 0.27 and 0.54 cms, respectively (Table 5.4). Moreover, the 1-day
low flow quantile results (Figure 5.3) revealed that the low flow value of 0 is attained in the long
term future (> 10000 years) and thus the river will not be dried-up.
Table 5. 4: Low flow analysis

Return period 1-day low flow quantile 7-day low flow quantile
2 0.60 3.15
5 0.37 1.03
10 0.27 0.54
25 0.18 0.29
50 0.13 0.22
100 0.10 0.18
Figure 5. 3: 1-day Low-flow frequency analysis for different return levels
The BaseFlow Index, or BFI, which measures the ratio of long-term baseflow to total stream
flow, was also determined. This analysis is very important for river basin ecology and water
resources planning and management. Table 5.5 shows the yearly baseflow index at the weir
site and the average BFI is estimated to be 0.58, indicating that a large fraction of streams
obtain the majority of their streamflow from baseflow, and demonstrating the importance of the
baselow component of streamflow. This hold true as most parts of the catchment are covered
with Forest. The hydrograph of the streamflow and baseflow of the study catchment is
presented in Figure 5.4.
Table 5. 5: Baseflow index (BFI) at the selected option site, weir
19 19 19 19 19 19 19 19 19 19 20 20 20 20
90 91 92 93 94 95 96 97 98 99 00 01 02 03
0.5 0.4 0.5 0.5 0.7 0.6 0.7 0.5 0.5 0.5 0.5 0.5 0.6 0.6
3 9 2 8 0 8 0 1 8 0 8 2 5 3
Figure 5. 4: Streamflow and baseflow hydrograph for the period 1990-2003

6. DESIGN FLOOD ESTIMATION FOR CROSS-DRAINAGE
In this section, the methodology used for estimating design flood for cross-drainage are
presented. Following this, the peak discharge at each crossing point was estimated based on
ERA 2013 drainage design manual.
6.1. Design Flood Estimation for Cross-Drainage Structures
Depending on the size of the catchment area of the cross drainages, the rational and the United
States Soil Conservation Service (SCS) method was applied. The rational method was
employed when the catchment area is less than 0.5 km 2 whereas the SCS method was used for
a catchment area greater than 0.5 km2. In this study we used SCS method since the
catchment’s of the CD are greater than 0.5. We also considered, a 25-year flood, which was
commonly used to design most of the drainage facilities. The location of cross-drainage (CD)
points and their associated catchment area are presented in Table 6.1. The spatial distribution
of these CD points in the main canal is shown in Figure 6.1.
Figure 6. 1: Location of cross drainage points and curve number (CN) of the command area
Table 6. 1: Location of cross-drainage points and their corresponding catchment area. The projection is
WGS-UTM-37N.
Location
CD
Name Easting Northing Area ( Km2)
CD-1 195421.7838 805402.764 41.0
CD-2 195792.3441 807028.2401 122.9
CD-3 197863.2606 800232.5814 7.0
CD-4 202118.2264 798685.3211 12.4
6.1.1. SCS Method

6.1.1.1. Input data
For a catchment area greater than 0.5 km 2, the following data are required for estimating floods
at each crossing point using the SCS method;
 Curve Number of the catchment area of CD

 Physical catchment characteristics including Time of Concentration (TC)
 Extreme rainfall quantiles for 25 and 50 years return period
Determination of the CN value of the specific catchment is a prerequisite for estimating design
floods when using the SCS method. For this analysis, the CN of the catchment area was
determined based on land use and soil hydrologic group. The land use in the main canal of the
Bonga command area was identified based on the 2016 land use/cover map. The CN value for
each land cover was then assigned based on the ERA 2013 drainage manual guideline once
their corresponding hydrologic soil group was identified.
In addition to the CN value, the SCS method required data pertinent to the physical catchment
characteristics of the study area. These data include rainfall depth, drainage area, mean slope
of the main channel, and channel length. The mean slope of the main channel and channel
length were used to estimate the time of concentration (TC, equation 7.1). The Kirpich equation
(1940), one of the most used empirical equations, was used to calculate the time of
concentration in a channel flow:
[ ]
0 .77
L
0 . 0195
S 0. 5
Tc 3=
60
6.1
Where:
Tc = time of concentration (hr)
L = the length of the catchment’s along the longest river channel (m)
S = the overall catchment’s slope in m/m = H/L
H = the difference in elevation between the most remote point and the outlet point
In light of this, extreme rainfall quantiles for different return periods were estimated using rainfall
frequency analysis. For this analysis, high rainfall quantiles were estimated for 2-yr,5-yr, 10-yr
25-yr,50-yr, and 100-yr return periods based on the best-fit distribution. A detailed description of
the best-fit distribution,PDF, and CDF for Bonga station is presented in section 3.3.2
Table 6. 2: Estimated high quantile extreme rainfall for different return periods at Bonga station,
which was used as input for estimating peak discharges
2-yr 5-yr 10-yr 25-yr 50-yr 100-yr
46.37 55.99 62.93 71.68 78.00 84.07
The aforementioned data were used as input the SCS method to compute direct runoff from 24-
hour or 1-day storm rainfall. The SCS method uses the following equations;
Q= (P- Ia)2 / (P - Ia) + S (6.2)

Where:
Q= accumulated direct runoff, mm
P= accumulated rainfall (potential maximum runoff), mm
Ia= initial abstraction including surface storage, interception, and
infiltration prior to runoff, mm
S= potential maximum retention, mm
The relationship between Ia and S was developed from experimental catchment area data. It
removes the necessity for estimating Ia for common usage. The empirical relationship used in
the SCS runoff equation is:
Ia = 0.2S (6.3)
Substituting 0.2S for Ia in equation above, the SCS rainfall-runoff equation becomes:
Q = (P - 0.2S)2 / (P + 0.8S) (6.4)
S is related to the soil and cover conditions of the catchment area through the CN. CN has a
range of 0 to 100, and S is related to CN by:
S = 1000/CN – 10 (6.5)
The peak discharge is determined by using 24 hrs return period rainfall (P) and the direct runoff
(Q). The unit peak discharge (Qu) is determined using Tc and Ia/P parameters. The peak
discharge is then calculated using the formula:
QPT = qu *A* QT (6.6)
Where:
QPT = Peak Discharge (T-years return period) m3/s
qu = Unit peak discharge m3/s/100ha/mm
QT = Direct Runoff (T-years return period), mm
q u=0.000431∗10¿¿
Where co=-2.23*(Ia/p)^2+0.47*(Ia/p)+2.52, c1=1.5*(Ia/P)^2-0.7*(Ia/p)-0.5),

c2=0.6*(Ia/p)^2+0.043*(Ia/p)-0.176
6.1.1.2. Developing Intensity-Duration-Frequency Curve for the Command Area
The design rainfall-intensity-duration relationships are obtained directly from the time distribution
of rainfall, simply by converting the rainfall for a given duration to rainfall intensity (mm/hr).
Rainfall intensity in mm per hour could be derived for various return periods of durations ranging
from 5 minutes to 120 minutes. The value of rainfall intensity is computed by considering the
proportion of the rainfall at the time of concentration using the 24-hr rainfall. Based on this
definition, it can be calculated using the equation:
R tc
i=
T conc
6.8
Where: i = rainfall intensity (mm/hr),
Rtc= the amount of rain falling during the time of concentration (m),
Tconc = the time of concentration for the sub-basin (hr).
An analysis of rainfall data by Hershfield (1961) method for different durations and frequencies
showed that the amount of rain falling during the time of concentration was proportional to the
amount of rain falling during the 24-hr period.
6.9
Where,
Rtc = the amount of rain falling during the time of concentration,
αtc = the fraction of daily rainfall that occurs during the time of concentration,
Rday= is the amount of rain falling during 24-hr.
The Transport and Road Research Laboratory (TRRL, UK), Laboratory Report 623, published
"Prediction of storm rainfall in East Africa" (D. Fiddes, J.A. Forsegate and A.O. Grigg). The
research study gave a valuable result of regional storm rainfalls in East Africa. The research
study used the following model to fit a large number of stations data:
a
I=
( b+ t )n
6.10
Where: I = intensity or rainfall in mm/hr
t = duration in hours
a, b & n are constant
The best fit was found for b equals 0.33 hours and the index n varies from 0.78 to 1.09 and
adopts 24-hour rainfall value as I24.
I=[(b+24)n/(b+24)n]*I24 6.11
Therefore, in this case, we adopt the value of “b” as 0.33 and “n” equals 0.9 (equals to the
average value of n). Based on this formula, the rainfall intensities of the stations are computed
and the IDF curve also developed as shown in Figure 6.2
6.2. Flood estimation for the cross drainage structures
Table 6.3 shows flood quantiles estimated for the cross drainage structures. Results revealed
that the design discharge for 25 years return period ranges from 5.4 to 44.8 cms.
Table 6. 3: Flood estimation for the cross drainage structures located at different points.
Catchment Design Unit Peak Design Discharge (m3/s)
Stream Characteristics discharge
Length TC
Name (m) slope CN (hr) 25yr 50yr 100yr 25yr 50yr 100yr
CD1 26933.26 0.037 68.500 2.986 0.04 0.04 0.04 44.76 70.05 113.02
CD2 20409.41 0.042 66.800 2.303 0.04 0.05 0.06 15.47 25.29 42.44
CD3 2537.87 0.190 65.34 0.315 0.25 0.29 0.33 13.29 22.78 39.29
CD4 6432.2 0.105 61.74 0.797 0.08 0.11 0.13 5.40 10.65 20.99
IDF Curve of Bonga Station

160.00
140.00
120.00
100.00
Intensity in mm/hr
2 years
5 years
80.00 10 years
25 years
50 years
60.00 100 years
40.00
20.00
0.00
0 20 40 60 80 100 120 140 160 180 200
Duration in Min
Figure 6. 2: IDF curve for Bonga Station

7. CONCLUSION AND RECOMMENDATION
This climatic and hydrological study aims to provide reliable hydro-meteorological data of the
study catchment that could be used as input for planning and designing water supply scheme,
and flood protection works. We used the state-of-the- art hydrologic model and rainfall
frequency analysis to determine flow at the selected water abstraction point and to quantify
floods for different return periods.
Based on the findings of this study, the following conclusion and recommendation are drawn;
 The mean annual rainfall of the study catchment is estimated to be 2217 mm.
 The long-term mean annual flow at the weir site is 209 MCM and this result is
comparable with the findings of the Omo-Gibe river basin master plan study.
 The 90 % long-term mean annual dependable flow at the weir site is 69.4 MCM with the
highest dependable flow was observed in August and lowest in February. It is thus
important to deal with this water scarcity issues in those months by using the ground
water as a source
 Compared to the maximum demand (around 1.8 MCM), water deficit is noted in the
month January, February and March but not for other months.
 The estimated flood quantiles at 25, 50 and 10 years return period are 67, 78 and 91
cms, respectively.
 The 1-day and 7-day low flow quantiles at 10 years return period is 0.27 and 0.54 cms,
respectively
 The baseflow index is estimated to be 0.58, indicating that a large fraction of streams
obtain the majority of their streamflow from baseflow, and demonstrating the importance
of the baselow component of streamflow.
 The design discharge for the cross drainage structures for 25 years return period ranges
from 5.4 to 44.8 cms.
 The long-term mean annual sediment yield at the weir site is estimated to be 86,000 m3
 Overall, the present study provides flow and sediment rates required for designing water
supply scheme and its appurtenant structures. However, these requirements may
change as additional information about local meteorology becomes available, thus
changing the properties of the likely extreme events. Further, as the watershed changes
due to development or natural shifts, the volume of runoff and the extreme event may
change, which may change the results of this study. Therefore, it is very important to
collect 5-years’ time-series flow and sediment data for the specific project site before
constructing the weir. Moreover, various watershed management intervention measures
shall be implemented at the catchment scale to reduce sediment inflow into the
reservoir.
8. REFERENCE
Chow V.Te. A general formula for hydrologic frequency analysis. Eos, Trans. Am. Geophys.
Union. 1951;32:231–237. doi: 10.1029/TR032i002p00231. [CrossRef] [Google Scholar]
ERA (2013). Drainage Design Manual, Addis Ababa, Ethiopia.
FAO. (1998). Crop evapotranspiration-Guidelines for computing crop water requirements-
Irrigation and drainage paper 56. Rome, Italy.
Halcrow Group Ltd and GRID (2009). Rift valley lakes integrated Resources development
Master Plan study Project.
Hershfield D.M. Estimating the probable maximum precipitation. Journal of Hydraulics Division.
1961;87:99–106. [Google Scholar]
Hosking, J.R.M., 1990. L-Moments: Analysis and Estimation of Distributions Using Linear
Combinations of Order Statistics. J. Roy. Stat. Soc.: Series B (Methodol.) 52, 105–124.
https://doi.org/10.1111/j.2517-6161.1990.tb01775.x.
International Water Management Institute:Water Resources and Irrigation Development in
Ethiopia, Working Paper 123, by Seleshi Bekele Awulachew, Aster Denekew Yilma, Makonnen
Loulseged, Willibald Loiskandl, Mekonnen Ayana and Tena Alamirew, 2007
Salas, J.D. and Fernandez, B. (1993). Models for data generation in hydrology: univariate
techniques. In Stochastic Hydrology and its Use in Water Resources Systems Simulation and
Optimization (pp. 47-73). Springer Netherlands.
Seleshi, Y., Zanke, U., 2004. Recent changes in rainfall and rainy days in Ethiopia. Int. J.
Climatol. 24, 973–983. https://doi.org/10.1002/joc.1052
Smith, A., Sampson, C., Bates, P., 2015. Regional flood frequency analysis at the global scale.
Water Resour. Res. 51, 539–553,http://dx.doi.org/10.1002/2014WR015814
9. ANNEX
Annex 1: Stationarity test results in time series data of Gonga kemisee Station. NS means not
significant
Critical values
Test (Statistical table) Result
Statistical test statistic a=0.1 a=0.05 a=0.01
Mann-Kendall 1.444 1.645 1.96 2.576 NS
Spearman's Rho 1.387 1.645 1.96 2.576 NS
Cusum 6 6.57 7.324 8.778 NS
Rank Sum -2.597 1.645 1.96 2.576 S (0.01)
Student's t -2.372 1.701 2.048 2.763 S (0.05)
Median Crossing 2.268 1.645 1.96 2.576 S (0.05)
Turning Point -1.365 1.645 1.96 2.576 NS
Rank Difference -3.448 1.645 1.96 2.576 S (0.01)
Annex 2: Stationarity test results in time series data of Kemisse Station. NS means not
significant
Test Critical values(Statistical table) Result

Statistical test statistic a=0.1 a=0.05 a=0.01
Mann-Kendall 2.42 1.645 1.96 2.576 S (0.05)
Spearman's Rho 2.409 1.645 1.96 2.576 S (0.05)
Linear regression 2.821 1.703 2.052 2.771 S (0.01)
Cusum 7 6.57 7.324 8.778 S (0.1)
Rank Sum -2.204 1.645 1.96 2.576 S (0.05)
Student's t -2.2 1.701 2.048 2.763 S (0.05)
Median Crossing 2.268 1.645 1.96 2.576 S (0.05)
Turning Point 0.455 1.645 1.96 2.576 NS
Rank Difference -2.522 1.645 1.96 2.576 S (0.05)
Annex 3: Stationarity test results in time series data of Keresi Station. NS means not significant
Statistical test Test Critical values(Statistical table) Result

statistic a=0.1 a=0.05 a=0.01
Mann-Kendall -1.182 1.645 1.96 2.576 NS
Spearman's Rho -1.137 1.645 1.96 2.576 NS
Linear regression -1.336 1.703 2.052 2.771 NS
Cusum 3 6.57 7.324 8.778 NS
Rank Sum -0.284 1.645 1.96 2.576 NS
Student's t -0.014 1.701 2.048 2.763 NS
Turning Point 0 1.645 1.96 2.576 NS
Rank Difference 0.319 1.645 1.96 2.576 NS
Annex 4: Stationarity test results in time series data Wushiwesh Station. NS means not
significant
Test Critical values(Statistical table)

Statistical test statistic a=0.1 a=0.05 a=0.01 Result
Mann-Kendall -0.732 1.645 1.96 2.576 NS
Spearman's Rho -0.748 1.645 1.96 2.576 NS
Linear regression -1.503 1.703 2.052 2.771 NS
Cusum 3 6.57 7.324 8.778 NS
Cumulative
deviation 1.149 1.118 1.238 1.456 S (0.1)
Rank Sum -0.502 1.645 1.96 2.576 NS
Student's t 0.252 1.701 2.048 2.763 NS
Median Crossing 0 1.645 1.96 2.576 NS
Turning Point 0.455 1.645 1.96 2.576 NS
Rank Difference -0.862 1.645 1.96 2.576 NS
Annex 6: Imputed monthly climatic data
A) Monthly rainfall for different rainfall stations for the period 1990-2018.
A1: Shadatura station
Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual
240. 337.
1990 36.9 165.4 198.5 183.2 4 210.0 235.6 2 277.7 74.4 136.5 65.8 2161.5
180. 269.
1991 78.5 175.2 193.0 199.7 6 257.4 256.2 2 233.8 75.9 46.1 64.1 2029.8
120. 278. 289.
1992 3 143.5 100.3 205.4 1 289.6 178.3 8 233.5 256.2 92.1 79.5 2266.6
118. 335. 149.
1993 6 96.6 121.1 238.6 0 209.4 245.4 0 246.8 138.8 52.3 13.7 1965.4
425. 283.
1994 8.5 7.3 158.2 222.8 1 260.7 309.2 7 131.6 109.4 46.4 36.2 1999.1
112.
1995 0.0 36.2 32.8 123.3 37.9 43.4 97.9 8 101.8 70.1 53.2 44.9 754.4
105. 132.
1996 85.4 10.9 72.5 68.2 7 170.2 39.5 1 90.8 65.4 6.0 9.2 855.7
187. 234.
1997 56.8 26.9 48.7 40.2 9 167.4 189.7 4 156.6 377.5 343.9 48.5 1878.5
125. 174. 412.
1998 0 90.9 136.0 234.6 6 172.7 500.7 8 208.5 515.3 54.4 0.0 2625.5
217. 268.
1999 0.0 16.2 120.5 297.7 3 183.8 285.1 3 228.1 596.1 19.7 80.5 2313.4
338. 121.
2000 0.0 7.0 87.4 307.1 0 270.9 148.9 2 324.6 313.8 64.3 43.9 2027.1
336. 448.
2001 49.3 10.3 176.8 142.4 4 237.9 194.6 6 277.9 183.1 86.0 50.0 2193.3
275. 244.
2002 85.1 45.7 145.3 0.0 35.2 244.5 218.1 7 183.7 166.2 25.9 7 1670.0
247. 403.
2003 70.1 21.8 10.2 202.7 4 355.9 398.3 7 324.6 80.3 41.1 3.8 2159.9
220. 100. 126.
2004 92.1 55.3 88.5 136.0 5 203.1 166.2 5 132.3 189.5 21.7 9 1532.6
136. 348.
2005 5 20.3 184.2 134.5 6 213.1 84.1 97.5 253.3 226.6 82.0 0.0 1780.7
254. 216. 199.
2006 30.4 79.6 191.8 157.6 3 204.7 460.2 8 364.5 189.0 148.0 6 2496.5
106. 272. 421.
2007 3 111.6 105.2 376.0 8 389.6 357.0 0 570.4 173.4 25.6 0.0 2908.8
455. 153.
2008 94.6 29.2 179.9 341.8 6 305.8 560.6 2 666.4 296.2 143.6 90.9 3317.8
111. 272. 204. 127.
2009 6 154.7 181.7 156.3 8 320.4 183.6 2 203.6 219.1 61.7 8 2197.6
515. 688.
2010 96.5 48.3 267.8 377.4 8 454.3 601.4 7 351.8 81.8 40.1 51.0 3574.9
208. 301.
2011 83.8 27.9 164.1 235.8 2 134.2 310.5 1 168.7 113.2 94.9 58.7 1901.1
146. 231.
2012 39.0 0.0 66.2 197.5 5 251.3 233.4 3 248.4 104.8 109.1 74.0 1701.4
372. 154.
2013 28.6 88.5 63.5 166.2 9 143.2 165.5 7 221.2 169.8 140.5 77.7 1792.1
203. 113.
2014 0.0 64.6 43.6 93.2 7 96.3 70.7 3 220.6 158.1 100.4 42.0 1206.6
213. 175. 111. 205.
2015 2 147.4 112.3 111.8 5 194.5 121.9 3 159.4 222.3 206.6 5 1981.7
152. 159.
2016 11.5 14.4 98.5 145.2 3 205.4 279.5 9 205.6 206.9 40.2 4.1 1523.5
2017 0.0 38.4 46.0 58.3 70.5 28.1 60.8 59.2 72.5 55.7 14.3 0.0 503.8
185. 207. 164. 172.
2018 7 161.4 104.7 124.8 9 95.5 146.4 3 192.7 183.5 131.7 0 1870.6
242. 238.
mean 71.2 65.4 120.7 182.0 0 217.7 244.8 5 243.2 193.5 83.7 69.5 1972.1
114. 136.
std 55.5 55.4 60.3 93.3 9 93.6 144.0 3 125.6 126.3 68.9 65.2 1139.3
CV 0.8 0.8 0.5 0.5 0.5 0.4 0.6 0.6 0.5 0.7 0.8 0.9 7.6
A2: Wushiwesh Station
1990 17.2 112 328.6 221.4 409.4 348.9 221.7 294.1 252.6 69.2 99 35.6 2409.7
1991 70.3 116 151.5 237.5 259.2 282.4 164.4 295.8 196.4 98.5 69.7 20 1961.7
1992 70.1 163 195.9 153.5 237.9 271.8 214.3 228.8 201.3 295.4 130 83.5 2245.5
1993 46.5 95.5 113.9 164.7 205.1 250.2 174.6 125.6 225.9 155 26.8 11.3 1595.1
1994 10.7 33.5 67.7 155 383.8 146.9 297.9 187.7 204.2 85.9 132 50.6 1755.9
1995 1.2 56.5 75.6 161.7 202.8 169.8 153.2 211 227 68.2 50.8 161.8 1539.6
1996 47.8 42.8 185.8 273.3 166 64.3 213.6 139.3 186.5 69.4 74.3 32.6 1495.7
1997 74.6 7.1 139.3 276.6 222.3 248.2 195.3 215.6 167 322.1 150.4 188.6 2207.1
1998 106.7 26.6 127.8 233.1 217.5 276.8 174.4 358.7 184.3 247 16.7 0.5 1970.1
1999 14.5 16.3 108.9 255.1 187.3 120 203.5 150.3 196.5 124.7 20.3 28.9 1426.3
2000 2.7 7.4 66.3 166.4 235.8 149.7 253.1 155.9 110.1 171.9 16.1 29.7 1365.1
2001 16.3 43.5 116.5 130.7 273.7 251.2 223.3 145.6 195.8 164.6 72 28.6 1661.8
2002 36.7 5.4 157 154.9 150.9 174.7 195.9 161.5 141.7 117.9 34.8 127.6 1459
2003 54.6 15.4 137.9 221.7 109 270.9 211.6 243.5 227.1 56.3 65.2 37.4 1650.6
2004 149.8 20.2 93.4 228.4 269.5 199.9 193.6 264.8 183.7 84.4 102.2 71.7 1861.6
2005 25.4 32.8 183.7 210.5 232.6 153 168.4 240.6 295.9 133.1 91 2.3 1769.3
2006 47.7 111.4 208.9 120.2 217.1 211.1 187.1 267.8 168.8 165.6 148.5 102.7 1956.9
2007 100 37.1 115.4 192.9 221.6 220.9 219.5 213.5 210.2 111 114.9 0.5 1757.5
2008 72.5 59 59.6 211.3 223.7 149.1 196.7 174.8 229.6 82.7 77.1 98.4 1634.5
2009 46.1 52.6 154.1 172.4 105.8 289.7 120.6 178.1 244.9 210.8 75.5 89.3 1739.9
2010 40.7 109.1 106.7 205.5 230.1 207.3 126.5 195.6 284.6 140.3 84.9 79.5 1810.8
2011 9.6 7.8 108.9 158.6 241.8 206.9 182.4 248 289.8 65.7 168.9 35.3 1723.7
2012 20.3 2.7 43.4 166.1 87.8 276.1 181.4 204.4 322 73.1 150.8 71.5 1599.6
2013 77.8 31.1 116.5 126.9 209.4 114.7 220.4 202.4 256 129.4 185.1 5.8 1675.5
2014 22.9 56.4 195.9 239.4 299.6 162.3 181.3 238.5 78.6 275 148.6 4.9 1903.4
2015 0.7 49.6 169.5 245.2 258.8 167 127.9 102.7 147.9 189.9 166 80.5 1705.7
2016 40.2 30.1 72.2 197 286 153 313.2 171.3 246.7 213.3 35.1 45.5 1803.6
2017 3.2 95.3 37.2 197 225.4 157.9 139.4 112.3 180.7 317.7 131.9 49.1 1647.1
2018 65.5 63.9 187.2 78.5 110.9 170.1 192 120.3 133.8 198.3 102.4 170.6 1593.5
mean 44.6 51.7 131.9 191.6 223.5 202.2 194.7 201.7 206.5 153.0 94.5 60.1 146.3
std 35.4 40.5 60.4 47.5 71.2 63.8 43.3 60.0 55.0 78.2 49.2 51.1 54.6
CV 0.8 0.8 0.5 0.2 0.3 0.3 0.2 0.3 0.3 0.5 0.5 0.8 0.5
A.3. Bonga
1990 60.6 134.6 138.3 143 254.4 173.4 189.6 233.5 216 74.4 85.5 65.2 1768.5
73.2 50.2 126.7 226.1 202.3 220 194.2 264.1 201. 53.4 52 51.8
1991 5 1715.5
88.3 69 102.2 103.6 155.6 257.8 191.5 194.9 159. 285 99.9 59.1
1992 3 1766.2
138.7 89.8 99.3 250 276.8 244.9 178 121.2 202. 183.1 11.4 2.7
1993 8 1798.7
10.2 9.1 71.6 0 245.7 122.4 236 139.9 126. 46.2 111.6 19.1
1994 3 1138.1
0 42 52.3 158.9 153.4 160.8 172.6 214.4 223. 56.8 31.6 151.8
1995 3 1417.9
1996 45.5 36.6 155.4 202 188.9 159.6 158 178.3 214 93.6 94.8 21.7 1548.4
86.2 12 133.5 231.7 205.9 212.7 181.8 146 138. 239.9 248.2 135.9
1997 4 1972.2
128.2 47.6 64.7 173.4 223.1 217 207.5 260.4 192. 154.2 14.5 0
1998 3 1682.9
1999 33.8 6 108.1 174.6 181.9 138.2 165.1 121.9 138 162.7 23 22.2 1275.5
6.3 4.6 101.4 194.3 214.1 161.6 232.1 135.1 147. 260.4 38.7 28.9
2000 5 1525
11.1 67.5 119.4 202.9 222.6 192.5 178.3 197.3 195. 118.1 69.7 6.4
2001 7 1581.5
36.1 22 172.5 131.1 102.2 253.2 142.1 159 166. 158.8 33.1 115.7
2002 4 1492.2
47.2 23.3 51.5 213.3 47.7 352.1 462.9 433.4 314. 26.8 46.4 48.7
2003 3 2067.6
95.4 16.1 95 225.6 255.2 82.8 142.2 281.3 275. 85.6 63.7 109.2
2004 5 1727.6
33.7 39.5 158 163.5 319.4 202.1 172.5 178.9 184. 140 89.7 0
2005 7 1682
31.2 68.5 155.8 88.6 236 185.7 287.9 206.4 183. 148.1 129.4 108.7
2006 7 1830
116.9 40.2 96.8 198.9 295.1 276.5 116.8 216 188. 87.4 56.7 0
2007 5 1689.8
31.5 55.4 59.4 188.8 214 202.8 240.7 278.6 150. 245.1 77.8 33.4
2008 1 1777.6
48.1 29 146.6 174.3 125.2 198.5 146 153.4 253. 281.8 104.5 97.2
2009 8 1758.4
54.7 79.2 157.2 199.4 282.1 227.4 233.9 342.1 421. 204.7 33.7 45.4
2010 5 2281.3
14.4 10.7 129.5 311.1 310.2 192.5 377.5 380.3 367. 74.4 102.5 12
2011 8 2282.9
26.1 6.9 90.2 137 183.8 258.9 232 227 277. 63.7 101.2 59
2012 5 1663.3
34.8 28.2 113.7 117.6 240.3 267.3 258.7 217.3 218. 139.6 177.5 35.8
2013 4 1849.2
2014 22.6 41.4 210.3 164.1 257.8 208.9 200.3 189.8 139 209 129.7 4.6 1777.5
32.5 49.7 144 287.8 262.6 231.1 234.9 143.5 127. 181 145.9 65.4
2015 3 1905.7
83.1 86.1 79.1 243.3 304 99.8 344.8 219.9 127. 155.2 66.8 41
2016 8 1850.9
6.1 95.3 118.8 250.9 356.8 247.8 214 199.4 383. 314.7 134.8 0
2017 3 2321.9
34.1 31 105.5 77.3 212.1 246.7 388.1 142.1 151. 167.9 58.5 125.4
2018 8 1740.5
49.3 44.5 115.8 180.5 225.1 206.7 223.4 212.9 209. 152.1 83.9 50.6
mean 9 1754.8
SD 36.8 31.1 37.7 64.5 66.5 55.8 79.3 74.4 77.8 78.0 51.5 45.0 58.2
CV 0.7 0.7 0.3 0.4 0.3 0.3 0.4 0.3 0.4 0.5 0.6 0.9 0.5
A.4. Gonga Kemise station

Annua
Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec l
17.5 91.4 219. 78 149 205. 196. 332. 221. 10.6 72.5 5.7 1600.6
1990 6 6 5 7 5
62.3 79 48.3 186. 286. 260 330. 266. 184. 45 0.7 62.6 1811.1
1991 1 2 3 4 2
1992 44.6 56.2 123. 141. 158. 247 191. 211. 175. 180. 102. 6.1 1638.8
2 5 9 5 4 1 6 7
26.7 88.3 70.6 172. 89.5 87 236. 228. 168. 179. 15.2 24.6 1387
1993 1 5 1 5 9
14.5 3.2 49.4 46.8 116. 174. 160. 183. 142. 38.5 0 5.2 934.1
1994 2 6 4 1 2
0 12.7 55.7 84.7 86.5 145. 96.3 160. 75.6 16.4 16.6 42.6 792.3
1995 1 1
20.7 28 115. 67.9 133 81.3 104. 65.2 230. 43.7 55.4 7.7 953.6
1996 6 8 3
37.1 0 23 98.8 117. 158. 155. 127. 123. 201 107. 5.7 1155.9
1997 6 8 4 3 5 7
7 9.9 47.7 70.3 100. 172. 128. 236. 148 170. 48 26 1165.4
1998 5 2 7 9 2
11 0 6.3 46.7 198. 163. 86.2 108. 108. 188 1.1 10.4 928.6
1999 4 5 7 3
0 1.3 17.4 52.5 168. 140 92.4 69.4 84.5 85.4 49.5 11.2 772
2000 4
3.4 43.1 55 52.1 97.8 147. 160. 128 182. 120. 29.3 10.3 1029.5
2001 4 6 2 3
43 4.8 40.9 29.1 147. 154. 159. 125. 90.6 28.5 7.6 23 854.6
2002 3 8 9 1
22.1 55.6 26.4 134. 20.4 332. 255. 135. 205. 2.6 4.9 12.3 1208
2003 9 7 2 5 4
19.8 4.5 23 182. 128. 272. 176. 265. 267. 29.6 10.1 10.3 1390.9
2004 8 9 4 9 1 5
3.3 0 123. 107. 141 188. 188. 143. 334. 67.8 40.4 0 1337.8
2005 1 2 4 2 7 7
16.8 46 142. 314. 148 143. 375. 459. 312. 116. 112. 93.8 2282.6
2006 9 5 2 5 8 8 9 4
55.4 49.8 79.9 59.2 169. 154. 290. 280. 115. 143. 29.7 0 1428.5
2007 2 7 3 6 9 8
8.9 4.1 18.2 57.6 321. 275. 255. 187. 70 120 40.6 4.2 1364.7
2008 8 9 6 8
34.8 36.5 63.3 126. 97.2 139. 232. 235. 162. 133. 7.5 26 1295.1
2009 8 7 3 4 1 5
3.7 32.5 32.8 37 240 341 228. 278. 660. 23.9 2.2 2.1 1882.5
2010 1 6 6
9.2 10 22.3 10.8 195. 192. 256. 170. 220. 131 150 53.1 1422.3
2011 7 6 3 8 5
70 52.2 34.7 85.7 155. 158. 137. 224. 197. 41 74.3 73.2 1304.8
2012 8 8 7 1 3
72.8 45.8 88.2 54.6 224. 176. 125. 119. 175 157. 153. 34.4 1427.9
2013 6 9 3 1 5 7
55 51.3 198 88.8 113. 133. 198. 143. 69.7 140. 22.6 88.7 1302.9
2014 1 5 1 5 6
0 10 46.5 52.5 149. 116. 254 223. 125. 112 97.2 43.2 1231.3
2015 9 5 6 9
9.2 18 46.1 87.3 169 177. 189. 215. 106. 181. 66.2 79.3 1345
2016 8 2 4 1 4
50 87.4 24.7 173. 331. 247. 214 199. 383. 314. 134. 0 2160.9
2017 7 1 8 4 3 7 8
34.1 31 105. 77.3 212. 246. 388. 142. 67.1 154. 58.5 133. 1650.3
2018 5 1 7 1 1 4 4
mea 26.0 32.8 67.2 95.8 160. 187. 202. 195. 186. 109. 52.1 30.9 1346.9
n 9 4 2 4 5 6
22.0 28.7 52.5 62.3 68.8 64.4 77.3 81.7 119. 72.9 46.4 33.9 60.9
std 8
CV 0.8 0.9 0.8 0.7 0.4 0.3 0.4 0.4 0.6 0.7 0.9 1.1 0.7
A.5 Keresi station

104. 292. 497.
1990 3 47.9 59.1 73.9 184 374.9 5 3 497 150.2 58.7 30.9 2370.7
255. 354. 412. 400. 361.
1991 52.8 11.7 85.3 9 3 337.8 4 3 8 129 47 24.2 2472.5
206. 442. 324. 456. 128.
1992 27.3 21.1 107.6 8 233 328.5 9 7 1 343.7 4 43.6 2663.7
108. 119. 387. 304. 233.
1993 37.7 63.1 72.4 5 4 324.2 9 3 8 149.3 15.3 2.5 1818.4
180. 381. 356. 360.
1994 70.3 0 38 5 449 303.3 4 4 2 103.7 57.7 26.7 2327.2
135. 194. 256. 465. 365.
1995 1.5 28.6 114.6 7 1 283.8 9 5 7 147.8 31.8 65.8 2091.8
107. 120. 279. 285. 368.
1996 9 19.7 128.4 5 2 328.5 1 2 275 152.4 91.1 60.5 2216.5
145. 403. 277. 385. 254. 141.
1997 93.2 6.3 88.4 8 2 349.7 6 6 1 222.7 2 14 2381.8
269. 312. 321. 358.
1998 8.5 11.8 119.6 23.4 3 344.8 5 8 3 200.5 24.2 1.7 1996.4
147. 437. 309.
1999 38.9 6.1 5.9 2 4 273.4 281 2 374 323.7 0 56.4 2253.2
115. 311. 352.
2000 17.4 0 22.8 6 3 348.5 232 365 7 337.1 80.7 26.3 2209.4
302. 313. 347.
2001 0.3 22 36.8 68.5 7 408.6 369 3 4 193.9 27.7 48.3 2138.5
2002 40.1 6 108.7 137. 134. 350.4 253. 334. 369. 199.7 42.4 101 2078.9
6 4 9 9 9
107. 205. 328. 368.
2003 7 35.1 133.4 1 176 422.5 5 8 1 139.1 60.5 38.8 2021.9
222. 406. 408. 441.
2004 16.5 58.5 108.3 191 5 464.4 5 3 4 199.9 80.6 0 2597.9
144. 215. 378. 350.
2005 90.8 12.7 108.8 4 1 331.8 2 4 298 168.2 70 2.3 2170.7
266. 393. 344. 412. 173.
2006 14.2 38.5 95.8 42.1 8 383.8 3 9 7 337.2 4 101 2603.6
151. 242. 342. 341. 363. 120.
2007 8.1 53.7 57.3 7 6 243 8 9 3 189.5 2 21 2135.1
103. 255. 363. 315. 355. 329. 138.
2008 5 0 34.1 1 5 313 2 2 7 202.4 9 14.9 2425.5
274. 401. 234. 623.
2009 5.4 64.8 127.3 7 93.2 374.2 2 7 3 190.2 52.1 27.9 2469
232. 364. 440. 113.
2010 71.5 27.3 34 75.7 7 354.8 9 7 372 174 5 97.6 2358.7
127. 339. 358. 424. 351. 133.
2011 29.9 12.4 46 9 7 161.3 7 2 6 92.2 6 4.7 2082.2
364. 331. 420. 201.
2012 4.8 1.7 73.1 64.8 1 296.5 8 2 464 115 8 14.9 2352.7
212. 382. 249. 345. 311.
2013 6.3 86.8 32.6 1 2 280.7 7 4 5 219.6 79.2 23.6 2229.7
193. 170. 268. 243.
2014 90 74.1 271.4 9 9 269.2 7 3 62.2 128.9 173 23 1968.6
214. 203. 157. 217.
2015 83.3 97.4 134.5 18.4 6 290.8 9 7 4 141.8 145 92.2 1797
241. 230. 243.
2016 16.8 40.3 68.4 90.3 287 243.3 7 2 5 108.1 17.4 0 1587
109. 277. 329.
2017 18.4 98.9 179.5 254 37.7 203.1 8 4 5 201 132 54.2 1895.5
146. 181. 163. 163. 318. 185. 151.
2018 3 2 150.8 1 5 149.5 4 6 75.1 242 8 152 2079.1
140. 256. 312. 339. 340.
mean 45.3 38.9 91.1 9 7 315.1 9 2 3 189.8 89.3 40.3 2199.8
101. 110.
std 40.6 39.8 54.4 69.1 6 70.1 75.6 77.4 4 69.2 54.0 37.2 66.6
CV 0.9 1.0 0.6 0.5 0.4 0.2 0.2 0.2 0.3 0.4 0.6 0.9 0.5
A.6. Kemise
229. 378. 450. 331.
1990 21.2 31.5 81.5 82.7 5 377 2 8 8 84.5 31.6 32.1 2132.4
219. 271. 337. 245. 274.
1991 52.1 33.6 54 2 8 162.1 3 1 1 20.9 1.8 21.9 1693.9
192. 226. 234. 325. 367.
1992 45.9 19.6 46.9 3 9 280.2 8 9 1 267.8 21.5 19.5 2048.4
211. 230. 214. 396. 424.
1993 17.6 67 103.1 7 1 294.9 4 9 6 172.9 18.2 0 2151.4
118. 316. 416. 378. 256.
1994 16.6 0 27.7 6 4 257.8 1 2 6 50.7 11.8 8.1 1858.6
162. 357.
1995 1.2 51.6 166.1 84.3 9 195.3 304 405 8 78.1 11.4 56.9 1874.6
117. 268. 381. 367. 358.
1996 86.8 6.9 168.5 7 3 275.3 5 9 7 129.6 44.4 34.4 2240
304. 367. 386. 309.
1997 58.1 10.6 77.3 212 1 485.1 4 2 1 309.7 3.4 17.3 2540.3
148. 452. 407. 349.
1998 12 13.3 103.9 51.4 1 432 4 9 5 312.2 65.2 2.5 2350.4
153. 270. 470. 473.
1999 2 4.4 7.5 2 327 307.2 9 5 9 266 0 46.2 2328.8
206. 279. 408. 234. 303.
2000 11.1 1.2 12.6 7 9 277.9 4 9 2 170.3 24.4 9.5 1940.1
236. 278. 262. 385.
2001 0 31.9 53.1 36.9 8 416.1 3 1 6 177.2 40.9 59.6 1978.5
117. 230. 339. 212. 192.
2002 23.5 0 114.4 5 4 321.8 6 5 9 151.7 4.6 34.5 1743.4
360. 275. 378.
2003 7.8 51.7 101.9 71.5 75.8 315.1 4 4 7 32.8 27 24.7 1722.8
110. 211. 221. 209. 303.
2004 15.8 11 45.3 2 1 267 3 3 1 139.7 70.4 32.9 1637.1
160. 210. 273. 345. 356.
2005 8 21.9 133.5 57.1 4 346.7 1 4 9 173 54.4 14.4 2147.6
214. 355. 291. 355.
2006 0 41.4 50.9 30.9 1 295.1 9 2 6 251.1 99.4 80 2065.6
110. 189. 275. 504. 432.
2007 15.9 32.9 99.7 2 2 417.1 7 2 8 198.2 28.5 0 2304.4
262. 445. 251.
2008 45.4 64.4 0 6 1 327 430 303 4 394.3 76.4 26 2625.6
100. 193. 577. 341. 451.
2009 48.3 5 100.7 3 25.5 371.5 5 9 8 226.6 72.6 16.8 2527
110. 340. 656.
2010 24.3 0 20.5 49.9 2 446.1 9 2 644 89.4 67.2 102 2550.5
434. 441. 242. 146.
2011 0 0 61.4 5 6 519.1 7 272 78.3 100.2 6 10.5 2306.9
2012 59.7 86.4 90.5 116. 264. 276 382. 301. 38.9 54.6 158. 216 2046.5
6 1 8 8 9
721. 168. 536. 227.
2013 0 2.2 0 74 6 359.7 5 4 4 118.1 96.4 69.1 2373.4
164. 223.
2014 65.4 148 192.7 6 7 194.3 155 102 82.3 159.2 0.5 0.8 1488.5
163. 100. 311. 305. 206.
2015 8 83 125.9 7 6 185.9 367 240 6 142.9 9 206 2439
102. 156. 401. 722. 354.
2016 6.4 6 31.9 4 2 230.1 6 7 9 556 6.3 36.7 2611.2
195. 206. 126. 183. 346. 109.
2017 15.1 81 54 7 9 385.1 4 7 4 303.5 8 24.8 2032.4
158. 238. 213. 393. 243. 156.
2018 4 98.9 143.3 5 6 285.7 6 3 80.9 206.1 3 139 2357.4
142. 250. 326. 347. 312.
mean 39.1 38.0 78.2 0 1 320.8 1 3 9 184.0 57.1 46.3 2142.0
125. 135. 127.
std 47.2 38.1 51.6 85.0 5 88.2 97.2 2 9 115.1 54.5 54.6 85.0
CV 1.2 1.0 0.7 0.6 0.5 0.3 0.3 0.4 0.4 0.6 1.0 1.2 0.7
B) Monthly maximum temperature for Bonga Station
Annua
year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec l
1990 28.16 26.80 27.63 28.07 27.88 26.19 26.02 26.74 25.89 28.68 28.60 29.30 27.50
1991 28.61 29.60 28.42 27.79 27.74 27.31 25.60 25.85 26.70 27.67 28.03 27.82 27.60
1992 27.85 27.84 29.15 28.21 28.34 26.23 25.81 25.19 25.81 26.45 28.03 27.94 27.24
1993 26.79 26.97 28.52 27.11 27.67 27.38 26.34 27.22 27.11 26.81 29.41 30.10 27.62
1994 31.15 32.04 29.34 28.45 26.78 26.82 25.37 25.51 26.83 29.10 27.01 27.92 28.03
1995 29.99 29.50 29.80 28.07 27.16 27.13 25.15 25.72 26.68 27.75 29.49 27.74 27.85
1996 26.56 29.22 28.25 26.66 26.43 25.52 24.27 24.08 25.60 27.35 27.72 27.05 26.56
1997 27.44 29.74 28.99 26.73 25.89 25.63 25.15 25.63 27.08 26.48 26.35 27.85 26.91
1998 28.00 28.39 27.38 29.14 26.58 27.23 25.03 24.56 25.78 26.42 28.51 29.28 27.19
1999 29.10 28.48 28.62 27.62 26.10 26.42 24.59 25.06 26.37 26.44 28.34 28.89 27.17
2000 28.80 28.93 28.62 27.68 26.58 26.61 25.24 24.81 25.56 26.38 26.45 27.37 26.92
2001 27.55 26.81 25.96 27.09 25.60 24.53 24.55 24.87 25.89 26.78 26.35 26.18 26.01
2002 27.12 28.50 26.61 26.69 26.27 25.88 26.03 25.28 25.86 26.85 26.66 25.80 26.46
2003 26.49 27.82 27.28 27.29 26.62 25.04 24.74 24.43 25.53 26.82 26.81 27.68 26.38
2004 27.28 23.93 30.34 28.80 21.69 25.29 25.94 25.64 25.75 26.59 27.72 27.64 26.38
2005 28.34 30.85 30.32 29.14 26.28 26.09 25.14 26.31 26.54 26.86 27.52 28.24 27.64
2006 29.05 29.68 28.39 27.50 26.77 26.61 25.39 24.61 25.37 27.17 27.37 27.21 27.09
2007 26.98 27.81 29.11 28.21 27.06 26.30 25.94 25.92 26.26 27.89 28.19 28.69 27.36
2008 28.08 29.38 30.65 27.64 26.48 25.67 25.66 25.13 26.16 27.26 27.43 28.18 27.31
2009 28.66 28.80 28.81 27.87 28.03 27.22 25.98 26.35 25.90 26.76 28.65 27.00 27.50
2010 28.11 28.27 28.71 27.92 26.97 26.10 26.60 26.34 26.82 27.61 29.03 28.53 27.58
2011 28.27 28.96 28.27 27.98 28.05 26.11 26.44 25.85 26.65 28.79 28.77 29.66 27.82
2012 30.50 31.26 30.37 27.98 27.40 26.37 25.29 26.53 26.10 28.40 28.67 29.52 28.20
2013 30.18 31.34 30.68 30.32 26.73 26.14 25.34 25.52 26.66 27.24 28.40 28.94 28.12
2014 29.16 29.43 27.53 27.29 27.31 26.65 26.67 26.36 26.56 26.84 26.82 28.39 27.42
2015 27.66 28.40 28.71 28.10 27.26 27.29 26.48 27.76 27.50 26.79 26.20 27.23 27.45
2016 28.08 27.95 30.73 27.93 27.05 27.55 26.56 25.73 27.35 27.56 29.23 29.60 27.94
2017 1.10 5.78 5.57 11.15 13.25 12.09 10.92 9.48 15.36 24.68 27.00 27.34 13.64
2018 30.44 32.13 30.16 30.38 31.40 31.02 12.81 21.19 25.34 27.75 26.48 26.29 27.12
Mea
n 27.43 28.09 28.03 27.41 26.46 26.02 24.66 24.95 25.90 27.18 27.77 28.05 26.83
SD 5.12 4.54 4.41 3.20 2.88 2.86 3.55 3.14 2.08 0.88 1.00 1.07 2.89
CV 0.19 0.16 0.16 0.12 0.11 0.11 0.14 0.13 0.08 0.03 0.04 0.04 0.11
C) Monthly minimum temperature for Bonga Station
Annua
year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec l
1990 10.42 12.87 13.04 14.04 14.41 14.20 13.36 13.49 13.27 12.19 11.62 10.20 12.76
1991 12.08 10.97 12.73 11.21 10.83 10.16 10.25 13.01 12.89 12.21 11.93 9.77 11.50
1992 8.69 10.84 12.20 11.14 12.51 10.96 10.99 10.25 10.92 11.04 12.19 11.98 11.14
1993 11.35 10.56 12.05 12.53 12.56 12.70 11.27 11.70 12.16 12.64 10.48 8.19 11.52
1994 8.95 10.79 12.96 12.99 13.88 11.85 13.71 13.71 12.62 8.31 10.09 9.18 11.59
1995 8.77 10.84 10.40 12.50 13.05 12.25 11.38 11.65 12.55 9.72 9.42 10.96 11.12
1996 11.47 11.07 11.09 9.62 10.68 10.12 10.47 11.00 11.52 11.94 10.85 10.23 10.84
1997 10.81 9.96 12.20 12.75 12.96 13.06 12.75 11.60 11.45 11.31 11.45 11.30 11.80
1998 11.11 12.19 12.27 11.42 11.25 12.11 11.81 11.94 11.41 12.11 9.58 8.21 11.28
1999 8.87 10.23 11.86 11.94 11.55 11.93 11.30 11.59 12.28 12.66 9.76 9.24 11.10
2000 9.50 7.77 12.03 12.31 12.51 12.85 12.95 13.64 12.92 13.29 11.96 8.90 11.72
2001 9.50 11.91 12.17 14.02 12.76 13.16 12.59 13.39 13.14 13.34 12.34 11.53 12.49
2002 11.82 10.13 12.01 13.91 13.42 12.95 13.24 12.94 13.16 12.85 12.73 14.05 12.77
2003 10.76 11.69 12.25 13.46 14.21 11.99 10.91 10.78 11.53 10.19 10.10 10.46 11.53
2004 10.47 10.60 12.49 15.02 14.64 14.42 13.02 14.00 13.72 11.76 11.11 13.94 12.93
2005 11.41 11.37 13.32 13.45 13.05 14.47 13.95 13.27 14.11 13.16 10.80 8.02 12.53
2006 11.77 12.30 13.14 14.06 13.15 14.13 14.07 15.04 13.94 14.28 13.15 12.95 13.50
2007 13.58 13.29 13.39 14.79 14.66 14.46 14.03 13.92 14.28 11.53 10.07 6.52 12.88
2008 10.65 9.84 9.81 12.23 12.13 11.98 11.82 12.19 12.42 12.03 9.13 8.26 11.04
2009 8.50 10.21 12.18 13.52 12.95 12.52 12.02 12.85 12.18 11.68 9.20 11.02 11.57
2010 11.00 10.96 11.37 12.30 13.08 12.18 12.23 12.60 12.27 10.72 8.46 8.95 11.34
2011 11.06 12.00 12.92 12.82 13.24 12.72 12.11 12.31 12.33 12.26 11.98 9.13 12.07
2012 8.05 11.10 11.98 12.32 11.74 11.53 11.06 11.11 11.55 8.47 10.81 10.60 10.86
2013 10.94 10.96 12.10 12.58 12.89 12.68 12.15 12.06 12.46 12.50 11.60 7.68 11.72
2014 9.19 9.72 12.27 11.87 13.23 12.72 11.76 11.73 12.30 12.87 11.61 11.05 11.69
2015 12.21 11.51 11.76 13.12 12.55 11.82 11.93 13.23 13.63 11.88 12.50 12.27 12.37
2016 12.21 11.36 12.08 11.96 13.94 13.43 12.28 11.55 11.73 13.87 9.63 9.64 11.97
2017 7.05 13.29 13.21 14.00 14.21 13.50 13.70 13.64 13.35 12.10 11.24 10.59 12.49
2018 10.42 11.36 10.89 11.54 11.37 12.05 26.89 12.01 11.71 11.04 13.15 11.80 12.85
Mea
n 10.43 11.09 12.14 12.74 12.88 12.58 12.76 12.49 12.54 11.86 11.00 10.23 11.90
SD 1.46 1.12 0.82 1.17 1.07 1.12 2.87 1.11 0.88 1.37 1.26 1.81 1.34
CV 0.14 0.10 0.07 0.09 0.08 0.09 0.23 0.09 0.07 0.12 0.11 0.18 0.11
D) Relative humidity of the study catchment based on CFSR

Annua
1990 0.46 0.53 0.57 0.68 0.67 0.66 0.71 0.73 0.63 0.59 0.58 0.48 0.61
1991 0.46 0.43 0.48 0.59 0.63 0.66 0.69 0.69 0.64 0.57 0.57 0.45 0.57
1992 0.43 0.44 0.43 0.55 0.63 0.67 0.68 0.71 0.67 0.64 0.53 0.54 0.58
1993 0.52 0.50 0.42 0.56 0.70 0.71 0.71 0.66 0.59 0.62 0.55 0.42 0.58
1994 0.31 0.32 0.44 0.62 0.70 0.71 0.70 0.69 0.61 0.59 0.59 0.45 0.56
1995 0.36 0.45 0.49 0.61 0.68 0.64 0.68 0.61 0.67 0.66 0.62 0.54 0.59
1996 0.49 0.40 0.53 0.66 0.69 0.78 0.79 0.76 0.74 0.67 0.59 0.49 0.63
1997 0.48 0.29 0.46 0.74 0.71 0.67 0.70 0.67 0.60 0.71 0.78 0.70 0.62
1998 0.65 0.48 0.52 0.62 0.71 0.70 0.77 0.72 0.64 0.68 0.55 0.40 0.62
1999 0.41 0.34 0.53 0.61 0.68 0.64 0.67 0.62 0.59 0.64 0.43 0.47 0.55
2000 0.33 0.26 0.39 0.58 0.63 0.60 0.62 0.63 0.58 0.66 0.54 0.43 0.52
2001 0.38 0.39 0.47 0.64 0.62 0.65 0.63 0.63 0.62 0.65 0.63 0.50 0.57
2002 0.47 0.32 0.52 0.61 0.69 0.61 0.57 0.60 0.56 0.62 0.54 0.60 0.56
2003 0.47 0.35 0.45 0.62 0.66 0.63 0.62 0.65 0.55 0.55 0.58 0.47 0.55
2004 0.50 0.43 0.41 0.69 0.65 0.65 0.63 0.62 0.61 0.58 0.62 0.54 0.58
2005 0.40 0.35 0.53 0.58 0.74 0.69 0.65 0.62 0.63 0.62 0.50 0.30 0.55
2006 0.37 0.41 0.50 0.68 0.71 0.68 0.67 0.71 0.61 0.69 0.64 0.66 0.61
2007 0.55 0.51 0.48 0.63 0.68 0.74 0.68 0.65 0.69 0.57 0.53 0.37 0.59
2008 0.42 0.37 0.46 0.62 0.64 0.65 0.66 0.63 0.63 0.67 0.56 0.42 0.56
2009 0.48 0.46 0.49 0.66 0.66 0.61 0.58 0.59 0.60 0.65 0.50 0.64 0.58
2010 0.51 0.55 0.61 0.71 0.76 0.71 0.72 0.70 0.70 0.69 0.54 0.47 0.64
2011 0.46 0.35 0.49 0.61 0.72 0.72 0.70 0.73 0.69 0.64 0.73 0.59 0.62
2012 0.43 0.32 0.43 0.70 0.72 0.71 0.74 0.71 0.69 0.65 0.67 0.62 0.61
2013 0.51 0.46 0.62 0.73 0.73 0.73 0.74 0.72 0.68 0.69 0.68 0.53 0.65
2014 0.49 0.47 0.58 0.68 0.71 0.69 0.71 0.63 0.63 0.67 0.56 0.42 0.60
2015 0.48 0.46 0.49 0.66 0.66 0.61 0.58 0.59 0.60 0.65 0.50 0.64 0.58
2016 0.51 0.55 0.61 0.72 0.76 0.71 0.72 0.70 0.69 0.68 0.53 0.48 0.64
2017 0.46 0.34 0.50 0.62 0.72 0.72 0.69 0.73 0.69 0.64 0.73 0.57 0.62
2018 0.43 0.32 0.43 0.70 0.72 0.71 0.74 0.71 0.69 0.65 0.67 0.62 0.61
Mea
n 0.46 0.41 0.49 0.64 0.69 0.68 0.68 0.67 0.64 0.64 0.59 0.51 0.59
SD 0.07 0.08 0.06 0.05 0.04 0.04 0.05 0.05 0.05 0.04 0.08 0.09 0.06
CV 0.15 0.19 0.12 0.08 0.06 0.06 0.08 0.07 0.07 0.06 0.13 0.18 0.11
E) Wind speed of the study catchment station at 2m based on CFSR

Annua
1990 1.84 1.57 1.71 1.64 1.79 1.69 1.56 1.60 1.60 1.92 1.71 1.92 1.71
1991 1.88 1.85 1.93 1.86 1.93 1.75 1.60 1.68 1.61 1.94 1.74 1.92 1.81
1992 2.00 1.89 1.86 1.88 1.83 1.64 1.57 1.48 1.62 1.78 1.77 1.74 1.75
1993 1.65 1.83 1.87 1.74 1.69 1.60 1.51 1.66 1.75 1.86 1.85 1.94 1.75
1994 1.92 1.96 1.82 1.84 1.79 1.68 1.51 1.54 1.62 1.89 1.92 2.00 1.79
1995 2.03 1.72 1.93 1.76 1.64 1.62 1.55 1.76 1.70 1.78 1.78 1.66 1.74
1996 1.69 1.75 1.78 1.75 1.71 1.42 1.45 1.47 1.47 1.84 1.77 1.84 1.66
1997 1.66 1.91 1.76 1.52 1.66 1.63 1.58 1.69 1.76 1.57 1.38 1.50 1.64
1998 1.36 1.64 1.71 1.73 1.66 1.63 1.62 1.50 1.61 1.84 1.87 1.98 1.68
1999 1.87 2.01 1.97 1.86 1.85 1.76 1.59 1.74 1.76 1.76 2.05 1.98 1.85
2000 2.02 2.14 2.04 1.94 1.87 1.80 1.70 1.76 1.81 1.80 1.87 1.84 1.88
2001 1.81 1.95 1.72 1.83 1.89 1.83 1.73 1.84 1.72 1.86 1.78 1.82 1.81
2002 1.89 1.81 1.83 1.98 1.92 1.78 1.86 1.75 1.88 2.10 1.93 1.57 1.86
2003 1.85 1.98 1.93 1.94 2.10 1.70 1.66 1.64 1.81 1.98 1.89 1.97 1.87
2004 1.75 1.90 1.87 1.74 1.91 1.76 1.72 1.68 1.68 1.88 1.83 1.80 1.79
2005 1.92 2.00 1.85 1.86 1.69 1.63 1.68 1.70 1.65 1.93 2.06 2.25 1.85
2006 1.93 1.93 1.75 1.78 1.81 1.68 1.70 1.53 1.65 1.75 1.81 1.78 1.76
2007 1.70 1.65 1.85 1.80 1.75 1.65 1.57 1.61 1.63 1.91 1.98 2.21 1.78
2008 1.87 1.90 1.98 1.98 1.87 1.75 1.55 1.54 1.67 1.90 1.95 2.00 1.83
2009 1.76 1.76 1.81 1.82 1.86 1.76 1.77 1.78 1.70 1.63 1.96 1.52 1.76
2010 1.61 1.47 1.78 1.54 1.56 1.45 1.43 1.58 1.50 1.75 1.90 1.92 1.62
2011 1.75 1.86 1.78 1.65 1.50 1.50 1.55 1.31 1.42 1.78 1.57 1.71 1.61
2012 1.93 2.03 1.83 1.61 1.70 1.53 1.34 1.45 1.47 1.74 1.64 1.66 1.66
2013 1.63 1.76 1.59 1.74 1.49 1.46 1.35 1.41 1.44 1.62 1.56 1.78 1.57
2014 1.67 1.71 1.62 1.60 1.50 1.46 1.46 1.54 1.67 1.90 1.95 2.00 1.67
2015 1.76 1.76 1.81 1.82 1.86 1.76 1.77 1.78 1.70 1.63 1.96 1.52 1.76
2016 1.61 1.51 1.76 1.51 1.57 1.46 1.43 1.56 1.51 1.77 1.90 1.90 1.63
2017 1.75 1.86 1.78 1.63 1.48 1.52 1.56 1.30 1.43 1.77 1.56 1.73 1.62
2018 1.92 2.04 1.83 1.61 1.70 1.53 1.34 1.45 1.47 1.74 1.64 1.66 1.66
Mea
n 1.79 1.83 1.82 1.76 1.74 1.63 1.58 1.60 1.63 1.81 1.81 1.83 1.74
SD 0.15 0.16 0.10 0.13 0.15 0.12 0.13 0.14 0.12 0.12 0.16 0.19 0.14
CV 0.08 0.09 0.05 0.08 0.09 0.07 0.08 0.09 0.08 0.06 0.09 0.10 0.08
Annex 7: Estimated monthly flow at the weir site for the period 1990-2018
Annua
1990 0.67 0.73 2.54 4.07 7.32 13.17 10.19 16.14 15.23 7.61 2.38 2.19 6.85
1991 5.61 8.78 7.63 1.07 0.96 2.05 13.21 15.35 15.54 5.62 1.74 1.40 6.58
1992 3.32 3.59 1.40 1.30 6.25 9.86 8.98 12.92 14.29 21.38 4.48 1.57 7.45
1993 1.55 2.11 2.03 4.74 10.49 13.68 12.18 9.43 11.79 9.97 2.90 1.65 6.88
1994 1.90 0.86 1.59 1.31 6.87 14.92 15.48 15.21 11.34 2.90 2.63 1.48 6.37
1995 1.11 0.98 0.48 1.46 3.22 6.36 9.45 13.29 14.75 5.26 1.47 1.36 4.93
1996 2.10 1.10 1.74 5.28 11.05 15.43 11.84 13.82 15.63 8.55 1.96 1.25 7.48
1997 1.50 0.81 2.10 6.47 10.58 19.60 16.17 11.32 8.80 15.85 17.02 7.17 9.78
1998 5.41 3.18 3.66 4.35 16.17 16.25 16.30 30.46 14.19 23.22 3.27 1.16 11.47
1999 1.29 1.23 1.45 2.07 3.85 5.82 9.37 7.59 6.41 10.39 2.23 1.01 4.39
2000 0.56 0.48 0.68 1.08 2.68 2.71 10.07 4.52 5.19 11.85 2.00 1.01 3.57
2001 0.65 0.66 0.77 1.30 6.55 14.35 11.77 13.63 16.91 8.94 2.63 1.45 6.63
2002 2.28 1.56 1.90 3.05 2.72 3.59 8.69 8.64 6.63 3.51 2.58 3.58 4.06
2003 2.08 1.70 1.64 3.12 2.64 10.59 11.83 14.61 14.99 4.35 2.33 2.07 6.00
2004 1.04 0.92 2.10 1.28 3.48 8.38 9.69 13.86 16.64 9.41 2.50 2.71 6.00
2005 4.74 5.55 4.59 2.06 4.59 7.90 10.25 6.93 6.74 8.85 2.06 2.74 5.58
2006 8.04 7.96 2.45 1.84 1.51 5.47 11.32 13.61 13.06 7.79 5.66 6.47 7.10
2007 2.26 1.51 1.91 2.78 5.50 3.87 15.95 22.15 13.35 7.08 2.61 1.45 6.70
2008 1.66 1.17 2.08 7.17 11.80 18.57 13.34 17.28 11.21 18.34 4.10 1.69 9.03
2009 2.06 2.94 2.47 2.05 7.40 12.87 10.79 10.73 13.53 5.92 1.99 1.73 6.21
2010 2.23 0.87 1.81 1.16 7.34 2.11 9.52 11.41 16.00 9.03 2.20 1.22 5.41
2011 0.64 0.84 1.87 6.09 9.28 9.87 9.76 11.27 9.19 7.31 3.27 1.57 5.91
2012 1.60 1.37 1.97 4.04 9.78 7.38 10.06 23.58 22.60 4.70 1.53 1.16 7.48
2013 0.65 0.60 1.04 2.51 5.12 15.47 9.93 9.96 10.29 10.53 5.49 1.55 6.09
2014 2.12 0.91 2.64 3.98 12.89 36.55 13.28 17.36 18.67 10.79 2.51 1.19 10.24
2015 1.25 0.95 1.88 2.72 5.86 2.43 11.46 9.94 6.27 36.55 12.47 1.14 7.74
2016 1.48 1.79 1.43 2.83 3.52 4.44 9.98 9.12 7.06 2.45 1.83 1.16 3.92
2017 0.73 1.21 1.88 1.96 5.55 10.97 17.93 22.02 8.83 3.33 2.63 1.23 6.52
2018 1.74 1.07 1.82 1.15 4.30 12.66 13.89 15.53 10.29 7.29 2.20 1.29 6.10
Mean
(m3/s 2.15 1.98 2.12 2.91 6.53 10.60 11.82 13.85 12.26 9.96 3.54 1.95
) 6.64
SD 1.71 2.04 1.31 1.72 3.64 7.08 2.50 5.40 4.20 7.12 3.27 1.44 3.45
CV 0.80 1.03 0.62 0.59 0.56 0.67 0.21 0.39 0.34 0.71 0.92 0.74 0.63

Untitled

Uploaded by

Copyright:

Available Formats

Untitled

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Untitled

Uploaded by

Copyright:

Available Formats

TABLE OF CONTENTS

Table 3. 1: List of meteorological stations representing the study catchment...........................8

 Developing Intensity-Duration-Frequency (IDF) for the selected rainfall stations

1.3. Scopes of the study

1.4. Review of the previous study

2. CHARACTERISTICS OF BONGA CATCHMENT

Figure 2. 1: Spatial distribution of terrain slope of Bonga catchment

2.1.2. Land use/cover

Table 3. 3: Locations of the proposed headwork site

1 Option -1 Guma R. 204200.00 800136.00 Weir Selected

3.2. Data Processing and Quality Control

3.2.2. Data Quality Control

Table 3. 4: Statistical test results in time series data of Bonga Station.

Table 3. 5: Statistical test results in time series data of Shadatura Station.

3.3.2. Rainfall Frequency Analysis (RFA)

Figure 3. 10: CDF of Bonga Station as determined from 17 candidate distributions

3.3.4. Other climatic parameters

4. Estimation of Flow and Sediment

4.1.1. Description of SWAT model

4.1.2. SWAT Input

Figure 4. 1: SWAT model setup and sub watershed discretization

4.1.3. Model calibration schemes

4.1.4. Model performance Evaluation

4.2. SWAT model results: Flow and Sediment

Table 4. 1: Optimized SWAT model Parameters for the study catchment

4.3. Estimating Dependable Flow

5.1. Flood estimation using statistical flood frequency analysis

5.1.1. Selection of the best-fit regional frequency distribution

The choice of frequency distributions is determined based on goodness-of-fit measures, which

5.1.2. Prediction of flood at the weir site (Ungauged site)

Table 5. 2: Annual Maximum series flow at Dincha Gauging station

5.2. Low-flow and baseflow analysis

Table 5. 4: Low flow analysis

100 0.10 0.18

Figure 5. 3: 1-day Low-flow frequency analysis for different return levels

Figure 5. 4: Streamflow and baseflow hydrograph for the period 1990-2003

6.1. Design Flood Estimation for Cross-Drainage Structures

CD-1 195421.7838 805402.764 41.0

CD-2 195792.3441 807028.2401 122.9

CD-3 197863.2606 800232.5814 7.0

CD-4 202118.2264 798685.3211 12.4

6.1.1. SCS Method

 Curve Number of the catchment area of CD

Q= (P- Ia)2 / (P - Ia) + S (6.2)

Where co=-2.23*(Ia/p)^2+0.47*(Ia/p)+2.52, c1=1.5*(Ia/P)^2-0.7*(Ia/p)-0.5),

IDF Curve of Bonga Station

Figure 6. 2: IDF curve for Bonga Station

Test Critical values(Statistical table) Result

Statistical test Test Critical values(Statistical table) Result

Test Critical values(Statistical table)

A2: Wushiwesh Station

A.4. Gonga Kemise station

A.5 Keresi station

D) Relative humidity of the study catchment based on CFSR

E) Wind speed of the study catchment station at 2m based on CFSR

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Where co=-2.23(Ia/p)^2+0.47(Ia/p)+2.52, c1=1.5(Ia/P)^2-0.7(Ia/p)-0.5),