Modelling of the Hydrology, Soil Erosion and Sediment Modelling of the Hydrology, Soil Erosion and Sediment

transport processes in the Lake Tana Catchmentstransport processes in the Lake Tana Catchments

of Blue Nile River Basin, Ethiopiaof Blue Nile River Basin, Ethiopia

Combining Field Data, Mathematical Models and Geographic Combining Field Data, Mathematical Models and Geographic

Information Systems (GIS)Information Systems (GIS)

ByBy Shimelis Gebriye SetegnShimelis Gebriye SetegnShimelis Gebriye SetegnShimelis Gebriye SetegnShimelis Gebriye SetegnShimelis Gebriye SetegnShimelis Gebriye SetegnShimelis Gebriye Setegn

Department of Land and Water Resources EngineeringDepartment of Land and Water Resources Engineering

Division of Hydraulic EngineeringDivision of Hydraulic Engineering

The Royal Institute of Technology, Stockholm, SwedenThe Royal Institute of Technology, Stockholm, Sweden

CALIBRATION and Validation of SWAT2005/ArcSWAT in A njeni Gauged Watershed, Northern Highlands of Ethiopia

Financed by Nile Basin Initiative Financed by Nile Basin Initiative –– NBINBI--ATPATP

Outline

�� Background (Project Summery)Background (Project Summery)�� General objectiveGeneral objective�� Specific objective/immediate activitySpecific objective/immediate activity�� Description of the study areaDescription of the study area�� Modelling tools and methodsModelling tools and methods�� Results and conclusionResults and conclusion

�� Soil erosion & Land degradation are major problems on Soil erosion & Land degradation are major problems on the Ethiopian highlands.the Ethiopian highlands.

�� Eroded sediments carry nutrients, herbicides, and Eroded sediments carry nutrients, herbicides, and pesticidespesticides�� degrade the quality of streams, rivers and lakes.degrade the quality of streams, rivers and lakes.

�� The soil depth is less than 35 cm in 34% of the land.The soil depth is less than 35 cm in 34% of the land.

�� Ethiopia loses an est. 1.3 billion metric tons of fertile soil Ethiopia loses an est. 1.3 billion metric tons of fertile soil every year.every year.

Background

� The poor land use practices and management system has played significant role.

� Effective measures are not taken sufficently to combat the soil erosion and sedimentation problems

� due to lack of relevant decision-making tools.

� Adequate research is lacking for understanding the process and effects of soil erosion and sedimentation

� It is important to model the physical processes in the Lake Tana watershed in particular and the Blue Nile River Basin in general

Background Cont.

General Objectives of the study

1. Evaluate the performance and applicability of the watershed model in predicting the hydrology, soil erosion and sediment transport

2. Modelling of the hydrology and soil erosion yield from the catchments and sediment transport into the lake

3. Water quality Modelling

4. Working towards the identification of best land and water management practices

�Specific Objective of the study

1. Calibration and validation of the SWAT2005 (ArcSWAT)1. Flow

2. Modelling of the hydrology, soil Erosion, sediment and pollutant transport in Lake Tana basin

2. Sediment3. Nutrient

Description of the study area

•• Ethiopia is known for its extensive water resources Ethiopia is known for its extensive water resources potential. potential.

•• The source of the Nile river and many transThe source of the Nile river and many trans--boundary boundary rivers such as rivers such as BaroBaro--AkoboAkobo, , WabiWabi ShebeleShebele, , TekezeTekeze. .

•• The total annual runoff is estimated at 110 Billion m3The total annual runoff is estimated at 110 Billion m3

Ethiopian Major River Basins

• Blue Nile River has an average annual run off of about 52 B m3.

• This basin contributes an average of 62% to Nile river.

Blue Nile River Basin

Lake Tana Basin • Lake Tana is the source of the Blue Nile

• The lake covers 3000 - 3630 km2 area

• Elevation of 1800 m and with a maximum depth of 15m.

• Rainfall averages 1315 mm/year

• Four perennial rivers and numerous seasonal streams feed the lake.

••Its altitude ranges from 2407 Its altitude ranges from 2407 --25072507 m m aslasl..

•• Hydrological catchment area is Hydrological catchment area is 113.4 ha.113.4 ha.

••Mean annual rainfall and Mean annual rainfall and temperature are 1690temperature are 1690 mm and mm and 1616 °°C.C.

Anjeni Gauged Watershed

Modelling tools and methods

( )∑=

−−−−+=t

igwseepasurfday QwEQRSWSW t

10

�� Runoff is predicted separately for each HRU using SCS Runoff is predicted separately for each HRU using SCS curve methodcurve method

�� Potential evapotranspiration was estimated using the Potential evapotranspiration was estimated using the Penman Penman MonteithMonteith equation. equation.

�� SWAT2005SWAT2005�� ArcGIS/ArcSWATArcGIS/ArcSWAT�� The hydrologic simulation is based on the water balance The hydrologic simulation is based on the water balance

equationequation

�� Calibration MethodCalibration Method�� ParasolParasol

�� Variation MethodVariation Method�� Multiply by value (%)Multiply by value (%)

�� Objective functionObjective function�� Sum of the Squared of the Residuals (SSQ)Sum of the Squared of the Residuals (SSQ)

�� Average and threshold CriteriaAverage and threshold Criteria�� AverageAverage

Modelling tools and methods Contd.

Model Input

� GIS input files used for the SWAT model include• the digital elevation model (DEM),• land cover, and • soil layers.

�� Soil physical propertiesSoil physical properties

�� Metrological DataMetrological Data�� 10 years precipitation, air temperature, solar 10 years precipitation, air temperature, solar

radiation, wind speed and relative humidity.radiation, wind speed and relative humidity.

�� 10 years River Discharge 10 years River Discharge

�� Land ManagementLand Management

Model Input Contd.

Sensitivity analysis

�� Twenty six hydrological parameters were tested for Twenty six hydrological parameters were tested for sensitivity analysis for the simulation of the stream sensitivity analysis for the simulation of the stream flow.flow.

�� Eighteen parameters were found to be sensitive.Eighteen parameters were found to be sensitive.

�� Eight parameters had shown no effect on the monthly Eight parameters had shown no effect on the monthly stream flow simulationsstream flow simulations

�� The most eight sensitive ones were considered for The most eight sensitive ones were considered for calibration processescalibration processes

Results and Discussion

8Soil depth [mm]Sol_Z

7Groundwater "revap"

coefficientGw_Revap

6Soil evaporation

compensation factorEsco

5Channel effective hydraulic

conductivity [mm/hr]Ch_K2

4

Available water capacity [mm WATER/mm soil]Sol_Awc

3Groundwater delay [days]Gw_Delay

2Base flow alpha factor

[days]Alpha_Bf

1Initial SCS CN II valueCn2

RankDescriptionParameter

27Temperature lapse rate [ °C/km]Tlaps

27Snow pack temperature lag

factorTimp

27Snow melt base temperature

[ºC]Smtmp

27Snowfall temperature [ºC]Sftmp

27

Melt factor for snow on December 21 [mm H2O/ºC-day]Smfmn

27Melt factor for snow on June

21 [mm H2O/ºC-day]Smfmx

27

Threshold water depth in the shallow aquifer for flow [mm]Gwqmn

27

Threshold water depth in the shallow aquifer for "revap" [mm] Revapmin

RankDescriptionParameter

Sensitive parameters considered Sensitive parameters considered for calibration processesfor calibration processes

Parameters shown no effect on the Parameters shown no effect on the monthly stream flow simulationsmonthly stream flow simulations

Model Calibration

�� Manual and automatic calibration methodManual and automatic calibration method

�� Seven years, Jan 1984Seven years, Jan 1984--Dec 1991, hydrometric Dec 1991, hydrometric flow data were utilizedflow data were utilized

�� A good agreement between observed and A good agreement between observed and simulated flows at Anjeni station. simulated flows at Anjeni station.

�� The coefficient of determinations (R2), 0.92 The coefficient of determinations (R2), 0.92 and the Nashand the Nash--SuttcliffeSuttcliffe simulation efficiency simulation efficiency (NS), 0.91(NS), 0.91

0.00E+00

2.00E-02

4.00E-02

6.00E-02

8.00E-02

1.00E-01

1.20E-01

1.40E-01

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59

Months (1987 - 1991)

Sur

face

Run

off,

m3/

s

Simulated-Calibrated Observed

Comparison between observed monthly flow and Simula tion using calibrated parameters for the calibration period (1987 – 1991)

Model Validation

�� Model validation involved reModel validation involved re--running the running the model using input data independent of data model using input data independent of data used in calibration. used in calibration.

�� Two years observed flow data from 01 January Two years observed flow data from 01 January 1992 to 31 December 1993 1992 to 31 December 1993

�� The validation process resulted in R2 =0.96 The validation process resulted in R2 =0.96 NS = 0.93NS = 0.93

Comparison between observed monthly flow and Simulation result

0.00E+00

2.00E-02

4.00E-02

6.00E-02

8.00E-02

1.00E-01

1.20E-01

1.40E-01

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Months (1992 - 1993)

Sur

face

Run

off,

m3/

s

Validated Observed

Scatter plot of monthly simulated versusobserved flow

ConclusionConclusion

�� The model was successfully evaluated through sensitivity The model was successfully evaluated through sensitivity analysis, model calibration, and model validation.analysis, model calibration, and model validation.

�� Curve number, Base flow alpha factor, and groundwater delay Curve number, Base flow alpha factor, and groundwater delay time are the most sensitive parameters in Anjeni watershed.time are the most sensitive parameters in Anjeni watershed.

�� The calibration and validation study has shown that the SWAT The calibration and validation study has shown that the SWAT model can produce reliable estimates of monthly runoff.model can produce reliable estimates of monthly runoff.

AcknowledgmentAcknowledgment

I am grateful toI am grateful to

�� Applied training Project of the Nile Basin Initiative Applied training Project of the Nile Basin Initiative for the financial support of this research project.for the financial support of this research project.

�� University of Bern, Switzerland University of Bern, Switzerland -- SCRP and Dr. SCRP and Dr. GeteGeteZelekeZeleke for providing the necessary data.for providing the necessary data.

�� Dr. Dr. BijanBijan DargahiDargahi for his advice and review of the for his advice and review of the paper.paper.

�� Dr. Dr. Raghavan Srinivasan (Srini) Raghavan Srinivasan (Srini) for his support in for his support in running the ArcSWAT interfacerunning the ArcSWAT interface

Thank you!