MODELING OF SEDIMENTATION PROCESS IN ASWAN HIGH DAM RESERVOIR

ABSTRACT MODELING OF SEDIMENTATION PROCESS IN ASWAN HIGH DAM RESERVOIR By Prof. Dr. Mohammed El-Moattassem 1 Assoc. Prof. Dr. Tarek M. Abdel-Aziz Dr. Hossam EL-Sersawy 3 The River Nile is the second longest river in the world, starting at its remotest headstream, the Luvironea River in Burundi, central Africa, and to its delta on the Mediterranean Sea in northeast Egypt. Most sediment carried by water from the watersheds within the Nile basin is received by the river during the flood season, and originates in the Ethiopian Plateau. Due to the construction of Aswan High Dam (AHD), it was estimated that more than 14 million tons per year were deposited annually in the reservoir decreasing its total capacity. Therefore the prediction of deposited sediment in Aswan High Dam Reservoir (AHDR) is essential to estimate the life time of the reservoir. The present paper aims to developing the approaches for simulating water flow and sediment transport in Aswan High Dam Reservoir (AHDR) to predict the problems associated with the sediment deposition in the reservoir especially delta growth. A two dimensional model was used to investigate the long term sedimentation in the Aswan High Dam Reservoir (AHDR). In this study, a general description of the area of study, water flow and sediment modeling approach, development of the computational mesh, calibration and verification of the numerical model, and the model results were introduced. The results showed that the predicted water surface elevations, velocity distributions, and bed elevations were in good agreement with the field measurements. Based on the computed results, the stability of delta deposits in the reservoir will be studied. It was concluded that the proposed modeling approach could be used to simulate sediment transport in a complex system as Aswan High Dam Reservoir. In addition, the results demonstrated that numerical models could be used to simulate the delta progress in the reservoir. KEY WORDS: Nile River Basin, Aswan High Dam Reservoir, Two dimensional models, Sediment Transport. 1 Professor, Nile Research Institute (NRI), National Water Research Center (NWRC),Cairo, Egypt. E-mail: MelQotb@hotmail.com Associate Professor, Nile Research Institute (NRI), National Water Research Center (NWRC),Cairo, Egypt E-mail: Aziztm@hotmail.com 3 Researcher, Nile Research Institute (NRI), National Water Research Center (NWRC),Cairo, Egypt. E-mail: hossam_elsersawy@yahoo.com 1

1. INTRODUCTION 1.1 General The Aswan High Dam Reservoir (AHDR) is the second largest man-made lake in the world, extending from the southern part of Egypt to the northern part of Sudan, about 500 km in length. It extends about 30 km in Egypt and almost 180 km farther upstream in Sudan. The prediction of the sediment deposition in the Aswan High Dam Reservoir (AHDR) is a major concern. Few studies have been conducted using a numerical model to simulate the sediment deposition in Aswan High Dam Reservoir (AHDR). Previous studies to simulate and predict the sediment deposition in Aswan High Dam Reservoir may be divided into two stages; the first stage started before the construction of the Dam till 1985 and the second stage from 1985 till present. Investigators concentrated during the first stage on collecting and analyzing field data to study the characteristics of the reservoir and to deduce relationships between flow and sediment load. While in the second stage they started to develop mathematical models to describe the motion of both water and sediment flow to simulate the water surface and bed profile in the longitudinal direction. Hurst, et al., (1965), estimated the average composition of the suspended matter carried by the flood based on the measured sediment concentrations during the period 199-1955. They concluded that there is no presence of coarse sand, and that 30% by weight is carried as sand fraction, 40% silt, and 30% as clay fraction. Shalash (1980) used the measured suspended sediment concentration during the period from 1968 to 1979 to study the sediment transport along Aswan High Dam Reservoir (AHDR). He concluded that the average annual rate of sediment inflow is 130 million tons, the average annual rate of outflow is 6 million tons, and therefore the average annual deposited sediment is 14 million tons. He used also the available data of the total discharge and sediment passing Kajnarity station during the period 199-1955 to develop a formula relating sediment discharge to water discharge. He estimated the deposited sediment to be 1570 million tons during the 15 years. Shalash also estimated the life time of the dead zone of the reservoir to be approximately 36 years. El-Moattassem and Makary (1988) studied the sediment balance in Aswan High Dam Reservoir (AHDR) during the period from May 1964 to December 1985. They used the sediment and water discharge data at Dongola during the period (1968-1985) to develop a formula that relates sediment load and discharge on daily and yearly values. Using this formula, and Lane-Koelzer formula for density of deposited sediment, they estimated the deposited volume to be 1650 million m 3. The calculated deposited volume from the hydrographic survey for the same period is 1657 million m 3 which is very close to the estimated one. El-Manadely (1991) and Abdel-Aziz (1991) developed a one-dimensional numerical model based on the continuity equation, the momentum equation, and the sediment continuity equation to estimate change in the river bed profile in the longitudinal direction. The model results of the total volume of deposits accumulated inside the reservoir were,650 billion m 3 for the period of operation of AHDR from year 1964 1988. This value is nearly equal to the estimated deposited volume based on field measurements which has a value equal to,760 billion m 3. As a result of these two models, it is concluded that the AHDR cross sections are highly irregular especially in the transverse direction and the change in water depth is large. Therefore, there is a need to develop a new approach based on two-dimensional models in order to predict the sediment deposition in the transverse and longitudinal directions.

1. Description of the Study Area AHDR Dongola G.ST. Aswan High Dam Study Reach Figure 1: Location of Aswan High Dam and Dongola station Aswan High Dam (AHD) is a rock fill dam, closing the Nile River at a distance of 6.5 km upstream of the old Aswan Dam, about 950 km south of Cairo as shown in Figure 1. The dam is 3600 m long and has a width of 40 m at the top and 980 m at the bed level. The maximum height of the dam is 111 m above the river bed. Construction began on the Aswan High Dam in 1960. By 1964 the river was blocked with a coffer dam, and the upstream reservoir began to fill. The construction of the Dam itself was completed in 1970. The construction of AHD upstream of old Aswan Dam, made it possible to have over-year water storage and thus create a reservoir upstream the dam. The length of AHD reservoir is about 500 km at its maximum storage level, which is 18 m a.m.s.l., with an average width of about 1 km and a surface area of 6540 km. The storage capacity of the reservoir has a volume of 16 km 3 divided into three zones: dead storage capacity of 31.6 km 3 between levels 85 m and 147 m, live storage capacity of 90.7 km 3 from level 147 m to 175 m, and flood protection capacity of 39.7 km 3 ranging between levels 175 m and 18 m, the maximum level of the reservoir. The reach analyzed in this study was chosen from km 500 to km 350 upstream Aswan High Dam with a total length of 150 km. Here, we focus at the mean bed channel which represents the area with the most intensive sediment deposition. 3

1.3 Problem Description The Nile River like many other large rivers has two distinct hydrological phenomena: (i) a short, but high flow period (muddy water) of about three months, and (ii) a long, but low (clear water) period of about nine months. Most sediment carried by water from the watersheds around the equatorial lakes is deposited in the lakes and the swamp areas. As a result, the sediment load brought into the river from the White Nile and tributaries is relatively small, accounting for approximately 5 percent of the total annual sediment load of the main river. Meanwhile, 95% of sediment load is received by the river during the flood season, and originates from the Ethiopian Plateau. Due to the construction of the Aswan High Dam in the year 1964, it was estimated that more than 14 million tons per year; Shalash (1980); deposit annually in the reservoir. The sediment deposition is concentrated upstream Aswan High Dam Reservoir (AHDR) as shown in Figure, which describes the longitudinal section of the lowest bed elevation of Aswan High Dam Reservoir (AHDR) from year 1964 to 003. It is observed that the thickness of the deposition layer may be more than 60 meters within the last 40 years at the entrance of the reservoir. Note: Elevation is in (m) a.m.s.l. (above mean sea level) Figure : Longitudinal bed elevation profile for Aswan High Dame Reservoir 1.4 The Main Research Objectives The main objectives of the present study are to (i) develop the methodology for simulating flow of water and sediment in Aswan High Dam Reservoir (AHDR), and (ii) predict the development of the delta growth in the reservoir from year 003 to year 015. 4

Water Level. METHODOLOGY.1 Archived Data for Aswan High Dam Reservoir (AHDR) Hydrographic (bathymetric), hydrologic, hydraulic, and sediment data were collected as input for the water and sediment flow model..1.1 Bathymetric Data Bathymetric data describe the geometry of Aswan High Dam Reservoir (AHDR), were based on data obtained from contour maps, produced from the hydrographic survey of the reservoir provided by the Nile Research institute (NRI, 1973-003). The channel geometry presented by Easting, Northing, and Elevation (E, N, and Z) was used for the mesh generation. The coordinates of the mesh were referenced to the WGS84 ellipsoid with Universal Transverse Mercator (UTM) Projection. In this study, the bathymetric data from the hydrographic survey of year 1999 of the Aswan High Dam Reservoir (AHDR) were used..1. Hydrologic and Hydraulic Data In addition to topographic data, the model requires hydrologic and hydraulic data such as stage and flow hydrographs, stream velocities, and rating curves to establish the initial and boundary conditions. These were split into a calibration and verification dataset: the period from 1999 to 001 were used for calibration and from 001 to 003 were used for validation. In this research, the records of discharges at Dongola station 780 km upstream Aswan High Dam were used as upstream boundary condition for the model. In addition, the water level hydrograph upstream Aswan High Dam was used as the downstream boundary condition for the study area as shown in Figure 3. Water Level Hydrograph at Upstream Aswan High Dam 190.00 180.00 170.00 160.00 150.00 140.00 130.00 10.00 110.00 1964 19651966 1967 19681969 1970 1971197 1973 19741975 1976 19771978 1979 19801981 198 19831984 1985 19861987 1988 19891990 1991 1991993 1994 19951996 1997 19981999 000 00100 003 Years Figure 3: Water level hydrograph upstream Aswan High Dam (1964-003) The daily discharge data at Dongola Station indicated that there are two stages for the Nile River discharges which inflow Aswan High Dam Reservoir (AHDR). The rising stage starts by the end of July and reaches its peak around the middle of September and the falling stage, of which the discharge starts to decrease during months October to June. In general, Dongola discharge records range from 000 to 6000 m3/sec. Also, the same trend of the minimum and maximum observed water levels at the upstream of Aswan High Dam was shown in Figure 3. 5

AUG. SEP. OCT. NOV. DEC. JAN FEB. MAR. APR. MAY JUN. JUL. AUG. SEP. OCT. NOV. DEC. JAN FEB. MAR. APR. MAY JUN. JUL. AUG. SEP. OCT. NOV. DEC. JAN FEB. MAR. APR. MAY JUN. JUL. AUG. SEP. OCT. NOV. DEC. JAN FEB. MAR. APR. MAY JUN. JUL. Discharge (M.m^3/day) Sediment Concentration ppm.1.3 Sediment Data After the construction of Aswan High Dam, Dongola Station has been selected to monitor suspended sediment. The suspended sediment data available for the period 1966-198 were used to establish a rising and falling stage rating curves for the Aswan High Dam Reservoir (AHDR) allowing for seasonal effects relating to the rising and falling stages ((El-Moatasem, et. al., 1988, Walling, et. al 1990). The equations for estimating the suspended sediment hydrograph at Dongola Station are as follows: (i) For rising stage flow discharge hydrograph: Q s = 5.753*10-6 Q 1.98 (ii) For falling stage flow discharge hydrograph: Q s =.695*10-7 Q.347 Where Q is the discharge at Dongola Station in million m 3 /day and Q s is the sediment load in 10 9 kg/day. By applying these equations, the suspended sediment concentration hydrographs at the inlet boundary of the study area of the reservoir (Dongola Station) can be estimated for the period from 1999 to 003 as shown in Figure 4. 000 1800 1600 1400 100 1000 800 600 400 00 0 Discharge and Sediment Concentration at Dongola Station ( 1999-003) Flow Discharge Sediment Concentration 5000 4500 4000 3500 3000 500 000 1500 1000 500 0 1999/000 000/001 001/00 00/003 Months Figure 4: Discharge and sediment concentration hydrographs at Dongola station (1999-003) From Figure 4, it is clear that the concentration of suspend sediment entering Aswan High Dam Reservoir also has a seasonal variation like the flow hydrograph, the peak discharge and peak suspended sediment concentration do not occur simultaneously. The suspended sediment concentration rises to a maximum (5000 ppm) many days before the peak of water discharge. The lag time between the peak of the water discharge and the suspended sediment concentration varies from year to year, and on average is approximately 10 days. 6

. Model Selection and Development..1 Model Selection The following criteria were used to select the most suitable model to be incorporated in the modeling of water flow and sediment transport in the AHDR: (i) the capability to perform hydrodynamic and sediment transport simulations, (ii) public domain software with the source code available, (iii) the ability to be linked to other models and (iv) technical documentation and user support and the record of practical applications. Finally, the TABS-MD models were found to be the most suitable models for simulating the complex reservoir system. The selected model is a continuous simulation model with the capability to include discharge, sediment load, and high precision bathymetry. The particular choice of the TABS-MD model is due to the model s ability to run on a desktop with the Surface-Water Modeling System (SMS) software (BYUEMRL, 00). The SMS software provides valuable tools for mesh generation, data interpolation, and graphical visualization. The SMS program was developed by Brigham Young University (BYU) in cooperation with USACE-WES... The TABS-MD Model Three modules (GFGEN, RMA and SEDD) of the TABS-MD were used in this study. The module GFGEN was used to create the finite element mesh of the study reach; the module RMA simulated the hydrodynamic conditions of the study area; and SEDD was used to compute sediment transport, scour, deposition and bed elevation changes within the study area. The RMA hydrodynamic model solves the depth averaged two-dimensional equations of continuity and momentum transport (Donnell et al., 000). As a result of the simulation, RMA produces water depth and velocity at each time step. Water depths and velocity fields produced by the RMA were then used by the SEDD model. The governing equations for modeling water flow are: For flow in the x direction, ( HU) t uuhuu (cos x cos z) x 1 gh ( uvhuv) y 1 (H ) (H ) xx xy HV bx sx - 0 x y And for flow in the y direction, the vertically integrated mass transport equation (continuity equation) is: H t (HU) x (HV) y q zb cos xgh x ( HV) t HV vv HVV (cos y cos z ) y 1 by sy (H x yx ) (H - y yy 1 ) gh 0 ( uvhvu) x z b cos ygh y 7

( HV) t vv HVV (cos y cos z ) y 1 ( uvhvu) x 1 (H yx) (H yy) HV by sy - 0 x y Where H = water depth; z = vertical direction; z b is the bed elevation; z s = z b + H = water surface elevation; U = horizontal velocity in the x direction; and V = horizontal velocity in the y direction; β uu, β uv, β vu, and β vv = momentum flux correction coefficients that account for the variation of velocity in the vertical direction; α x = arc tan ( z b / x ), α y = arc tan ( z b / y ), α z = arcos (1-cos α x cos α y ); g = gravitational acceleration; Ω = Coriolis parameters; ρ = water mass density, which is considered constant; τ bx and τ by = bed shear stresses acting in the x and y directions, respectively; τ sx and τ sy = surface shear stresses acting in the x and y directions, respectively; and τ xx, τ xy, τ yx, and τ yy = shear stresses caused by turbulence; q = unit source (inflow) or a unit sink (outflow) term, t = time. The SEDD model calculates the suspended sediment concentration using the convectiondiffusion equation presented in Ariathurai et al. (196) as following: gh z b cos ygh y C t C u x C v y D x x C x D y y C 1C y Where C = concentration; t = time; U = flow velocity in x direction; x = primary flow direction; V = flow velocity in y direction; y = direction perpendicular to x; D x = effective diffusion coefficient in x- direction; Dy = effective diffusion coefficient in y- direction; α 1 = a coefficient for the source term; α = the equilibrium concentration portion of the source term. Several options are available in the SEDD model for computing bed shear stresses using τ b = ρ(u*), where; ρ = water density and u* = shear velocity calculated by using either a smooth wall log velocity profile, or the Manning shear stress equation. The sediment load transport determined by using Ackers-White formula was adopted for this model. The following empirical formula was proposed: S t Kbu d 35 u u *, b n Y Y Ycr cr m In which: K e.86ln D 0.434ln D * * 8.13 Y cr 0.3 0.14 D.5 * D 1 g 3 * d35 9.66 m 1.34 D Y u * *, gd b 35 n 0.5 5.66 u 10R log d35 b 1n 8

Where S t = total sediment load transport, u - = mean flow velocity, Δ = relative water density (1.65), b = flow width, υ = kinematic viscosity, u *,b = bed shear velocity, R b = hydraulic radius, d 35 = particle diameter of bed material, g = gravity acceleration..3 Mesh Generation A preliminary finite element mesh was developed, using the SMS 8.0 software (U.S. Department of Transportation 00). A finite element mesh is defined as a network of triangular and quadrilateral elements constructed from nodes. In this study, the Map Module in SMS was used to define the study area boundaries and water features using aerial photographs of year 1991 in scale 1:50,000 and land sat satellite image for AHDR in year 1998). Then, SMS Model automatically generated a mesh or grid network from the map module and then interpolated the bathymetric data into the mesh. The mesh contains 7,738 elements and 3,787 nodes. The mesh of the study reach represents approximately 150 kilometer length, and located approximately 350 kilometer upstream Aswan High Dam (AHD) as shown in Figure 5. The built-in interpolate command in the mesh creator module of SMS was used to assign a depth for each individual node using the bathymetric survey for Aswan High Dam Reservoir (AHDR) at 1999 as shown in Figure 5. Figure 5: The Network with distances in km upstream AHD and bed elevation in (m) of study area in August 1999. 3. MODEL CALIBRATION AND VALIDATION 3.1 Hydrodynamic model calibration and verification Calibration of the hydrodynamic and sediment transport models was performed for the period where most observed data were available. Inflow discharge, water elevations, and velocities of flow were considered to select a time period for the model calibration. The data at the boundary condition of the study reach at the period from 1999 to 001 were used for the 9

calibration of the hydrodynamic (RMA) and sediment transport (SEDD) model. The following parameters were used during the calibration of the hydrodynamic simulation of Aswan High Dam Reservoir (AHDR): material properties, including roughness coefficients (Manning s n) and eddy viscosity values (turbulence exchange factor), for each element in the mesh. The model was calibrated using the available data for 1999, 000, and 001 by using inflow discharges at Dongola Station and water level upstream Aswan High Dam respectively as upstream and downstream boundary conditions as shown in table 1. The comparison of the observed and predicted water surface elevations at Dongola Station shows good agreement as shown in table 1. Table 1: The inflow discharges and water levels for the calibration process The inflow discharges in U.S boundary and water levels in D.S boundary Date Discharge (M.m3/day) Water Level (m) The observed and calculated water levels at U.S boundary Observed Water Level (m) Calculated Water Level (m) Difference July 1, 1999 183.50 175.84 176.5 176.35 0.10 May 1, 000 98.00 178.44 178.50 178.61 0.11 July 6, 001 19.00 175.74 175.97 176.15 0.18 (m) For the validation, the data from 001 to 003 were used. The RMA model was simulated as a dynamic simulation (unsteady state). It takes the time step for the runs as 10 days. The results of these simulations were used later as the input for the sediment model (SEDD). The main flow simulation characteristic during this period such as, the water surface elevations, water depth, velocity distribution and the bed shear stress were calculated as shown in Figure 6. 3. Sediment model calibration and validation The SEDD model used the existing mesh geometry, nodal velocity, and water depth output from the RMA model simulation to calculate bed shear stress and transport/deposition characteristics of the specified streambed. The global sand bed characteristics including sediment particle size, specific gravity of bed material, bed thickness, sand grain roughness, diffusion coefficients, settling velocity, gravity, shear stress equation (Manning s shear stress equation), bed change threshold, and fluid density were used for model calibration and validation. Settling velocity was estimated from the relation between the sphere diameter and settling velocity. Effective diffusion was based on Roig et al. (000). 10

The particle-size characteristics were determined from sediment samples collected and analyzed for the particle-size distribution. The RMA model outputs were used as an input for the SEDD dynamic hydrograph simulations. However, to run SEDD for dynamic simulation RMA has to be run first for steady state simulation. Therefore, RMA was run for 4 to 5 times to allow for sediment transport to occur along the whole reach of study area and to get stable condition. Applying this method, sediment concentrations and bed shear stress will be more stable during the initial phase of SEDD simulation. Calibration Process Figure 6: Velocity magnitude in (m/s) and water surface elevation in (m) a.m.s.l. Figure 7: Suspend sediment concentration in ppm (mg/l) and bed changes during the period of flood (August- October) 1999. Of flood (Aug-Oct. 1999) 11

LEVEL(M) LEVEL(M) LEVEL (M) Comparison of the Observed and Modeled X.S No. (10) at Year 003 Km (415.5) U.S HAD 00 190 180 170 160 150 Observed Model 0 100 00 300 400 500 600 700 800 900 1000 1100 100 DISTANCE (M) Comparison of the Observed and Modeled X.S No. (3) at Year 003 Km (378.5) U.S HAD 00 190 180 170 160 150 Observed Model 0 100 00 300 400 500 600 700 800 900 1000 1100 100 DISTANCE (M) Comparison of the Observed and Modeled X.S No. (D) at Year 003 Km (37) U.S HAD 00 190 180 170 160 Observed Model 150 0 100 00 300 400 500 600 700 800 900 1000 1100 100 1300 1400 1500 1600 DISTANCE (M) Figure 8: Comparison of observed and predicted cross sections in AHDR for year (003) in Validation Process The calibration of the sediment transport model for the Aswan High Dam Reservoir was based on suspended sediment concentrations during yearly fieldtrips and by comparison of simulated bed elevations along chosen cross sections with the measured ones. The period of the calibration was chosen from 1999 to 001 based on the availability of data. Figures 7 describes the suspend sediment concentrations distribution and the bed elevations changes within the area of study at the maximum flood event from August to November for the year 1999. The changes in bottom elevation generated by the SEDD model at the end of each time step were calculated for nodes located on the selected cross sections. Predicted cross-sections were compared with observed ones. Figure 8 shows that there is reasonable agreement between field and simulated data. This indicates that the model is well calibrated, and provides confidence that the model can reproduce concentration for real-time simulations on a time scale of days or weeks and years. 1

Elevations (m) Elevations (m) Elevations (m) Elevations (m) The adopted hydrodynamic (RMA) and sediment transport (SEDD) model parameters were used for the simulation of the sediment deposition in Aswan High Dam Reservoir for the period 001 to 003 for validation of the model results. The water level stages and the velocities calculated by the hydrodynamic model RMA were used as input to the sediment transport model. At the end of each SEDD simulation, a new bathymetric configuration was produced. This new bathymetry was imposed on the hydrodynamic model for the next cycle. Repetition of sequence was repeated until the ends of simulations were made. Predicted cross sections were compared to those observed in year 003 as shown in Figure 8; this comparison shows reasonable agreement with field data in the area of the study. 3.3 Model Simulation and Prediction 175.00 Comparison between Longtudinal Section for the High Aswan Dam Reservoir (HADR) at 001 Flow Direction 175.00 Comparison between Longtudinal Section for the High Aswan Dam Reservoir (HADR) at 003 Flow Direction 170.00 170.00 165.00 165.00 160.00 160.00 155.00 155.00 001- Measured 003- Measured 150.00 150.00 001- Model 003- Model 145.00 500 480 460 440 40 400 Distance from High Dam (km) 380 360 Figure 9: Comparison of measured and predicted longitudinal profile for AHDR 001 and 003. The models were applied from 1999 to 003 to predict new bathymetric data for the bed of Aswan High Dam Reservoir AHDR. The longitudinal section of the sedimentation deposition along the lowest point in the bed elevations of the Aswan High Dam Reservoir was generated for 001 and 003 as shown in Figure 9. In addition, the models were used for future prediction of sediment deposition to predict the delta growth in AHDR. A time series of seven consequent high flood years (004 to 010) and five low flood years (011 to 015) was simulated. This is similar to the historical records of AHDR during the period 197-1983 (Abdel-Aziz 1997). The longitudinal bed profile of AHDR was predicted for different years and shown in Figure 10. 340 145.00 500 480 460 440 40 400 Distance from High Dam (km) 380 360 340 175.00 Longtudinal Section for the High Aswan Dam Reservoir (HADR) From 003 to 010 Flow Direction 175.00 Longtudinal Sections for the High Aswan Dam Reservoir (HADR) From 011 to 015 Flow Direction 170.00 170.00 165.00 165.00 160.00 160.00 155.00 155.00 003 005 007 010 011 01 013 014 015 150.00 150.00 500 480 460 440 40 400 380 360 340 500 480 460 440 40 400 380 360 340 Distance from High Dam (km) Distance from High Dam (km) Figure 10: Prediction of the Delta Growth in AHDR (003-010) and (011-015) 13

4. DISCUSSION The comparison between the observed and modeled cross-sections (Figure 8) indicates that there is a good agreement between the modeled and the measured cross-sections, although some slight differences are observed which are probably linked to that the deposited sediment distribution in the transverse direction is similar to the velocity distribution. In order to increase the accuracy of simulation, the locations of measuring velocity should increase in the transverse direction. From the comparison of the modeled and measured longitudinal section of AHDR in 001 and 003; it was noted that the modeled bed level is higher than the measured one in the whole inlet zone of the reservoir (i.e. from km 500 to km 400 upstream the dam) although there is a good agreement between modeled and measured longitudinal profile in the rest of the reservoir (Figure 9). This is may be explained by the fact that part of the sediment is probably trapped in the Sudanese reservoirs before entering AHDR, which is not included in the model. For the prediction of delta progress in the dam direction until year 010, a time series of seven successive years of high flood were simulated (Figure 10). The prediction indicates that the bed level will rise along the whole reservoir by a value ranging between 3.5 and 1.5 m in the inlet zone until km 370 upstream the dam, and by a value ranging between 1.5 and 0 m in the rest of the length. These seven years of flood are predicted to be followed by five successive years of low flood. The model predicts a raise of the bed level by.0 to 1.0 m in the inlet zone until km 370 upstream of the dam, and by 1.0 to 0 m in the rest of the length. 5. CONCLUSIONS The present study discusses the sediment deposition across the Aswan Dam High Reservoir, and shows how two-dimensional hydrodynamic and sediment transport models were used to predict sediment deposition and delta migration. The dataset consisting of four years (i.e. 1999 till 003) of measurements of water discharge and sediment deposition was split into a calibration (i.e. 1999-001) and validation (001-003) dataset for the application of the SMS model. The comparison of the modeled and measured cross-sectional data indicated that the sediment deposition can be predicted within an error range of 15% for the calibration data and by 10% for the validation dataset. A scenario was developed to predict the delta migration in the ADHR for the coming 1 years. A stationary scenario was designed, which predicts future water discharge and sediment load to be similar to the past 1 years of measurements at the entrance of ADHR. Accordingly, a time series consisting of seven flood years (003-010) followed by five low flood years (011 015) was simulated. For the seven high flood years, the whole bed level will raise by approximately 3.5 to 1.5 m in the inlet zone (i.e. from km 500 till km 370 upstream AHD), and by 1.5 to 0 m in the rest of AHDR. During the successive five low flood years, the bed level is predicted to increase by a value between.5 and.0 m at the inlet zone and by.0 to 0 m in the rest of AHDR. The modeling approach has proven to be a useful tool to monitor future water flow and sediment deposition in the reservoir, which is of uttermost importance for the operational plan and the maintenance of the AHDR. Different scenarios can be developed based on future climate predictions for the region, and/or based on changes in sediment production and sediment delivery to the main stream. It is clear that changes in watershed and stream management upstream will have a profound impact on the discharge and the sediment load 14

entering AHDR. Watershed models, which link the sediment production and delivery in the upstream catchments to the sediment transport and deposition in the river channels and reservoirs, will allow us to predict the future behavior of AHDR. ACKNOWLEDGEMENTS The authors appreciate deeply the financial support of FRIEND /Nile Project through Sediment Transport and Watershed Management Component (STWMC). They are of great thankful for the support of the UNESCO Cairo office staff members. The financial and technical supports of the Flemish government are highly appreciated. REFERENCES Abbott, M. B. 199. Computational Hydraulics. Ashgate Publishing Company, Brookfield, Vermont 05036, USA. Barrett, K. R. 1996. Two-dimensional Modeling of Flow and Transport in Treatment Wetlands: Development and Testing of a New Method for Wetland Design and Analysis, Ph.D. Dissertation, Northwestern University, 1996. 53 pp. B. P. Donnell, Joseph V. Letter, Jr., L. C. Roig,,W. A. Thomas, W. H. McAnally, and S. A. Adamec, Jr.,000, User Manual of SEDD, McLean, Virginia, U.S.A. Chanson, H. 1999. The Hydraulics of Open Channel Flow: An Introduction. John Wiley and Sons Inc., New York. DeVries, M., Klaassen, G. J., and Struiksma, N. 1989. On the Use of Movable Bed Models for River Problems: State of Art, Symp. River Sedimentation, Beijing, China. Donnell, B. P., Letter, J. V., Jr., and Teeter, A. M. 1991. The Atchafalaya River Delta; Report 11, Two Dimensional Modeling, Technical Report HL-8-15, US Army Engineer Waterways Experiment Station, Vicksburg, MS. El-Manadely, M.S., 1991, Simulation of sediment transport in the Aswan High Dam Lake, Ph.D. Dissertation, Irrigation and Hydraulic department, Cairo University, Egypt. EL-Moattassem, M, and Abdel-Aziz, T.M., 1988, A Study of the Characteristics of Sediment Transport in Aswan High Dam Reservoir, Report No. 117, Cairo, Egypt. Finite Element Surface Modeling System, Two Dimensional Flow in a Horizontal Plane. User s Manual, Brigham young university environmental modeling Research laboratory, 00, McLean, Virginia, U.S.A. NRI, 1973-003, Annual report of Sedimentation in Aswan high Dam Reservoir, Nile Research Institute report, National water Research Center, Cairo, Egypt. Shalash, S., 1980, Effect of Sedimentation on Storage Capacity of Aswan high Dam Reservoir, Nile Research Institute Report, National water Research Center, Cairo, Egypt. 15