![]() | Only 14 pages are availabe for public view |
Abstract the Monitoring, Forecasting and Simulation project (MFS) was initiated in 1990 by the Minis. J of Public Works and Water Resources (MPWWR) for the purpose of providing planners and decision makers in Egypt with (a) Timely forecasts ofthe Nile River inflows into the High Aswan Darn (HAD) reservoir, (b) Real-time information about hydrological and meteorological processes occurring in the whole Nile Basin, and (c) Tools to simulate flow regime of the Nile and assess the possible consequences of changes, man-induced or natural, in the Basin. The simulation component of the MFS system is a distributed model with the sub-basin represented as connected grid points. Connectivity of the grid points was determined by a software which uses Geographic Information System (GIS) based on streams network definition and elevations. TI,e runoff simulation model has 3 parts: a water balance model which produces surface subsurface runoff; a hiUslope routing model to route surface runoff within the cell to the stream, and a channel routing model for moving water from grid cell to grid cell (i.e. the stream channel network). The channel routing model has 5 parameters: the channel friction slope (s), the Manning’s roughness coefficient (n), the channel length (L), and the parameters (a and b) which depend on channel shape. The study presents the folJowing:(a) an approach to estimate parameters a and b at Khartoum location using available data for this cross section. Then, the GIS (geographic information system) is employed to set the value of parameter ”a” at each grid cell as a function of the catchment area of each sub-basin in order to have cross section for the channel close to the real cross section, (b) a sensitivity analysis for the channel roughness coefficient at Diem station with the new cross section parameters, (c) parameter optimization for the Manning’s roughness coefficient(n) based on the new parameters of the cross section (a.b), (d) a comparison of MFS simulation with new simulation. The Extended Stream flow Predictions (ESP) is employed to simulate the future based on historical rainfall patterns and current watershed conditions to produce forecasting. Besides, the study compares the MFS forecasting with a simpler statistical model to check and validate the behavior of tile forecasting system. TI,e study shows that tbe distributed parameters for the channel cross section gives better simulation results than the lumped parameter. It shows also that the best results can be achieved with the case of triangular section. The sensitivity analysis shows that the simulated hydrograph for the case of triangular section is mort; sensitive to the changes in the roughness parameter than the case of rectangular section. The comparison between the ESP and the regression model shows that the ESP forecasts are more reliable than the regression model forecasts. More specifically, the comparison results indicate that the ESP is less biased and include much less uncertainties. In June and July, the ESP forecast is overestimated indicated by positive mean error. This could be due to an error in the initial state variable of the system or due to an error in the historical rainfall used as an input to the ESP for these particular months. It recommended to consider both the ESP forecasts and the regression model forecasts for June and July. |