Numerical simulation of bed level changes of the ganges

Abstract

Simulation of sediment transport rate at the river Ganges and variation of bed level along the river is carried out by using a one dimensional morphological model. Noncohesive sediment transport module of MIKEll is used for the simulation. The upstream boundary of the model is taken at 97 Ian upstream of Mathabhanga offtake (Ganges 0) and downstream boundary is taken at Padma 100 Ian. The river system is schematized by five branches, one inflow boundary at upstream, one lateral inflow and four lateral outflow boundaries and one outflow boundary at downstream (water level boundary). In order to make convenient for sediment calculation with hydrodynamic computation, the river is represented by equivalent cross-sectional shape and longitudinal profile. This is achieved by considering that the hydraulic properties of the representative cross-section is the same as that of the actual cross section. Simulation period is taken from April 1985 to October 1991. Simulation is carried out for hydrodynamic calibration and for transport rates. The cross-section interval varies from 2 Ian to 10 Ian in the schematization and the total length including the Padma river is 314 Ian. The time interval used for the morphological computation is four hours for hydrodynamic computation and four days for sediment transport computation respectively. Both for hydrodynamic and sediment transport, the results are saved after every four days. Calibration is carried out against field observations (both water level and discharge) from 1985 to 1991. Water level comparison for simulated and observed values are taken at five key locations (Hardinge Bridge, Sengram, Mohendrapur, Baroria and Mawa) whereas discharge is compared at Baroria only. The results showed satisfactory agreement with observed values. For morphological computation, a time series data of sediment flow is used at the upstream boundary. This time series is generated after a logarithmic transfomation of observed transport rate and observed water discharges measured by BWDB. With this generated transport rates at the boundary, the sediment transport calibration is performed by comparing simulated rates. The results show very good agreement with the observed rates of sediment transport. The simulated sediment load is again verified with the loads at Hardinge Bridge as computed by various investigators. After calibration of the model for both hydrodynamic and sediment transport, the net amount of erosion and siltation along the river reach is computed. Finally, longitudinal variation of bottom level for the monsoon season from 1988 to 1991 is drawn from the model results.