Limitations of hydrodynamical models with limited data available case study: Sondu river basin (Kenya)

Abstract

The Surface Water Modeling System (SMS) model has been incorporated in the Sediment Transport and Watershed Management Component of the FRIEND/NILE project as a research tool for presenting practical solutions for sediment transport problems. In order to predict sediment transport and deposition in selected river sections within the Nile river basin, the SMS model is being tested for its suitability for hydrodynamic modeling in rivers and reservoirs. The paper therefore discusses the use of SMS to predict water depth, stream velocity, sediment loadings and bed elevation changes in the Sondu Miriu River in Kenya. In addition the paper also discusses the problems experienced with SMS in modeling meandering rivers with low volumes of discharge as well as limited cross-sectional data regarding the river channel width, and water level data corresponding with the discharges at specified times, and discusses how the limited data availability affects the quality of the results obtained by SMS. The paper discusses the possible measures that were taken to resolve the problem of data gaps. Comparing the results from the model run with field data provided some validation; this shows that the dryness of the channel that is predicted by the model is also observed in the field. The paper also gives recommendations on the workability of the SMS model on various rivers.