Flow Forecasting on the Blue Nile Using Lumped System & Conceptual Models

Abstract

Rainfall-Runoff modeling is essential to provide basic information on those activities related to rivers flow, such as irrigation development, flood control, hydropower generation, hydraulic structures and drinking water supply. Many different types of rainfall-runoff models have been reported in literature review with their mathematical development to understand the physical concepts of water behavior on the ground during hydrological cycle. This study presents the application of the recently developed software, which is known as Galway Flow Forecasting System (GFFS). Several models contained in this software but only seven models namely Parametric and Non-parametric Simple Linear Model, Parametric and Non-parametric Linear Perturbation Model, Nonparametric Linearly Varying Gain Factor Model, Artificial Neural Network Model (system type models) and Soil Moisture Accounting and Routing (SMAR) model (conceptual models) are applied on the Blue Nile catchment at Eldeim (254230 sq. km) during seven years (1990- 1996) of real data of daily rainfall, evaporation and regulated river flow. The results indicate that the Linear Perturbation Model (Parametric and Non-parametric) gave superior performance than others models during the applications. That indicates to the seasonality in the behavior of the Blue Nile catchment. Also the results suggest that the Linear Perturbation Model is a best model to forecast the Blue Nile flow