Salinity‐Intrusion Forecasting System for Gambia River Estuary

Abstract

A methodology is presented for forecasting the response of salinity movement in a tidal estuary to seasonal rainfall and freshwater inflows. The forecasting procedure uses linked stochastic and deterministic models to provide information to aid decision makers. These models include: (1) Multisite stochastic rainfall data generation models used to generate long‐term synthetic records of 10‐day rainfall for stations in the upper river basin; (2) a deterministic rainfall‐runoff multiple regression model used to compute a long‐term record of 10‐day mean flow on the river's main stem based on real‐time initial flow and rainfall data and synthetic rainfall records; and (3) a one‐dimensional finite difference salinity intrusion model used to compute the movement of the 1 part per thousand (ppt) salinity level for each year of the computed long‐term flow record. A cumulative probability distribution of the maximum salinity flushing distances along the estuary is developed as a tool for decision makers. The Gambia estuary in West Africa was used as a case study.