Abstract
n this study, potential impacts of ENSO and Indian Ocean SSTs indices on variability of inflow to the Roseires Dam are investigated. The objective is to investigate the relationship between these climatic indices and the river flow to enable building a forecasting tool to give a lead time prediction of inflow to the dam using lead time information of these indices. Ninety one years of annual flow (1914-2004) were used in the analysis. Correlation results showed that the annual inflow to the dam has negative relation with the ENSO-SST index and SST indices from two regions in the Indian Ocean. Two probabilistic prediction models were developed using the concept of conditional probability to forecast the inflow to the dam. Modelling results showed significant improvements in predictability of inflow when these indices are used as predictors. The models developed could contribute to better water management and operation of the dam reservoir.