Widandi Seotopo
This paper discussed the using of the data series from multiple rainfall gauging stations to predict the annual rainfall of a particular station. The prediction model is a multiple regression using 22 stations. With 22 possible independent variables, then the number of variations of the regression models is quite enormous. Therefore a random sampling technique combined with specific selection methods are employed to produce the predicted annual rainfall values. The best results from these predicted values are then compared to those of several familiar predicting methods. While the results show that the use of multiple series record is not up to the best, but it still has the potential to be investigated further.
Department of Water Resources, Faculty of Engineering, University of Brawijaya, Malang-65145, Jl. M.T. Haryono 167, Indonesia
