A univariate model for long-term streamflow forecasting
Academic Article
Overview
Identity
Additional Document Info
Other
View All
Overview
abstract
This paper, the second in the series, verifies the entropy-based univariate model developed in the first paper for long-term streamflow forecasting on five rivers from different regions of the world. The results of the model are compared with the corresponding results of ARIMA and state-space model. The Lagrange multipliers of the univariate model are found similar to autocorrelation coefficients of the ARIMA model. Forecasts by ARIMA and univariate models were comparable for periodic streamflow, but for forecasting of highly variable streamflows the univariate model was superior. 1991 Springer-Verlag.