An exploration of artificial neural network rainfall-runoff forecasting combined with wavelet decomposition. (1st April 2004)

• Record Type: Journal Article
• Authors: Anctil, François; Tape, Doha Guy
• Is Part Of: Journal of environmental engineering and science. Volume 3(2004)Supplement 1
• Page Start: S121
• Page End: S128
• Publication Date: 2004-04-01
• Subjects: environment; Environmental engineering -- Periodicals; Sanitary engineering -- Periodicals; Environmental engineering -- Canada -- Periodicals; Environnement, Technique de l' -- Périodiques; Technique sanitaire -- Périodiques; Environnement, Technique de l' -- Canada -- Périodiques; Environmental engineering; Sanitary engineering; Canada; Electronic journals; Computer network resources; Periodicals; 628.05
• DOI: 10.1139/s03-071
• Languages: English
• ISSNs: 1496-2551

Abstract

Abstract : This study compares the one-day-ahead stream flow forecasting performance of multiple-layer artificial neurons and a neuro-wavelet hybrid system at two sites. Morlet power spectra are used to identify the period-scale structure of the available rainfall and runoff time series. The time series are wavelet decomposed into three sub-series depicting the rainfall-runoff processes: short, intermediate, and long wavelet periods. Then, multiple-layer artificial neurons are trained for each wavelet sub-series. Results show that the short wavelet periods are responsible for most of the final neuro-wavelet hybrid forecasting error. Short period fluctuations are thus the key to any further improvements in artificial neural network (ANN) rainfall-runoff forecasting models. The final performance of the neuro-wavelet hybrid forecasting system and of the classic forecasting multiple-layer artificial neuron system is very similar. The slight advantage in performance of the neuro-wavelet system may be attributed to a better usage of the evapotranspiration time series. Key words : surface-water hydrology, rainfall-runoff, artificial neural networks, wavelet decomposition.