|
Titel |
Applying a Genetic Programming Approach to River Flow Forecasting |
VerfasserIn |
A. Y. Shamseldin, R. J. Abrahart, L. M. See, C. W. Dawson |
Konferenz |
EGU General Assembly 2009
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 11 (2009) |
Datensatznummer |
250024491
|
|
|
|
Zusammenfassung |
This paper explores the use of Genetic Programming (GP) for river flow forecasting in two contrasting catchments in different parts of the world. GP is a data-driven non-parametric approach that is powerful in modelling input-output relationships without making any a priori assumptions about the relationship. GP is based on the principles of evolution and essentially allows equations to be bred from a series of input variables and operators. The two contrasting catchments and their points of forecast are: (i) the River Brosna, at Ferbane, in the Republic of Ireland; and (ii) the Upper Han River, at Shiquan, in east-central China. The daily rainfall and the previous observed discharge are used as inputs to the GP model to produce the river flow forecasts for a 1-day lead time. The results of the GP model are compared to those of a traditional linear transfer model. The results indicate that the GP model has considerable promise in river flow forecasting. |
|
|
|
|
|