A Neutral Network Approach to Rainfall Forecasting in urban environments
- Publisher:
- AA Balkema Publishers
- Publication Type:
- Chapter
- Citation:
- Neural networks for hydrological modelling, 2004, 1, pp. 177 - 195
- Issue Date:
- 2004-01
Closed Access
Filename | Description | Size | |||
---|---|---|---|---|---|
2006008644OK.pdf | 2.18 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
An effective flood warning system in urban areas must provide the warnign with sufficient lead time for an appropriate response by the relevant emergency services and the affected community. The requirement poses a critical problem as most urban catchments are characterised by a fast hydrologic response to storm events. The approach used here to forecast rainfall over the Upper Paramatta River Catchment in Sydney is based on the application of a pattern recognition technique using an artificial neural network. It assumes that the future rainfall is a function of a discrete number of past spatial and temporal rainfall records; an important task, therefore, is the determination of the number of spatial and temporal rainfall records necessary for accurate prediction of future rainfall. The rainfall prediction model performed best when an optimal amount of spatial and temporal rainfall information was provided to the network.
Please use this identifier to cite or link to this item: