Title : ( A combination of neural networks and hydrodynamic models for river flow prediction )
Authors: Nigel G. Wright , Mohammad Taghi Dastorani , Peter Goodwin , C.W. Slaughter ,
Full TextAbstract
This research has investigated the application of artificial neural networks (ANN) to improve the accuracy of the results obtained from a hydrodynamic model of river flow. The study area was Reynolds Creek Experimental Watershed in southwest Idaho, which has 239 km2 drainage area and semi-arid climate conditions. Hydrological processes in this catchment are extremely variable because of the high variation in elevation, and consequently climate condition, within the river catchment. After the calibration of a 1D hydrodynamic flood routing model of the main river reach, a MLP neural network model has been adopted to optimise the outputs of the hydrodynamic modelling procedure. Using ANN in this way, the error produced by the hydrodynamic model was predicted and thereby, the results of the model were improved. In addition, the results of a hydrodynamic model affected by the suspension of flow gauging are improved by neural networks. Combination of these two techniques for this specific application uses the potential of both methods and shows a good performance
Keywords
, hydrodynamic models, neural networks, flow prediction
@inproceedings{paperid:1059471,
author = {Nigel G. Wright and Dastorani, Mohammad Taghi and Peter Goodwin and C.W. Slaughter},
title = {A combination of neural networks and hydrodynamic models for river flow prediction},
booktitle = {Fifth International Conference on Hydroinformatics},
year = {2002},
location = {Cardiff, UK, ENGLAND},
keywords = {hydrodynamic models; neural networks; flow prediction},
}
%0 Conference Proceedings
%T A combination of neural networks and hydrodynamic models for river flow prediction
%A Nigel G. Wright
%A Dastorani, Mohammad Taghi
%A Peter Goodwin
%A C.W. Slaughter
%J Fifth International Conference on Hydroinformatics
%D 2002
