PREDICTION OF ESTUARINE INSTABILITIES WITH ARTIFICIAL NEURAL NETWORKS

This paper shows how a feed-forward back-propagation type of neural network can be used to predict the flow conditions when interfacial mixing in stratified estuaries and fjords commences. This was achieved by training the network to extrapolate data from laboratory experiments on stratified saline flows up to full-size conditions. Because little data is available from real estuaries, and what theory exists is not precise, the neural network was first tested on laboratory and field data for stratified thermal flows. This type of stratified flow is formed in rivers downstream of a power plant's heated water outlet. Here, the stability of the interface is precisely known and this condition can be used to test the accuracy to which neural networks can extrapolate data. The extrapolated laboratory data for stratified saline flows gave results which compared favorably with field data from three fjords and with the stability conditions calculated from turbulent flow theory using laboratory results for interfacial friction factors.