Forecasting stream temperature using adaptive neuron- Fuzzy logic and artificial neural network models

All biological processes in water are temperature dependent. The plunging depth of stream water and its associated pollutant load into a lake/reservoir depend on stream water temperature. Lack of detailed datasets and knowledge on physical processes of the stream system limits the use of a phenomenological model to estimate stream temperature. Rather, empirical models have been used as viable alternatives. In this study, models using artificial neural networks were examined to forecast the stream water temperature from available solar radiation and air temperature data. Observed time series data were non-linear and non-Gaussian, thus the method of time delay was applied to form the new dataset that closely represented the inherent system dynamics. Mutual information function indicates that optimum time lag was approximately 3 days. The four-layer back propagation neural network (4BPNN) optimized by micro-genetic algorithms showed that the prediction performance was optimum when data are presented to the model with one-day and three-day time lag, respectively. Air temperature was found to be the most important variable in stream temperature forecasting; however, the prediction performance efficiency was somewhat higher if short wave radiation was included.