Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Prediction of sea surface temperature in the Gulf of Thailand using neural network approach

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  Paper number  032003,  2003 ASAE Annual Meeting . (doi: 10.13031/2013.14935) @2003
Authors:   Ramesh Kumar Gautam, So Kazama, Suranjan Panigrahi
Keywords:   Gulf of Thailand, sea surface temperature, back propagation, radial basis function, modular neural networks

Back propagation, modular and radial basis function neural networks were developed to predict the variation of sea surface temperature at different buoy stations in the Gulf of Thailand. Each model was trained and tested with the same number of datasets.

It was found that both modular and radial basis function neural networks outperformed the back propagation model in all the buoy stations. The prediction of radial basis function neural network was highest at Sichang and Rayong buoy stations with average prediction accuracy and correlation coefficient being 98.86% and 0.82, respectively. The performance of modular neural network was highest in Platong buoy station. Both modular and radial basis function neural networks showed similar prediction accuracies and correlation at Kochang buoy station. The average prediction accuracy and correlation coefficient at Kochang buoy station from modular neural network was 98.48% and 0.85, respectively. The corresponding values from RBFN were 98.54% and 0.84, respectively.

(Download PDF)    (Export to EndNotes)