A neural network model for predicting the bulk-skin temperature difference at the sea surface

Publisher Website
Google Scholar
Abstract
Night-time radiometric sea surface temperature (SST) observations were carried out on a research platform in the North Sea during the second campaign of the ASGAMAGE experiment. An extensive series of atmospheric measurements was also made, allowing a comparison between measurements of the bulk-skin temperature difference, Delta T, and several current theoretical models. An artificial neural network (ANN) was empirically designed and trained on a subset of the net heat flux and wind speed parameters. The remaining dataset was then applied to the output of the ANN. The neural network-based model reproduced the observed Delta T values with a higher level of accuracy than any of the other current models.

This publication has 0 references indexed in Scilit: