Mapping sea surface observations to spectra of underwater ambient noise through self-organizing map method.

Abstract

This letter presents a model for mapping sea surface observations to the spectra of underwater ambient noise through the self-organizing map method (SOM). The data used to train and test the proposed model include observations of wind speed, atmospheric pressure, and significant wave height from public databases, as well as observations of ambient noise from two deep-water experiments. SOM extracts nonlinear relations from the data and is more suitable for the study of nonlinear dynamics in the ocean than conventional methods. Results indicate the proposed model is reliable with coefficients of determination above 0.9 and root-mean-square errors below 1 dB.