Reconstructing Sound speed profiles worldwide with Sea surface data

Sound Speed Profile (SSP) is the key factor affecting underwater acoustics and it is of great value to obtain SSP in near real-time. In this paper, the sea surface data were used to reconstruct the SSP with the single empirical orthogonal function regression (sEOF-r) method in a global scale. Argo floats data as well as the altimeter data and sea surface temperature (SST) in the year 2010-2013 were used to establish the regression dataset. Argo profiles worldwide were grouped into 2°×2° longitude/latitude grid cells. Then EOF vectors were obtained in each grid, and the regression coefficients for the vectors were obtained with the sea surface data. Analysis showed that SSP anomalies differ from place to place. An assumption was made that the difference was due to the dynamic eddy activity and the eddy kinetic energy (EKE) map was pictured. Results suggested that the two variables correlated with each other. The larger the EKE, the larger the SSP anomalies. However, compared to the absolute value of the SSP anomalies, the error estimation improvement ratio remained relatively stable in most places.