Fast Estimation of Full Depth Sound Speed Profile Based on Partial Prior Information

The distribution of underwater sound speed profile (SSP) is of great significance for the design of underwater acoustic communication and positioning systems. SSPs can be accurately measured by conductivity–temperature–depth profiler (CTD) or sound velocity profiler (SVP), but it spends a long time. SSPs can also be estimated by inversion methods, such as matching process, artificial neural networks, but the inversion result is not as accurate as CTD or SVP. With the development of expendable conductivity–temperature–depth profiler (XCTD), the SSP can be quickly and accurately measured with the help of autonomous vehicles (AUV), however, the measured SSP could not cover the full ocean depth. In order to obtain a full–depth SSP based on prior information obtained by XCTD, we propose a fast SSP estimating method, called empirical–orthogonal–function–based partial sound speed profile matching extension (EOF–PSSP–ME). The feature coefficients are obtained through projecting prior information on partial–depth feature vectors, then the full–depth SSP is constructed by combining the feature coefficients with full–depth feature vectors. Results shows that the full–depth SSP can be estimated accurately based on partial prior information.