Hierarchical Construction of Sound Speed Profile based on Argo and Sea Surface Data

Abstract

Sound speed profile is an important content of hydroacoustics research, real-time acquisition of large-scale, high-precision sound speed profile is of great significance to ocean surveys. This paper studies the Indian Ocean from 2 degrees to 24 degrees north latitude and 50 degrees to 77 degrees east longitude. The data of the Argo sound speed profile; the sea surface temperature SST (Sea Surface Temperature) data and the sea level anomaly SLA (Sea Level Anomaly) data from September 2013 to 2015 are selected. Through correlation analysis, it is found that the sea surface data is feasible to reconstruct the sound speed profile and has a high degree of correlation with the upper layer (mixed layer and thermocline); therefore, through the establishment of a layered model, the upper layer adopts the sEOF-r method. The first two-order coefficients and sea surface data are used to construct a local temperature profile regression database; the lower layer uses the Empirical Orthogonal Function (EOF) method to obtain the full-sea depth temperature profile. Combined with the stable temperature-salt relationship and the empirical formula of sound speed, the sound speed profile is calculated. The analysis shows that the overall root mean square error of the upper layer of the temperature profile constructed by layers is 1.6234, and the overall root mean square error of the lower layer is 0.5439. The overall reconstruction result is accurate and conforms to the characteristics of the water area.