Abstract: During the long working time in the sea, the Argo profiling float’s conductivity sensor may drift, which leads to the measurement error and affects the quality of the measure data. This paper utilizes the error theory and symbolic statistics method including regression analysis and curve fitting, to detect and solve the conductivity sensor drift problem of Argo profiling float. This paper also calibrates the data of a number of Argo profiling floats from the Pacific , Atlantic and Indian ocean, which has the conductivity sensor drift problem, and obtains the satisfying result.
Key words: Argo profiling float; Conductivity sensor; Drift; Calibration