Objective analyses of hydrographic data for referencing profiling float salinities in highly variable environments

The development of a broad-scale array of about 3000 autonomous profiling floats, known as Argo, has been underway since 2000. This array will deliver up to 100,000 vertical profiles of temperature, salinity and other parameters from the surface to depths up to 2000 m. While floats are expected to give good measurements of temperature and pressure, salinity measurements sometimes show significant sensor drift with time or offsets. Unless a float is recovered before the battery fails, recalibrations cannot be performed and a remote calibration method is required. Such a quality control system has been set up for the North Atlantic to identify and correct salinity sensor drifts by using historical hydrographic data. An objective mapping method is used that takes the spatial and temporal variations in water mass properties into account. These scales aim to represent the hydrographic structure of the North Atlantic, which follow the large-scale contours of the potential vorticity. The float measurements of each profile are compared to the mapped salinities in potential conductivity space by weighted least-squares, giving one correction for each profile. It is assumed that any conductivity offset changes slowly over time, so that a linear fit of the profile based corrections over the float time series is done. The result is a set of calibrated salinity data with corresponding uncertainties.