What Observation Scheme Should We Use for Profiling Floats to Achieve the Argo Goal for Salinity Measurement Accuracy? Suggestions from Software Calibration

Abstract

To meet the Argo Project?s accuracy goal for salinity measurements (±0.01), several measurement schemes for profiling float observations were examined using the standard Argo software calibration for salinity. In this study, salinity calibration errors are evaluated in data series observed by employing several different measurement schemes. In the North Pacific the accuracy goal can be achieved by programming for constant measurements to a depth of 1500 db or more; however, measurements in the mixed water region require the adoption of a scheme that constantly measures salinity profiles to the marginal capability depth of floats (e.g., 2000 db), so that calibration errors can be minimized. In contrast, in the tropical regions of the Pacific and eastern Indian Ocean, 1000-db profiles provide salinity data of sufficient quality. When floats use a Park and Profile scheme, it is recommended that they be programmed to measure deeper profiles every observation cycle. Otherwise, it is impossible to detect and correct suspicious data, such as a salinity jump in which a salinity measurement suddenly differs from previous measurements by ±0.02?0.03 or more.