Glider and profiling float CTD dynamic performance and correction algorithms through a strong thermocline

Mark Halverson, Jean-Michel Leconte and Greg Johnson, RBR Ltd, Ottawa, ON, Canada
All sensors on a CTD have a transient response to variations in temperature and salinity, but the response depends on the details of the CTD design. Therefore, the dynamic response of new CTD designs must be evaluated in order to achieve high quality data.

In particularly strong temperature gradients, the errors caused by the transient response in conductivity and temperature can exceed the manufacturer's calibration accuracy. Dynamic errors in conductivity and temperature are usually small relative to natural variations. However, they sometimes compound into large relative errors in derived variables such as salinity because, in many regions of the ocean, salinity varies to a lesser extent than conductivity and temperature. Dynamic errors often result in salinity spikes creating signatures of density instability in profiles, falsely implying that there was active mixing.

Although the physics underlying the dynamic performance is common amongst all CTDs, the relative importance of various corrections depends on the CTD geometry and sensor layout. There is a rich literature detailing different methods to correct for dynamic errors for various CTD models, however, they do not deal with the specific case of an RBR CTD with an inductive conductivity cell. In this presentation, we present results from a series of experiments conducted in a laboratory tank designed to support a very sharp and thin thermocline. Two new RBR CTDs were tested: 1) an RBRargo3 CTD, and 2) an RBRlegato3 glider CTD. At low profiling rates (e.g., Argo floats), dynamic errors are dominated by thermistor inertia and conductivity cell thermal inertia, while at high profile rates, dynamic errors are dominated by thermistor inertia.