Abstract
Abstract A simple method is shown to reduce the problems caused by the mismatch in response times of Neil Brown CTD sensors. Assuming that the response time of the platinum resistance thermometer (PRT) is much longer than that of the conductivity cell, the only relevant parameter is the PRT response time. Thermodynamic variables computed with raw CTD data show two undesired effects: a bias or offset when the CTD is moving throughout constant gradient regions, and spurious variations (spikes) when moving throughout regions of variable gradients. Both effects are directly related to the phase shift and attenuation of the PRT signal and can be reduced significantly although with the loss of some vertical resolution. The method consists of attenuating the PRT signal with the digital version of a single-pole filter but backward in time to recuperate phase, and the conductivity and pressure signals are filtered forward as well as backward in time with the same filter. The aim is to produce equally smoothed temperature and conductivity series with null phase shifts with respect to the in situ series. The salinity difference between the descending and ascending T-S relationship is a variable whose variance and mean are used to estimate the appropriate time constant of the filter. This procedure produces profiles with a similar vertical resolution than other correcting methods with a final 2 db blocking average. Since it does not use the fast response thermistor signal, it is particularly useful for the simpler Neil Brown Smart CTD.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call