Inferring density from temperature via a density-ratio relation

A method is presented for determining salinity and density from temperature data in conjunction with historical or contemporaneous (but not collocated) CTD observations. The horizontal density ratio r(z) is determined from the temperature and salinity differences at each depth (δT, δS) between pairs...

Full description

Bibliographic Details
Main Authors: Käse, Rolf H., Hinrichsen, Hans-Harald, Sanford, T. B.
Format: Article in Journal/Newspaper
Language:English
Published: AMS (American Meteorological Society) 1996
Subjects:
Online Access:https://oceanrep.geomar.de/id/eprint/4156/
https://oceanrep.geomar.de/id/eprint/4156/1/1520-0426%281996%29013_1202_idftva_2.0.co%3B2.pdf
https://doi.org/10.1175/1520-0426(1996)013<1202:IDFTVA>2.0.CO;2
Description
Summary:A method is presented for determining salinity and density from temperature data in conjunction with historical or contemporaneous (but not collocated) CTD observations. The horizontal density ratio r(z) is determined from the temperature and salinity differences at each depth (δT, δS) between pairs or ensembles of profiles. These differences are expressed as a density ratio r=αδT/βδS, where α and β are the expansion coefficients for temperature and salinity, respectively. Salinity at a site where only temperature is measured, as with an expendable bathythermograph (XBT), is computed based on the temperature and salinity at a reference station (SR,TR); that is, S=SR+(T−TR)δS/δT. The method is restrictive in its application because it is most accurate when all water masses in the region of a survey are linear extrapolations from the water masses at each of the reference stations. In reality, it provides useful results when the T and S fields are not simply linear functions of horizontal distance. This approach is particularly useful in regions where, the T(z)−S(z) relation is nonunique, as in the Mediterranean Water in the North Atlantic. The corresponding expression for the lateral density difference for an observed temperature difference (δT) is δρ=−αρ0δT(1−r−1). Observations from regions offshore and along the coast of Portugal are used to evaluate the method. Errors of less than 0.05 psu are exhibited in the evaluation of salinity determined from T-5 XBT drops compared with nearly simultaneous CTD casts. A comparison of water properties and cyclostrophic velocities is made using XCP temperatures and XCP velocities in a meddy.