The [formula omitted] float—measuring stretching vorticity directly

Abstract

By changing the volume of an isopycnal float by a small, prescribed amount, one can force it to rise or sink to neighboring isopycnal surfaces. The pressure difference between these is a simple measurement of stratification, Δϱ Δz . This technique is adapted to the RAFOS float technology to measure stretching vorticity changes along the trajectories of floats in Gulf Stream meanders. The measurement uncertainty of the depth of an isopycnal is only a few meters, but the uncertainty of the layer thickness estimate, when measured over about 200 m, is about 7 m due to differential advection of small scale density structure. This includes all periods between the 8 × day −1 sampling rate of the prototype float and 1 day. The records from two floats are discussed in detail. As the floats up- or downwell and are expelled from the Gulf Stream, order 10–20% changes in stretching vorticity are observed. Particularly along the cyclonic edge of the current the changes can be quite abrupt corresponding to 1–2 × 10 −6s −1 horizontal divergences. The changes in curvature vorticity, estimated from the float trajectories, are, by and large, weaker and tend to evolve more gradually. On the assumption that potential vorticity is conserved, changes in shear can be estimated. For the two cases discussed, one shows suggestive similarity with what might be expected using a “frozen field” model of downstream velocity. The other retains strong negative shear after expulsion from the stream, which may be indicative of an anticyclonic circulation just south of a meander crest.