J. O'Dwyer et al., Does the potential vorticity distribution constrain the spreading of floats in the North Atlantic?, J PHYS OCEA, 30(4), 2000, pp. 721-732
Float trajectories are compared with the distribution of climatological pot
ential vorticity, Q, on approximate isentropic surfaces for intermediate wa
ters in the North Atlantic. The time-mean displacement and eddy dispersion
are calculated for clusters of Boats in terms of their movement along and a
cross Q contours. For float clusters with significant mean velocities, the
mean flow crosses Q contours at an angle of typically less than 20 degrees-
30 degrees in magnitude in the ocean interior. The implied Peclet number in
the ocean interior ranges from 1 to 19 with a weighted-mean value of 4.4.
This mean Peclet number suggests that there is significant eddy mixing in t
he ocean interior: tracers should only be quasi-conserved along mean stream
lines over a subbasin scale, rather than over an entire basin. The mean flo
w also strongly crosses Q contours near the western boundary in the Tropics
, where the implied Peclet number is 0.7; this value may be a lower bound a
s Q contours are assumed to be zonal and relative vorticity is ignored. Flo
at clusters with a lifetime greater than 200 days show anisotropic dispersi
on with greater dispersion along Q contours, than across them; Boat cluster
s with shorter lifetimes are ambiguous. This anisotropic dispersion along Q
contours cannot generally be distinguished from enhanced dispersion along
latitude circles since Q contours are generally zonal for these cases. Howe
ver, for the null case of uniform Q for the Gulf Stream at 2000 m. there is
strong isotropic dispersion, rather than enhanced zonal dispersion. In sum
mary, diagnostics suggest that floats preferentially spread along Q contour
s over a subbasin scale and imply that passive tracers should likewise pref
erentially spread along Q contours in the ocean interior.