Ship-mounted ADCP and buoy data are used to fit an azimuthal velocity profi
le to a N.E. Atlantic mesoscale eddy chosen as a site for a Lagrangian biog
eochemical survey, the cruise forming part of the UK Plankton Reactivity in
the Marine Environment (PRIME) initiative. Together with the buoy-derived
locus of the eddy centre, the ADCP-derived velocity held allows observation
s of a sulphur hexafluoride (SF,) tracer release within the eddy to be put
into a non-translating, non-rotating frame. Analysis of the transformed dat
a suggests that the patch of tracer, though spreading, remained coherent th
roughout the nine-day survey, A linear regression on the square of the radi
al patch width versus time gives an estimate for the effective horizontal d
iffusion coefficient of 22 +/- 10 m(2) s(-1). This is consistent with previ
ous estimates of diffusion rates at the O(10 km) lengthscale of the patch.
Theory predicts a corresponding along-streamline spreading time of similar
to 14 days. This implies that the tracer patch mixed little with the surrou
nding waters during the first five days of the survey, suggesting that biog
eochemical processes were little affected by lateral mixing during this per
iod. The theory is inapplicable at later times because the area was struck
by a storm at the start of the sixth day, which resulted in the halving of
SF, concentrations. Using a model of a;circular eddy with the calculated ve
locity profile, the dispersion of the initial,patch of tracer is simulated
for different diffusivities, the results confirming the estimate of along-s
treamline mixing time and demonstrating the enhancement of azimuthal diffus
ion by the radial shear of the flow. (C) 2001 Elsevier Science Ltd. All rig
hts reserved.