The spatial structure of surface currents collected using a Doppler Ra
dar system off the Florida Keys has been investigated, Surface current
vectors were acquired within a rectilinear grid containing 700 nodes,
where each node was spaced 1 km apart. Principal component analyses i
ndicate that at least 63% of the variance of the surface currents at t
he 700 nodal locations can be accounted for by only three principal co
mponents, The principal modes of circulation during two separate exper
iments were extracted and found to be basically the same, although the
first two modes were reversed between the two experiments. Coherence
spectra were determined for nodal spacings between 1.76 km and 10.48 k
m, and the results of these analyses indicate that over most of the ex
perimental area, approximately 60% of the variance is coherent over se
parations of 10.48 km, Application of a synoptic time-series model [1]
indicates that accurate prediction of the mean currents over each of
the 65-h subrecord lengths was the dominant factor in controlling mode
l performance, and that on the order of 10% error could be expected in
using the time-series model to predict the low-frequency fluctuations
, This indicates that time-series modeling of surface currents may be
feasible and useful in estimating the long-term mixing characteristics
of contaminants transported in the surface layer.