We map sea surface height anomaly on 1/6*1/6 degree grids at 5-day intervals, using TOPEX/POSEIDON, ERS and coastal tide gauge data, by a two-stage optimal interpolation. Our present focus is on the Leeuwin current, an anomalous poleward-flowing eastern boundary current, with an energetic eddy field, off the west coast of Australia

Altimetric geostrophic velocity estimates compare well with the 6-hourly krigged WOCE SVP drifter velocities, and both moored and vessel-mounted ADCP observations. The rms in-situ surface velocity off western Australia is 0.48m/s (N=4438). The rms altimetric estimate (linearly interpolated in time and space to the observation points) is 0.33 m/s, and the rms vector difference of the two is 0.34m/s. If only T/P is used, the rms altimetric estimate drops to 0.26m/s and the difference from in-situ observations increases to 0.39m/s. ERS is clearly helping.

Animations of LAC AVHRR thermal imagery also provide convincing evidence that the altimetric velocity estimates are essentially correct, even at the meso-scale.

One interesting observation that was made while comparing drifter tracks with altimetry was that drifters appear to spend a lot of time in cyclonic eddies. One stayed so close to the eye of a cyclone for 225 days (from 113E to 105E at 33S) that without altimetry, you cannot tell it is in an eddy.

The motivation for the present study is to see why yearly-averaged coastal sea level at Fremantle (Perth) correlates well with yearly-averaged settlement of lobster larvae. Both are high in La Nina years, when the Leeuwin flows strongly. We are using the altimetric velocity fields to test competing theories of why this is so.