SUMMER SURFACE CIRCULATION ON THE NEWFOUNDLAND SHELF AND GRAND-BANKS - THE ROLES OF LOCAL-DENSITY GRADIENTS AND REMOTE FORCING

Summer surface circulation on the Newfoundland Shelf and Grand Banks i s diagnosed from observed density profiles using the method proposed r ecently by Sheng and Thompson (unpublished manuscript). To assess the accuracy of the predictions they are compared against all available ne ar-surface current measurements for the region. In accord with earlier studies, local density gradients alone cannot account for the summer surface circulation in this region. The surface currents diagnosed fro m the density field are much weaker than the observations, particularl y near the shelf break. To explain the discrepancy we use a simple inv erse method to infer the optimal inflow boundary conditions for a baro tropic model from the differences between the observed and diagnosed c urrents. The surface circulation is then modelled as the sum of the di agnosed baroclinic component and the remotely-forced barotropic compon ent Overall the predicted circulation pattern agrees reasonably well w ith the current observations and maps of the mean surface circulation based on a variety of data sources. The predicted circulation pattern has identifiable inshore and offshore branches of the Labrador Current that split and merge as they encounter the main banks and saddles of the region. The mean square of the observed currents (J(obs)) is reduc ed by 52% on removal of the combined effect of local density gradients and remote forcing. Removing the baroclinic component alone reduces J (obs) by 40%. Our conclusion is that both remote-forcing and local den sity gradients make significant contributions to the circulation in th is region. For example the strong shelf-break flow on the northeast Ne wfoundland Shelf appears to be due mainly to remote-forcing while the circulation over Flemish Cap, and the narrow coastal jet off southeast Newfoundland are due mainly to local density gradients.