R. Beardsley (WHOI), R. Limeburner (WHOI) and C. Flagg (BNL)

The presentations by Smith (A) and Flagg (C) above describe briefly the moored array deployed on the eastern end of Georges Bank during late 1998-1999 to investigate the presence of Scotian Shelf water there. Earlier work during GLOBEC by Bisagni and others indicated that cold, low salinity water could flow directly from the Scotian Shelf across the Northeast Channel and onto the eastern flank of Georges Bank during winter. It was not clear if such "cross-over" events could occur during spring and summer, since the nominal salinity difference between Scotian Shelf water and Georges Bank water, which is large during winter, becomes much weaker during summer. To help study this phenomenon, a moored array was deployed on the eastern end of Georges Bank. The long-term mooring site on the Northeast Peak ("NEP") was augmented with two mooring sites located on the northern flank ("NFS" and "NFD") and one on the eastern flank ("EF") (Figure A1). The northern and eastern flank moorings were designed to study the vertical structure of flow and water properties at these potential pathways for Scotian Shelf water onto the bank.

Smith (Figure A2) and Flagg (Figure C3) show time series of the near-surface salinity at these mooring sites, which indicate that a number of Scotian Shelf water cross-over events occurred during 1998/1999 using the criterion S 32 to identify Scotian Shelf water. As noted by Flagg, the frequency and duration of periods of low salinity water of Scotian Shelf origin at the various mooring sites are surprising in light of our preconceived ideas. This is particularly true at the eastern flank mooring EF.

We give here a preliminary description of the vertical structure of flow and water properties at EF (Figure A1). A surface toroid buoy supporting VMCMs at 5 and 25 m and SeaCats at 1, 15, 35, and 80 m was deployed at EF on November 17 in 94-m water depth. The toroid dragged its anchor about 1.1 km towards the southeast along the same isobath during a short but intense storm on December 17-19. At some later point, the toroid mooring appeared to have moved approximately 3.7 km towards the southwest, into water 97 m deep. The mooring was recovered and the instruments reset using a surface discus. There is no obvious indication of the mooring movement in the VMCM and SeaCat data, so that we analyze the initial 112-day period covered by the toroid deployment.

Time series of the wind stress at NDBC 44011 and the EF current at 5 m and temperature and salinity at 1m and 80 m are shown in Figure D1. Water with S<32 is found at EF during much of December and part of February. The December crossover event (or series of events) was prolonged, and occurred early in winter before Scotian Shelf waters had reach minimum temperatures. Smith shows the track of drifter 15338 (Figure A3a) that moved from Browns Bank onto eastern Georges Bank during the period February 7-16. This event carried much colder Scotian Shelf water to EF. Shortly after this cross-over, on February 22, warm, high-salinity water moved passed EF, with nearly uniform temperature and salinity over the water column, with a maximum salinity of about 34.8. The T-S diagrams (Figure D2) for the 38-d periods November 17 - December 26 and January 30 - March 9 illustrates the change in the water properties at EF from early winter to deep winter. The latter period is characterized by clear mixing curves between Scotian Shelf water and the deeper slope water, with the February 22 intrusion suggesting the on-bank movement of the shelf/slope front.

The semidiurnal tidal components dominate the 5- and 25-m currents at EF (Figures D3 and D4, Table D1). Approximately 78% (87%) of the current kinetic energy at 5 m (25 m) lies in the tidal band. The subtidal currents are weaker in amplitude but also coherent with depth at EF. The lowest EOF, representing 88% of the current variance, exhibits a reduction of about 33% and a small anti-cyclonic rotation of only 3° between 5- and 25-m. Future analysis will look at the relation between this subtidal current structure, the surface wind stress, and mixed layer variability. Smith suggests that a comparison of the moored array, drifter, and AVHRR data does not support the hypothesis that Scotian Shelf crossovers are forced by wind stress events. At EF, the subtidal currents and wind stress are uncorrelated at zero lag, with large subtidal currents occurring during periods of weak and strong wind stress forcing. While the wind may play some role in perhaps triggering a crossover, this role is not obvious in our preliminary investigations. A more detailed analysis of the relation between wind stress and subtidal currents using all the moored data is needed to address this question.

Table D1. Tidal constituents at EF with major axis greater than 4 cm/s. Analysis period covered the 112-day toroid deployment. For each constituent, the 5-m values are listed above the 25-m results, and 95% confidence limits are enclosed in parentheses. Units: period (hr); major and minor axes (cm/s); inclination (degrees CCW from East); and phase (degrees Greenwich).