Seasonal and interannual covariance of the Gulf Stream/North Atlantic Current north wall and shelf-slope front positions between 75W and 50W

Rossby (1999) suggests that the position of the Gulf Stream (GS) and North Atlantic Current (NAC) may be controlled, in part, by a "spilling" of colder, less-saline shelf water into slope waters south of Nova Scotia, resulting in both seasonal and interannual variability in the position of the GS/NAC system. Analysis of up to 29 years (1973-2001) of GS/NAC "north wall" and shelf-slope front (SSF) positions, derived from weekly frontal charts from satellite-derived sea surface temperature, allows a test of the shelf water "spilling" hypothesis. Time series containing up to 348 mean monthly GS/NAC and SSF positions were produced along each of 26 longitude lines (75W-50W) by digitizing individual frontal charts and computing mean monthly latitudinal positions over the 29-year period. Seasonal variability of the GS/NAC and SSF position at each longitude was computed by subtracting an overall mean frontal position from 12 long-term mean monthly GS/NAC and SSF frontal positions. Least-squares bisector linear regression between GS/NAC and SSF seasonal variability at each longitude shows a maximum zero-lag correlation (p<0.05) which accounts for 75% of the variance at 61W. Interannual variability of the annual mean GS/NAC and SSF position anomaly at each longitude was computed by subtracting the long-term mean monthly position from each corresponding month's value over the full 29-year period, followed by computation of the annual mean anomaly for each year. Linear regression between GS/NAC and SSF interannual variability shows a maximum zero-lag correlation accounting for 62% of the variance at 62W and contained within a broad peak of significant (p<0.05) correlation from 63W-58W. Maximum correlation from both the seasonal and interannual variability analyses occur along lines of longitude which intersect Nova Scotia and are approximately equal to the longitude for Cabot Strait (60W), between Newfoundland and Cape Breton Island. Thus, covariance of the GS/NAC and SSF south of Nova Scotia at seasonal and interannual time scales supports the shelf water "spilling" hypothesis and at least some thermohaline control of the GS/NAC position from 63W-58W. Furthermore, over seasonal timescales, the southern-most position of the GS/NAC and SSF by early winter may be related to maximum outflow of the St. Lawrence estuary through Cabot Strait during the previous August, after allowing for a 3-4 month lag for the freshwater to cross the continental shelf south of Nova Scotia. Interestingly, analyses of low-frequency variability of the GS/NAC and SSF positions from these same data show a change in the spectra between 60W-62W, displaying dominant periods of approximately 8 years and 3 years west of this location, with only a single peak near 8 years east of this location. Both 8 and 3-year periods are found in the North Atlantic Oscillation (NAO) wintertime index during the1972-2001 period. Rossby, T., 1999. On gyre interactions. Deep-Sea Research II, 46: 139-164.