Low‐frequency fluxes of momentum, heat, salt, and nutrients at the edge of the Scotian Shelf

Abstract

Measurements of current, temperature, and salinity from a single mooring at the edge of the Scotian Shelf are analyzed for the period from December 1975 to April 1976. Major contributions to the subtidal Reynolds stresses consist of the wind‐driven flows (2‐ to 8‐day periods) and offshore forcing effects (10–30 days). These may be distinguished by a difference in sign of the onshore momentum flux in each band. Uniform divergence of these stresses across the shelf would not be significant to the general circulation. The shoreward eddy transport of salt and heat are maximum at 50‐m depth and are concentrated in the lowest frequency bands. Considered uniform at the open boundary, the 50‐m fluxes would support temperature and salinity gradients of 1.5°C and 0.5‰ per 100 km along the Scotian Shelf. Also, a method is presented by which estimates of nutrient fluxes at the shelf break are made by establishing a relationship between nutrients and TS properties through another variable, the concentration of dissolved oxygen. The subtidal shoreward transport of nitrate is found to be 12 ± 4 μg atom N m−2 s−1, which meets Fournier's (1977) estimate (15.5 μg atom N m−2 s−1) of the spring and summer phytoplankton requirements on the shelf. Finally, current meter and hydrographic data and satellite imagery are used to estimate the net salt and heat transports produced by the entrainment of surface shelf water by a Gulf Stream eddy. On an annual basis (6 events/yr) the volumetric exchange (∼104 km3 yr−1) is greater than the estimate for a similar event in the Mid‐Atlantic Bight (Morgan and Bishop, 1977) but consistent with figures for the transport in shelf water filaments along the Gulf Stream (Kupferman and Garfield, 1977). The resulting contributions of salt and heat to the shelf waters are 3 × 1013 kg yr−1 and 6 × 1019 cal yr−1, which are equivalent to the measured subtidal transports at the shelf break. These estimates exceed by a factor of 6 those of Voorhis et al. (1976) and Horne (1978) for small‐scale frontal mixing.