Modification of halocline source waters during freezing on the Beaufort Sea shelf: evidence from oxygen isotopes and dissolved nutrients

During some, but not all winters, waters on the Mackenzie shelf of the Beaufort Sea become sufficiently saline to ventilate the halocline of the adjacent Canada Basin. This occurred in March 1988, at which time a survey of the temperature, salinity, dissolved nutrient and ¹⁸O properties of the ventilating waters was completed. The regional hydrography of 1988 was very similar to that of 1981, when ventilation also occurred in this area. The δ¹⁸O-salinity properties of the cold, saline shelf waters revealed that in the winter of 1987–1988, ice was grown from water initially more saline by about 1.5 psu] than is typical for the area. The higher initial salinity appears to have been a consequence of a two-stage conditioning of shelf waters by storms in the autumn of 1987. Since the amount of ice growth, and consequent salt rejection, over the winter of 1987–1988 was abnormally low, this conditioning played a crucial role in the formation of the ventilating water mass. Nutrient concentrations in ventilating waters were the same as those of waters unaffected by freezing. Thus significant regeneration of nutrients within the cold saline shelf waters did not occur during their 6-month period of formation. In consequence, the nutrient signatures carried into the arctic halocline by winter shelf waters from this area tended to erode, rather than to reinforce the nutrient maxima. For this reason they are not the dominant source of supply to the arctic halocline. Waters in the Chukchi and northern Bering Seas during the same period had δ¹⁸O values intermediate between those on the Mackenzie shelf and those in the arctic halocline. Thus winter shelf waters are supplied to the arctic halocline with a range of nutrient, temperature, salinity and δ¹⁸O properties. On average, the southern Canada Basin is an impressive net producer of sea ice. The net rate of production from waters in the upper 350 m in this area is about 2 m y⁻¹, approximately twice the net rate of production in the central Arctic Ocean.