| Summary: | Deep water formation may be triggered by the density effect of brines that are released during sea-ice formation. But are sea-ice related brines also a likely mechanism to transport low δ18O signals into deep and bottom waters? Glacial foraminiferal δ18O records in the Nordic seas, covering the period 60 to 15 kyr show anomalously high-amplitude depletions in both benthic and planktic δ18O, particularly during the so-called "Heinrich events" (e.g. Rasmussen et al., 1996). Brine formation has been suggested as a tool to transport these low δ18O signals into the deep and bottom waters during these times (Dokken and Jansen, 1999). In many discussions and publications brines are now generally implied to have a low δ18O signal. Only some papers specifically address the fact that brines and δ18O signal are not necessarily connected (Bauch and Bauch, 2001; Meland et al., 2008; Rasmussen and Thomsen, 2009a, b; Thornalley et al., 2010a,b). Here we discuss a modern analogue situation from the Arctic Ocean. The shallow arctic shelf areas are seasonally ice covered and large amounts of sea-ice are formed here by initial freeze-up and repeatedly during winter in polynyas that open under certain wind conditions. In the Kara, Laptev and East Siberian Seas a low δ18O signal introduced by river water is transported via brine formation into the shelf’s bottom layer at a final salinity of ~30 to 32. These waters are exported into the Arctic halocline and can be identified there at ~30-50 m water depth. In the Barents Sea and the Chuckchi Sea initial salinities are higher and sea-ice formation introduces brines to a water body that has little to no δ18O signal. In these areas brines may penetrate into deeper layers of the water column, but with no effect on the δ18O as indicated also by measurements within the Arctic Ocean deep and bottom waters.
|
|---|