In 2006 Wallace Broeker first suggested that the sudden interruption of emergence from the last glacial maximum by a frigid climate about 12.8 ka was due to a massive release of fresh water to the North Atlantic that shut down its thermohaline ‘conveyor’ (see The Younger Dryas and the Flood in June 2006 issue of EPN). He resurrected an earlier idea that a vast lake of glacial meltwater (Lake Agassiz) to the north-west of the Great Lakes of North America burst down the St Lawrence Seaway, instead of quietly escaping to the Gulf of Mexico along the Missouri-Mississippi system. His hypothesis was that the resulting freshening of surface water in the North Atlantic and decreased density stopped the formation of cold dense brines that sink and drag warm water northwards. Setting aside the notion by some enthusiastic authors that a trigger for the Younger Dryas was an exploding comet and a kind of ‘nuclear winter’ (see Whizz-bang view of Younger Dryas and Impact cause for Younger Dryas draws flak in EPN July 2007 and May 2008) Broeker’s hypothesis is widely accepted. However there are few signs, if any, of a catastrophic glacial-lake outburst through the Great Lakes region and down the St Lawrence. An alternative is that Lake Agassiz drained northwards towards the Arctic Ocean. (Since the North American ice sheet covered Hudson’s Bay that could not have been the destination.) At the end of the last last full glaciation there was a corridor with relatively little glacial cover between the main ice over the Canadian Shield and that mantling the Rocky Mountains, roughly along the course of the modern Mackenzie River. That route would serve the hypothesis well, and there is clear evidence that an outburst flood followed it (Murton, J.B. et al. 2010. Identification of Younger Dryas outburst flood path from Lake Agassiz to the Arctic Ocean. Nature, v. 464, p. 740-743).
Sediments of the huge Mackenzie Delta of NW Canada contain a sharp erosion surface overlain by gravels that belie the low-energy of deposition today. Optically stimulated luminescence dating of sediment immediately below and above the erosion surface range from 13.4 (below) to 12.7 ka (above), the latter approximating the onset of frigid Younger Dryas conditions. The surface occurs all the way along the Mackenzie into its major tributary the Athabasca River. Near Fort MacMurray, 20 km north of what was the northern shore of Lake Agassiz, there is a terrace composed of massive boulders. Further evidence comes from the apex of the Mackenzie delta in the form of a 25 km long, 2 km wide spillway scoured of all loose sediment and with topographic features reminiscent of the famous Channelled Scablands of Washington State in the NW USA. Numerous beach lines record the drainage of Lake Agassiz, the highest being dated at the start of the Younger Dryas and giving a clue to the volume involved in the initial outburst flood: around 9500 km3. Dating of other features suggest that a second flooding into the Arctic Ocean occurred during the Younger Dryas around 11.5 ka, during its last stages, and a third at 9.3 ka. One effect of the Younger Dryas was a regrowth of the main ice sheet that allowed Lake Agassiz to refill periodically perhaps allowing quieter flooding events down the Mississippi and through the Great Lakes. There are no signs in the climate record of any major perturbation at 9.3 ka.
Broeker received the news graciously, commenting that a freshening of the Arctic Ocean would have been more effective at shutting down North Atlantic thermohaline circulation than a spillway down the St Lawrence, because the sites of modern day sinking of dense cold brine lie well to the north of its outlet. The only way additional water in the Arctic Ocean could escape would have been into the northernmost North Atlantic.
See also: Schiermeier, Q. & Monastersky, R. 2010. River reveals chilling tracks of ancient flood. Nature, v. 464, p. 657.