23 - Past the point of no return: the Younger Dryas
Glacial's last hurrah
The Younger Dryas stage (YD; 12.85 - 11.7 ka) has been fascinating scientists for decades. It was so named when 19th century Scandinavian pollen analyses of the sediments overlaying the Bølling-Allerød (BA) discovered a relative abundance of Dryas octopetalys, an indicator species of arctic-alpine regions, signifying a post-BA return to colder, arctic climate conditions. But it wasn’t until the late-20th century ice coring projects and the cultural focus on climate change that interest in the stage truly surged. Oddly, the earlier “Older Dryas” stage between the Bølling and the Allerød warm periods that is characterized by the same Dryas octopetalys did not benefit from this renewed interest: many scientists preferred (and still prefer) to think of the YD as a unique example of catastrophic climate change. A previous post however demonstrated the YD is in fact the end member of a series of progressively colder European climates that started with the Older Dryas.
This renewed YD interest is both a blessing and a curse. The good news is the significant surge in relevant YD data and studies that are designed to either confirm or debunk the competing climate change theories; a good, fairly comprehensive summary can be found in Carlson (2013)[1]. The bad news is that most authors propose low probability, internally contradictory theories that explain some of the data, but ignore, omit or attempt to ‘splain away the large volume of other, conflicting data, using words such as “possibly” and “perhaps” to ‘splain why their data zigged not zagged.
NH Summer Solar Insolation continues to increase
Mid-latitude Northern Hemisphere (NH) summer solar insolation continued to increase during the YD, prolonging the rise that started around 16 ka. Global sea levels also continued to rise during the YD, suggesting that solar insolation increases were very likely largely responsible for the vigorous (North American) YD melting.
The North Atlantic winter sea ice cover re-advances to its LGM extent
A previous post established that geothermal heat increases along the mid-Atlantic ridge around the start of the BA (14.7 ka) caused the North Atlantic sea ice cover to rapidly retreat to just-offshore Greenland, thereby reducing Europe’s and central Greenland’s distance to open oceans and their warm weather systems, and shifting the regions from the “Slow Growth” into the “Goldilocks" Zone. It also documented the gradual regrowth of the North Atlantic sea ice cover over the course of the BA, causing central Greenland accumulation rates and temperatures to gradually decrease in lockstep as Greenland slid back into the “Slow Growth” Zone during the late BA and YD. This gradual decrease/regrowth was interrupted by 3 very rapid sea ice cover advances - during the Older Dryas (ORD; 14.05 - 13.9 ka), the 13.3 ka event, and the YD - that followed large eastern US meltwater influxes into the North Atlantic.
The European climate abruptly cools and glaciers re-advance
Faunal, floral, speleothem and oxygen isotope temperature proxies all indicate that much of Europe suffered a 2–6 ºC cooler climate during the YD[1]. The Scandinavian Ice Sheet and Swiss Alpine glaciers re-advanced, indicating much of Europe had shifted from the “Battle Zone” into a mainly colder “Goldilocks Zone” following the re-advancing YD sea ice cover. According to Isarin et al.[4]:
Sea-ice in the North Atlantic Ocean played a decisive role in the [YD] climate of northwestern and central Europe by cooling the surface air temperatures
They also identify a subtle climate shift over the course of the YD due to a sea ice cover retreat:
A subdivision of the Younger Dryas into a phase of maximum cold and humidity followed by a less cold and relatively dry phase may be explained by a slight north- ward shift of the mean Atlantic sea-ice margin during winter
Following the initial St Lawrence freshwater influx and rapid sea ice cover advance at the start of the YD the North Atlantic sea ice cover gradually retreated again.
North American ice sheets continued to melt vigorously
North Atlantic sea ice cover had relatively little impact on the mid-western North American glacial melting that was charaterized by the further broadening of the Oldest Dryas interglacial corridor, which acted as an expanding low albedo melting nucleus during the BA and YD. The Cordilleran Ice Sheet had disappeared by the end of the YD: its voluminous YD meltwaters cooled the Northwest Pacific and the western US[1] via the California Current.
Routing of the Lake Agassiz freshwater discharge
During the YD the large volumes of mid-west Canadian meltwaters could not effectively drain via the usual drainage paths (Mackenzie, Mississippi, St Lawrence) or via the ice-blocked Hudson Bay and therefore continued to partially pond in the expanding glacial Lake Agassiz.
Five phases of Lake Agassiz drainage have been identified. During the initial (BA) Lockhart Phase (13,560-12,875 BP) Lake Agassiz mainly drained in a southerly direction, towards the Mississippi, causing freshwater discharge increases and sea surface temperature decreases during the late-BA (below). During this phase the Laurentide ice sheet had not yet retreated from Lake Superior.
During the subsequent (~YD) Moorhead Phase (12,875-11,690 BP) Gulf of Mexico freshwater discharge dropped and sea surface temperatures remained constant, indicating relatively low Lake Agassiz drainage via the southerly (Mississippi) route. Lake Agassiz experienced a sudden lake level drop during the Moorhead Phase when the ice dam blocking the Kaministikwia route to the east retreated, opening the eastward drainage route through Ontario via Lake Superior to the St Lawrence River. This route would be gradually blocked again by a Laurentide ice sheet advance during the late-YD that progressively closed off this eastward route, thereby slowly reducing the St Lawrence River freshwater influx into the Northern Atlantic. Carlson[1] confirms:
Four independent geochemical freshwater source and amount proxies and mollusk 𝛿18O from the St. Lawrence River and Estuary support the eastward routing of western Canadian Plains freshwater at the start of the Younger Dryas. Modeling of the geochemical records indicates a base-flow freshwater discharge increase of 0.12 Sv (Sv = Sverdrup, 106 m3 s-1)
A back of the envelope calculation demonstrates that a 0.12 Sv increase in freshwater discharge is realistic. Assuming a ~20 m global sea level rise over the YD implies 20 m * 3.3 1014 m2 (Area of Oceans) = 6.6 1015 m3 of meltwater entered the oceans over the 1150 year YD duration. This averages out to 5.7 1012 m3 per year, or 0.18 106 m3 s-1 (Sv). A sudden Lake Agassiz lake level drop at the start of the YD could therefore plausibly supply part or all of 0.12 Sv (and very likely more) at the YD start.
Note that unlike the Lake Missoula megafloods such 0.12 Sv flows are not necessarily catastrophic: estimates of the Lake Missoula flood flows are on the order of 1 Sv, while 0.12 Sv is slightly lower than the current total river runoff into the Arctic Ocean. In addition, the drainage path along the Great Lakes (fill and spill) would very likely smooth out any Lake Agassiz flood surge.
The relationship between YD temperatures and ocean currents
A previous post demonstrated that temperatures generally cooled in the Northern Hemisphere (NH) and warmed in the Southern Hemisphere (SH)[1] during the YD. Antarctic temperature reconstructions show a ~3-5 °C increase, very likely due to the weakening of the Atlantic Meridional Overturning Circulation (AMOC): the NH to SH export of cold water diminished and SH temperatures returned to their pre-BA levels.
Many authors attribute the YD cooling of the NH to a weakening of the AMOC, e.g. [6] It is however very unlikely that any AMOC weakening was caused by a weakening or cooling of the North Atlantic Gyre (red loop above) as southeastern US temperatures warmed during this period, indicating the Gyre/ Gulfstream strengthened and warmed rather than weakened and cooled during the YD. A more likely explanation for the European and central Greenland temperature drops is that the European and Greenland distance to open oceans and their warm weather systems abruptly increased at the start of YD due to the expansion of the North Atlantic sea ice cover.
Similarly, the sea ice cover advance at the start of the YD likely caused northeastern US/Great Lakes area temperatures to decrease again over the course of the YD, triggering a Laurentide ice sheet re-advance in these areas. This YD re-advance in the Lake Superior area ice-dammed the eastward Lake Agassiz drainage path, which in turn reduced freshwater influx into the North Atlantic via the St Lawrence river.
West coast US temperatures very likely decreased due to a cooling of the California Current due to the large influx of cool meltwaters from the Cordilleran ice sheet.
What happened to the AMOC during the YD?
A previous post demonstrated that the North Atlantic Deep Water (NADW) current strength - as reconstructed from sediment cores near the Bermuda Rise - decreased dramatically following the large BA St Lawrence River freshwater meltwater influxes. Currently, the low salinity St Lawrence river waters must travel several hundreds of kilometers south or thousands of kilometers east in the North Atlantic before effectively encountering - and mixing with - the warmer, more saline waters of the North Atlantic Central Surface Water (NACSW). The 0.12+ Sv freshwater pulse at the start of the YD (or Older Dryas or 13.3 ka event influxes) would have temporarily overwhelmed the NACSW: a large volume of cold, low saline St Lawrence river water would have been density-trapped at surface in a North Atlantic Low Density Water Mass (NALDWM) in a manner similar to the present-day river waters flowing into the Arctic (above). Such a low salinity, cold water mass at surface would have frozen over relatively easily, causing the re-advancement of the North Atlantic sea ice cover. The NALDWM formed between Newfoundland and northwest Spain (blue and green areas above), but gradually became denser and areally smaller over the course of the YD as St Lawrence river freshwater influx decreased (due to Lake Superior ice damming), and evaporation and ice-formation increased its salinity.
The bulk of the NACSW remained in the North Atlantic Gyre, effectively caught in a high temperature, low density loop. Just offshore Newfoundland a relatively weak northeastward wind component would have allowed a relatively low velocity NACSW to sink beneath the even lower density NALDWM, thereby generating the very weak and relatively warm North Atlantic Deep Water (NADW) current.
Summary
The Younger Dryas was characterized by an abrupt shift to a “a phase of maximum cold and humidity followed by a less cold and relatively dry phase” in Europe, Greenland, and northeastern US. This shift was very likely due to a regrowth of North Atlantic sea ice cover between Newfoundland and northwest Spain that “played a decisive role in the climate of northwestern and central Europe by cooling the surface air temperatures”. A 0.12+ Sv freshwater influx into the North Atlantic via the St Lawrence Seaway around the start of the YD generated a “density-trapped” cold and low saline North Atlantic Low Density Water Mass. This NALDWM formation caused the re-advancement of the North Atlantic sea ice cover as well as a reconfiguration of the AMOC and an abrupt decrease in the North Atlantic Deep Water (NADW) current strength. Over the course of the YD the NALDWM gradually became smaller, ultimately disappearing around 11.7 ka, due to the progressive reduction of the freshwater influx via the St Lawrence river, the evaporation and ice-forming of the NALDWM, the retreat of the sea ice cover, and the warming of the water mass via the NACSW / Gulfstream.
References
[1] Carlson, A., 2013, The Younger Dryas Climate Event. In: Elias, S.A. (editor). The Encyclopedia of Quaternary Science 3: 126–134. Amsterdam: Elsevier.
[2] Andersen, B., Borns, H., 1994, The Ice Age World, Oxford University Press. ISBN 978-8200218104
[3] Hughes, A. et al., 2015, The last Eurasian ice sheets – a chronological database and time-slice reconstruction, DATED-1. Boreas, 45, 10.1111/bor.12142.
[4] Isarin, R., 1998, The impact of the North Atlantic Ocean on the Younger Dryas climate in northwestern and central Europe. J. Quaternary Sci., 13, 447–453. ISSN 0267-8179
[5] Williams, C. et al., 2010, Deglacial abrupt climate change in the Atlantic Warm Pool: A Gulf of Mexico perspective, Paleoceanography, 25, PA4221, doi:10.1029/2010PA001928.
[6] Meissner, K. and Clark, P., 2006, Impact of floods versus routing events on the thermohaline circulation, Geophys. Res. Lett., 33, L15704, doi:10.1029/2006GL026705.