Finding the source of Meltwater Pulse 1A

The most recent glacial period reached its climax around 21,000 years ago.  Glacial melt-back started between about 16,000 and 14,500 years ago (different times in different places), and the major North American and Eurasian ice sheets had receded back to their present-day extents by about 6,000 years ago.

The melting of glacial ice had two major consequences for sea levels – namely the eustatic (ie. global) rise in sea water levels (related to the melt-water), and the regional-scale isostatic adjustment of land levels (related to changes in the weight of ice-cover).  These isostatic adjustments were regional because the crust rebounded only in the regions around the ice sheets (ie. in North America, Antarctica, Greenland and Eurasia).

Over the past 16,000 years the total eustatic sea level rise has been in the order of 105 m – equivalent to around 6 mm/y.  We know that the rate of melting varied during this time, and that there were some pulses of rapid melting and rapid sea-level rise (SLR) (Bard et al., 1990).  The best known of these pulses is referred to as Meltwater Pulse 1A (mwp-1A), which started at around 14,200 years ago and lasted for close to 500 years.  Over this short period, sea level rose approximately 20 m – or around 40 mm/y – almost 7 times the average rate.  While the existence of mwp-1A has been known for some time, there has been uncertainty about which ice sheet(s) melted quickly to produce the pulse.

Geoscientists from Oregon State University, the University of Toronto and the University of Durham in England (Clark et al, 2002) have approached the problem by analysing sea-level data for the mwp-1A period, and then combining the eustatic SLR predictions with various different predictions based on isostatic adjustments for different regions.  A model for combined eustatic and isostatic sea-level adjustment in the case where most of the melting for mwp-1A is from the Laurentian Ice sheet (eastern Canada) is shown on Figure A.  In this scenario Barbados should be affected by significant isostatic adjustment (due to the loss of ice load in North America) while Sunda Shelf (Indonesia) should not.  According to Clark et al., strong melting in North America should have produced a 25 m SLR at Barbados and a 38 m SLR at Sunda Shelf during mwp-1A.  The data show that both Barbados and Sunda Shelf experienced an equivalent overall SLR of close to 25 m during mwp-1A.

The Barbados and Sunda Shelf observations are more consistent with the model shown in Figure B, where most of the mwp-1A melting comes from Antarctica, and relatively little comes from North America.

One of the important implications of this work is in our understanding of the effects of rapid melting on the circulation of water in the Atlantic Ocean.  It is postulated that a large pulse of melt water from the Laurentian Ice Sheet would have put enough fresh water into the Atlantic Ocean to shut down the formation of NADW.

It has been observed that a decrease in NADW circulation did not take place until around 1000 years after mwp-1A, and this is consistent with the suggestion that most of the fresh water from mwp-1A was derived from Antarctica, and not from North America.  

This work also suggests that the date for the initial melting of Antarctic ice needs to be revised, to several thousand years earlier than was previously thought (Clark et al.).

References

Clark, P., Mitrovica, J., Milne, G. and Tamisiea, M., Sea level fingerprinting as a direct test for the source of global melt water pulse 1A, Science, V. 295, p. 2438-41, March 2002.

Bard E., Hamelin B., Fairbanks R. G., and Zindler A.,  Calibration of the 14C timescale over the past 30,000 years using mass spectrometric U-Th ages from Barbados corals, Nature 345, 405-410 (1990)

Steven Earle, 2002. Return to Earth Science News