Sea-level probability for the last deglaciation: A statistical analysis of far-field records

Abstract

Pulses of ice-sheet meltwater into the world ocean during the last deglaciation are of great current interest, because these large-scale events offer important test-beds for numerical models of the responses of ocean circulation and climate to meltwater addition. The largest such event has become known as meltwater pulse (mwp) 1a, with estimates of about 20 m of sea-level rise in about 500 years. A second meltwater pulse (mwp-1b) has been inferred from some sea-level records, but its existence has become debated following the presentation of additional records. Even the use of the more ubiquitous mwp-1a in modelling studies has been compromised by debate about its exact age, based upon perceived discrepancies between far-field sea-level records. It is clear that an objective investigation is needed to determine to what level inferred similarities and/or discrepancies between the various deglacial sea-level records are statistically rigorous (or not). For that purpose, we present a Monte Carlo style statistical analysis to determine the highest-probability sea-level history from six key far-field deglacial sea-level records, which fully accounts for realistic methodological and chronological uncertainties in all these records, and which is robust with respect to removal of individual component datasets. We find that sea-level rise started to accelerate into the deglaciation from around 17 ka BP. Within the deglacial rise, there were two distinct increases; one at around the timing of the Bølling warming (14.6 ka BP), and another, much broader, event that just post-dates the end of the Younger Dryas (11.3 ka BP). We interpret these as mwp-1a and mwp-1b, respectively. We find that mwp-1a occurred between 14.3 ka BP and 12.8 ka BP. Highest rates of sea-level rise occurred at ~ 13.8 ka, probably (67% confidence) within the range of 100–130 cm/century, although values may have been as high as 260 cm/century (99% confidence limit). Mwp-1b is robustly expressed as a broad multi-millennial interval of enhanced rates of sea-level rise between 11.5 ka BP and 8.8 ka BP, with peak rates of rise of up to 250 cm/century (99% confidence), but with a probable rate of 130–150 cm/century (67% confidence) at around 9.5 ka BP. When considering the 67% probability interval for the deglacial sea-level history, it is clear that both mwp-1a and -1b were relatively subdued in comparison to the previously much higher rate estimates.