Improving reconstructions of Northern European and Arctic Holocene Relative Sea Level: a data-model synthesis

Abstract:

Holocene sea level records contain important information about the dynamics of ice sheets and oceans during the current interglacial and inform our understanding of when, how much, and how fast Earth’s ice sheets melted during the deglaciation, given the long-term deformational effects of glacial isostatic adjustment (GIA). Recent Norwegian Sea radiocarbon calibrations suggest that the collapse of the Eurasian Ice Sheet Complex (EIS) coincided with Meltwater Pulse 1a, a period of accelerated sea level rise during the last deglaciation (Brendryen et al., 2020). This alignment implies that an ice sheet larger than the present-day West Antarctic Ice Sheet can disintegrate in fewer than 500 years. This newly proposed deglacial history differs from previous deglacial reconstructions in Northern Europe and the two alternate scenarios would have influenced relative sea level differently during the Holocene, an interval with abundant, highly resolved sea level records in this region.

To better understand Eurasian ice sheet collapse and sea level in Northern Europe and the Arctic during the Holocene, we reconstruct Northern European relative sea level from 12 ka to present using an empirical hierarchical Bayesian framework. We augment an existing database of 2756 sea level index points with newly compiled index points from Norway and Svalbard following rigorous quality-control data standards. Spatiotemporal covariance between paleo sea-level data is derived from a set of forward GIA models to form an empirical covariance function. This allows us to include in the inversion knowledge about physical processes that drive sea level change. The combination of data and Bayesian framework results in a posterior estimate of Holocene sea level in northern Europe. We use this sea level reconstruction to test and compare a suite of existing Eurasian ice sheet models and Earth viscosity profiles through maximum likelihood estimation. We further construct additional synthetic ice reconstructions with different timings of EIS collapse to explore which scenario of this ice sheet’s contribution to MWP-1A is most consistent with RSL data. In this analysis, we also consider how systematic biases in the marine radiocarbon reservoir age affect the ages of certain sea level indicators.