Ice Age Sea Level Change on a Dynamic Earth

Abstract:

Changes in global mean sea level (GMSL) are a sensitive indicator of climate variability during the current ice age. Reconstructions are largely based on local sea level records, and the mapping to GMSL is computed from simulations of glacial isostatic adjustment (GIA) on 1-D Earth models. We argue, using two case studies, that resolving important, outstanding issues in ice age paleoclimate requires a more sophisticated consideration of mantle structure and dynamics.

First, we consider the coral record from Barbados, which is widely used to constrain global ice volume changes since the Last Glacial Maximum (LGM, ~21 ka). Analyses of the record using 1-D viscoelastic Earth models have estimated a GMSL change since LGM of ~120 m, a value at odds with analyses of other far field records, which range from 130-135 m. We revisit the Barbados case using a GIA model that includes laterally varying Earth structure (Austermann et al., Nature Geo., 2013) and demonstrate that neglecting this structure, in particular the high-viscosity slab in the mantle linked to the subduction of the South American plate, has biased (low) previous estimates of GMSL change since LGM by ~10 m. Our analysis brings the Barbados estimate into accord with studies from other far-field sites.

Second, we revisit estimates of GMSL during the mid-Pliocene warm period (MPWP, ~3 Ma), which was characterized by temperatures 2-3°C higher than present. The ice volume deficit during this period is a source of contention, with estimates ranging from 0-40 m GMSL equivalent. We argue that refining estimates of ice volume during MPWP requires a correction for mantle flow induced dynamic topography (DT; Rowley et al., Science, 2013), a signal neglected in previous studies of ice age sea level change. We present estimates of GIA- and DT-corrected elevations of MPWP shorelines from the U.S. east coast, Australia and South Africa in an attempt to reconcile these records with a single GMSL value.