MATCHING THE DATING RESOLUTION OF GLACIER BEHAVIOR TO DATING RESOLUTION OF POTENTIAL FORCING (Invited Presentation)

Numerous efforts correlated glacial moraines to ice core records to test for causal relationships. The success may hinge on the uncertainties inherent in the dating technique. For example the uncertainties in a set of 10Be ages approximate the duration of the climate events during the late Glacial but varve sequences have the same resolution Radiocarbon falls in between Further an often employed hidden assumption is that moraines, formed at a stationary glacier configuration, represent glacial advances under cold conditions. Stationary ice margins can form for other reasons. The lag between the start of a cold time and a glacier achieving equilibrium scales with the size of the glacier but can lie within the uncertainty of some dating approaches.

A sequence of moraines formed by the former Laurentide Ice Sheet located west of Lake Superior illustrate these issues. Here 14C and 10Be dating along a 400 km transect, with sample sites largely between moraines, reveals two relevant findings: 1) the glacier margin was retreating at rates from 39 to 48 m/yr retreating during the Younger Dryas. A retracting ice sheet during “known” cold times implies either a different climate history in North America than recorded in the Greenland Ice Core, or a mis-correlation because of the different dating resolutions. Since the dating possibility cannot be eliminated any climate insight is speculative. 2) Bayesian analysis indicates that the moraines collectively constitute only 10% of a 9 ka long record. They do not represent the majority of time nor conditions of the ice sheet. A test of the first finding comes from a developing varve record across moraines in the center of the transect. The annual resolution shows faster rates of ~68 m/yr but preliminary dating also indicates a retreating ice sheet during the Younger Dryas.

This case study suggests multiple lines of evidence can converge, but that the uncertainties associated with the respective dating must be near that of the potential forcing being tested. Further we point out that if our sampling was restricted to the moraines, these insights would not have emerged.