Synchronizing ice-core and U / Th timescales in the Last Glacial Maximum using Hulu Cave 14C and new 10Be measurements from Greenland and Antarctica

Between 15 and 27 kyr b2k (thousands of years before 2000 CE) during the last glacial, Greenland experienced a prolonged cold stadial phase, interrupted by two short-lived warm interstadials. Greenland ice-core calcium data show two periods, preceding the interstadials, of anomalously high atmospheric dust loading, the origin of which is not well understood. At approximately the same time as the Greenland dust peaks, the Chinese Hulu Cave speleothems exhibit a climatic signal suggested to be a response to Heinrich Event 2, a period of enhanced ice-rafted debris deposition in the North Atlantic. In the climatic signal of Antarctic ice cores, moreover, a relative warming occurs between 23 and 24.5 kyr b2k that is generally interpreted as a counterpart to a cool climate phase in the Northern Hemisphere. Proposed centennial-scale offsets between the polar ice-core timescales and the speleothem timescale hamper the precise reconstruction of the global sequence of these climatic events. Here, we examine two new Be-10 datasets from Greenland and Antarctic ice cores to test the agreement between different timescales, by taking advantage of the globally synchronous cosmogenic radionuclide production rates. Evidence of an event similar to the Maunder Solar Minimum is found in the new Be-10 datasets, supported by lowerresolution radionuclide data from Greenland and C-14 in the Hulu Cave speleothem, representing a good synchronization candidate at around 22 kyr b2k. By matching the respective 10Be data, we determine the offset between the Greenland ice-core chronology, GICC05, and the Antarctic chronology for the West Antarctic Ice Sheet Divide ice core (WDC), WD2014, to be 125 +/- 40 years. Furthermore, via radionuclide wiggle-matching, we determine the offset between the Hulu speleothem and ice-core timescales to be 375 years for GICC05 (75-625 years at 68% confidence) and 225 years for WD2014 (25-425 years at 68% confidence). The rather wide uncertainties are intrinsic to the wiggle-matching algorithm and the limitations set by data resolution. The undercounting of annual layers in GICC05 inferred from the offset is hypothesized to have been caused by a combination of underdetected annual layers, especially during periods with low winter precipitation, and misinterpreted unusual patterns in the annual signal during the extremely cold period often referred to as Heinrich Stadial 1.