A Warmer and Earlier High Arctic Holocene Thermal Maximum from Agassiz Ice Cap Core , Ellesmere Island
A Warmer and Earlier High Arctic Holocene Thermal Maximum from Agassiz Ice Cap Core , Ellesmere Island
Abstract ID#: 34409
English Abstract:
The Holocene Thermal Maximum (HTM) in the Arctic occurred during the first half of the present interglacial and represents a period when air temperatures were warmer than present-day. The melt-layer and oxygen-18 record from Agassiz Ice Cap (Ellesmere Island, NU) represents the most northerly continuous climate record that can quantify temperature changes throughout the Holocene. Therefore, these records have a pivotal role in our understanding of recent High Arctic climate change (Fisher et al.,1995). However, estimates of temperature changes from these two Agassiz records have yielded inconsistent results: the melt record indicates a well-defined peak in temperatures around 11ka (Fisher et al.,2012), which is compatible with other proxy records from the region, whereas the O-18 time-series (Vinther et al.,2009) indicates a maximum temperature stretched over the 10 to7 ka period. Here, we present an new elevation corrected O-18 and melt temperature reconstructions of the Agassiz Ice Cap that resolves this apparent discrepancy between the two records. We also extend the O-18 time-series to present-day. Our corrected O-18 record now exhibits an earlier and warmer peak temperature of +4.5 C at 10.5ka (2σ uncertainty 3.4 – 5.3 C) and includes temperatures up to 6.5 C higher during the early Holocene (11.7 - 9ka) compared ~ +3C from previous studies (Vinther et al., 2009). Additionally, the updated O-18 time-series from the new shallow Agassiz ice core suggests that present-day air temperatures are at their warmest in 7.1–7.8ka.
The warmer and earlier HTM in our results has implications on the evolution of the northern Greenland Ice Sheet, where a revised model reconstruction finds a significantly enhanced thinning of this sector of the ice sheet which is compatible with observational constraints (Lecavalier et al., 2014).
