Climate Audit

In my search for high-resolution ocean sediment records, I stumbled across an interesting 1995 article by Domack et al (Domack of the Larsen Ice Shelf) discussing cores on the west side of the Antarctic Peninsula which were dated over the last 2000 years.

Here is the abstract but some of the detail and some of the methodology prove even more intriguing.

The Muller Ice Shelf is the northernmost ice shelf on the west side of the Antarctic Peninsula, in Lallemand Fjord at 67S. They mention in passing that the nearby George VI Ice Shelf is “thought to have been non-existent during a mid-Holocene warm period, but has subsequently re-formed since 6000 BP (Clapperton 1990). ” The Larsen Ice Shelf is on the east side of the Antarctic Peninsula. Here is a location map of Muller Ice Shelf and cores 72 and 75, to which reference will be made. The cores were drilled in 1990-1991 (austral summer). They were Karsen cores, which were said to preserve the top portion much better than piston cores.

The Muller Ice Shelf had been receding since 1974 (as proven by air photos). Core 72 was drilled in an area formerly located beneath the leading edge of the Muller Ice Shelf, while Core 75 was drilled in an open marine environment. They found that the uppermost 60 cm of Core 72 was in an ice shelf facies (sandy mud), overlying an open marine facies (silt). Thus there had been a change in Core 72 depositional process, but not in Core 75.

They discuss the benthic foraminifera record in core 75, observing a decreased number of forams per gram above 80 cm, with a calcareous rich interval from 101-273 cm dominated by B aculeata and Fursenkoina spp. (both associated with “warm
CDW.)

Chronology
Their approach to chronology is very interesting – and the approach taken here needs to be thought about in other cases where we see cores dated entirely by C14 dates. They report that core-top C14 dates for cores 72 and 75 were 2260 BP and 2590 BP respectively. (One notices this sort of date on many occasions in paleoclimate cores – not just ocean sediments – think about some of the peats in pro-glacial lakes in the Andes). They note that these ages are about 1000 years older than the then generally accepted reservoir age for C14 in the area.

They also carried out Pb-210 analyses on the upper 70 cm of both cores from which they estimated sedimentation rates in the upper 70 cm of the cores to vary from 2.1 mm/year (core 72) to 1.7 mm/year (core 75). (For comparison, rates in the Arabian Sea and Sargasso Sea cores – ~ 11 cm/1000 years in the Arabian Sea and ~ 19 cm/1000 years in the Sargasso Sea.- are about one-tenth of this). Rates based on C14 dates are 1.2-1.3 mm/year. So a 2 cm interval here meets the 2 mm standard that I was looking for.

Their conclusion was that the core 72 (150 cm) represented the past 1250 years and core 75 (300 cm) the last 2100 years respectively.

Results
Now they didn’t do analyses on 2 cm intervals, but the resolution was sufficient to permit a number of very interesting conclusions. They stated that the “sedimentology reflected a relatively recent advance of the Muller Ice Shelf”, which they date to ~400 years ago, noting that it is “correlative with the onset of the LIA, a widely recognized NH neoglacial event.” and that “advance of glacial margins during the LIA is also recognized from other SH high latitude locations [Clapperton 1990]”.

They go on to state:

The high abundance of B. aculeata and Fursenkoina spp throughout the lower section of core 75 indicates the presence of highly productive warm CDW occupying Lallemand Fjord to ~700 BP. The abundance of plankronic foraminifera N. pachyderma, although low, exceeds its occurrenes in modern sediments on the Bellingshausen SEa margin of the Antarctic Peninsula indicating marine conditions even more open than today… A decrease in the abundance of B aculeata from 84-41 cm (625-315 BP) and significant increase in the abundance of Nonionella spp indicate a progressive cooling of bottom waters, perhaps responsible for the reduced abundance of calcareous benthic foraminifera in the upper interval of core 75.

They suggest that it was an absence of CDW that contributed to the advance of glacial margins in Lallemand Fjord in the LIA, while today Lallemand Fjord is filled with CDW with a temperature of greater than 1 deg C. They note an inconsistency with Thompson ice cores at Siple Dome, which indicate warmer temperatures in the LIA. They suggest that this evidence is inconsistent with the Lallemand Fjord evidence, as “even a modest amount of warming would not be consistent with the advance of the Muller Ice Sheet” reported in this article for the LIA. They speculate that the Siple ice core data may not be representative of the rest of the Antarctic Peninsula and that increased accumulation accompanied by greater storm frequency may be another important factor to keep in mind.

It’s interesting to compare this approach to chronology to the approach taken in Domack’s famous recent article in Nature about the recession of the Larsen Ice Shelf, where he simply presents uncorrected C14 dates – no reservoir adjustment, no nothing. I’ve seen some evidence of a warm Holocene Optimum in other Antarctic cores, whcih I’ll try to get to. So it would be worth double-checking the dating of the formation of the Larsen Ice Shelf.

Reference: Domack et al 1995 Late Holocene advance of the Muller Ice Shelf, Antarctic Peninsula, Antarctic Science 7, 159-170. http://journals.cambridge.org/article_S0954102095000228