PP11C-01 INVITED 08:00h
What can the ice core record tell us about high frequency climate changes?
Ice cores show climate change on all frequencies from annual up to Milankovich frequencies, and the record now extends back to 800 kyr (but with much less time resolution in the longest records). In this paper, I will briefly review the different periods that can be observed in ice cores. The new European Project for Ice Coring in Antarctica (EPICA) Dome C core shows a clear shift in the pattern of climate about 400,000 years ago, with much warmer interglacials from MIS11 to the Holocene than were observed in MIS19 to 13. The record also shows the familiar 100 kyr cycles that dominate late Pleistocene climate. Within the last glacial period, we are all familiar with the millennial scale cycles (Dansgaard-Oeschger events) observed in Greenland cores, and we have the opportunity to test whether these are common to earlier glacials by looking for their subdued Antarctic counterparts, and for their accompanying methane excursions in the Antarctic records. It is harder to identify clearly millennial scale cycles in the Holocene, but the new EPICA core gives us a wide range of interglacials to investigate further. Finally, some ice core indicators in short, high resolution records may be able to show the persistence of very short period fluctuations such as ENSO, and I will briefly discuss the prospects for this. All of these frequencies are of course high frequency in the geological context! The paper will draw on work by many ice core colleagues
PP11C-02 08:15h
Late Quaternary Biosiliceous Laminated Marine Sediments From Antarctica: Seasonality During a Period of Rapid Climate Change
The Antarctic ice sheet plays a key role in global oceanic and atmosphere systems. One of the most dynamic regions of the continent is the Antarctic Peninsula (AP) where ecological and cryospheric systems respond rapidly to climate change, such as the last deglaciation ($\sim$12-13 kyr BP). Here, deglacial laminated diatom-rich marine sediments are well known, e.g., Palmer Deep (64$\deg$S 64$\deg$W; ODP Hole 1098A) comprising a distinctive 3 m thick sequence of deglacial `couplet' laminations. The East Antarctic margin (EAM), however, has received less attention than the West Antarctic margin (WAM) in palaeoceanographic studies yet its role in deep ocean circulation and, therefore, the global ocean system is significant. Recent sediment cores recovered from EAM sites during NSF Polar Programs-funded cruise NBP0101 in February and March 2001 (e.g. Mertz Drift \{66$\deg$S 143$\deg$E\}, Svenner Channel \{69$\deg$S 77$\deg$E\} in Prydz Bay, Nielsen Basin \{67$\deg$S 66$\deg$E\} and Iceberg Alley \{67$\deg$S 63$\deg$E\}), reveal that a similar sedimentary facies was deposited along the EAM, in similar geomorphological settings to Palmer Deep, during the same timeframe. These rich sediment archives reveal clues about circum-Antarctic palaeoceanographic change during the last deglaciation, a time of both high silica flux and rapid climate change. Microfabrics and diatom assemblages from scanning electron microscope backscattered and secondary electron imagery analysis of coeval deglacial varves from Palmer Deep (WAM), Mertz-Ninnis Trough and Iceberg Alley (EAM) are presented and compared. The varves from these localities are characterised by laminae to thin beds of orange-brown diatom ooze up to $\sim$8cm thick alternating with blue-grey diatom-bearing terrigenous sediments up to $\sim$4cm thick. The orange-brown oozes are dominated by resting spores and vegetative valves of {\itHyalochaete Chaetoceros} spp., resulting from spring sedimentation associated with stratified surface waters promoting exceptionally high primary productivity. The blue-grey terrigenous sediments contain mixed open-water Antarctic diatoms resulting from summer/autumn sedimentation associated with increased terrigenous input and the annual trend to more oligotrophic conditions. The significance of these biogenic seasonal and annual cyclicities will be discussed in terms of interannual palaeoclimate oscillations. Sub-seasonality and a related deglacial trend are apparent within the spring lamine from the EAM and in the summer/autumn laminae from the WAM. This circum-Antarctic asymmetry indicates differences in the nature and timing of oceanographic and cryospheric systems operating around the Antarctic continent at this time.
PP11C-03 INVITED 08:30h
El Niño, Past And Future
Abstract I review forecasts of the future of El Niño and the Southern Oscillation (ENSO), a coupled instability of the ocean-atmosphere system in the tropical Pacific with global impacts. The physics of the ENSO cycle is briefly discussed, with special attention to the Bjerknes feedback, the instability mechanism which figures prominently in ENSO past and future. Our knowledge of ENSO in the paleoclimate record has expanded rapidly within the last 5 years. The ENSO cycle is present in all relevant records, going back 130 kyr. It was systematically weaker during the early and middle Holocene, and model studies show that this results from reduced amplification of anomalies in the late summer and early fall, a consequence of the altered mean climate in response to boreal summer perihelion. Data from corals shows substantial decadal and longer variations in the strength of the ENSO cycle within the past 1000 years; it is suggested that this may be due to solar and volcanic variations in solar insolation, amplified by the Bjerknes feedback. There is some evidence that this feedback has operated in the 20th century and some model results indicate that it will hold sway in the greenhouse future, but it is very far from conclusive. The comprehensive general circulation models used for future climate projections leave us with an indeterminate picture of ENSO's future. Some predict more ENSO activity, some less, with the highly uncertain consensus forecast indicating little change.
PP11C-04 08:45h
El Nino Activity During MIS 5e in Peru
Oceanography and climate along the coast of Peru is fundamentally linked to ENSO variability with stronger lithic flux into the sea and reduced marine bioproductivity during El Nino events. A 19 m long piston core with laminated marine sediments has been recovered on the edge of the Peruvian shelf (12 03'S, 77 40W, 184 m waterdepth) during cruise Sonne-147 in 2000. We present the lower 6 meters of this core that cover the time between 100 to 130 kyr before the present (BP). SST has been estimated from alkenone analysis with a mean temporal resolution of 300 years. Color logging along the core at 2 mm intervals revealed high resolution proxy data (3 to 30 a) for the precipitation on the continent (fine-grained lithics) and marine bioproduction (photosynthesis pigments: chlorines, carotenoids). Proxy data show that a major change occurred around 123 kyr BP. Mean sedimentation rate which is largely controlled by lithics dropped from 40-70 cm/kyr to 18 cm/kyr after 123 kyr BP. Contemporaneously SST start to decline towards the early Glacial level that is reached between 118 to 116 ka during the time of glacial inception. We conclude that stronger El Nino floods occurred before 123 kyr BP and El Ninos were weaker during the second half of MIS 5e. The change to weaker ENSO activity in Peru is therewith roughly contemporaneous with the beginning cooling in Greenland.
http://www.uni-mainz.de/FB/Geo/Geologie/sedi/index.html
PP11C-05 09:00h
The End of The Last Interglacial in Central Europe: Evidence From Varve Counted Maar Lake Sediments in The Eifel/Germany
We present in this paper evidence from 14C and OSL dated laminated Eifel maar lake records which cover the time from 20-140 ka. The greyscale variations in these cores show all MIS3- North Atlantic/Greenland interglacials, which allows tuning and direct comparison with the NorthGRIP ice core. The longest core reaches MIS6 and includes the last interglacial, which is represented by abundant interglacial tree pollen. Varve counting and parallel pollen analysis reveal a fully developed interglacial temperate forest from 126-112 ka. The onset of glaciation in North America and a 50 m global sea level drop started, however, already at 118 ka, which implies that the later part of the central European interglacial is contemporaneous with the North American glaciation. Abrupt cooling and a sharp climate change to a shrub tundra with frequent loess storms occurred in the Eifel/Central Europe during the C24 event at 110 ka thus 8000 years after the first ice sheet growth started at a summer insolation near 450 Wm2 at 65N, thus 50 Wm2 lower than today. Thus, abrupt climate change in central Europe occurred not before the time of the first surge of the Laurentide glacier and is most probably caused by rapid cooling of the sea surface temperatures in the North Atlantic drift (Gulf stream). The very end of the last interglacial in central Europe thus occurred at boundary conditions very different than today and the abruptness of the end of the Eemian is no analogue for the climate evolution of the next centuries.
http://www.uni-mainz.de/FB/Geo/Geologie/sedi/index.html
PP11C-06 INVITED 09:15h
Timing and Duration of Terrestrial Interglacials in SW Portugal: Direct Land-Sea Comparisons
While a broad correspondence between marine and terrestrial stages has been noted for some time, the lack of sufficiently precise terrestrial chronologies impedes the assessment of phase relationships between changes on land and variations in global ice volume. One way around this problem is to examine marine sequences containing a reliable vegetation signal. The linking of land and sea records through joint pollen and marine proxy analyses from the same samples allows an in situ assessment of phase and amplitude relationships, bypassing correlation issues. The Portuguese margin, where the combined effects of major river systems and narrow continental shelf lead to the rapid delivery of terrestrial material to the deep-sea environment, has in recent years emerged as a prime research area for linking marine and terrestrial records. In particular results from core MD95-2042 have led to a major revision of our understanding of the timing and duration of the last interglacial in southern Europe (Shackleton et al., 2002 Quat. Res. 58, 14). This new scheme proposed that the onset of interglacial conditions on land occurred inside MIS 5e and that forest persisted in southern Europe for 16 kyr, well into the interval of ice growth (MIS 5d). Here we attempt to establish whether similar (i) diachroneity between marine and terrestrial stages and (ii) patterns of extended terrestrial interglacial duration are observed during earlier periods. Foraminiferal oxygen isotope and pollen analyses were undertaken on the interval 190-350 ka from a new core MD01-2443, retrieved in the same area as MD95-2042. Contrary to expectations, what emerges is that the duration of terrestrial temperate stages does not follow any consistent pattern, with tree populations sometimes declining well before ice inception. The pollen record reveals the occurrence of abrupt changes during intervals of low ice volume. What emerges is that while the broad timing of interglacials is consistent with orbital theory, their specific duration appears to be dictated by millennial variability.
PP11C-07 09:30h
Are Millennial Oscillations a Feature Inherent to Interglacials in the North Atlantic-Arctic System?
Several paleoceanographical proxies (sedimentological and micropaleontological) in interglacial sediments of the northern North Atlantic provide clear indications of millennial-scale oscillations notably during isotopic stages 1 and 5e (Bond et al. Science, 2001; Hillaire-Marcel et al. Nature, 2001). However, it is not obvious that these oscillations are widespread beyond this region. Also unclear is whether these oscillations are synchronous, even within the North Atlantic domain. Actually, the most significant paleoceanographical results suggest that the oscillations consist of expanding sea-ice cover and/or pulses of lower salinity in surface waters. Analyses performed with a centennial time-resolution demonstrate that such oscillations are out of phase between records off the eastern margin of Canada (Labrador Sea) and those of the central North Atlantic (south of Iceland). The analyses even show opposite long term trends from the west to the east, with surface salinity increasing along the Canadian margins from the base to the top of the Holocene sediments, whereas it decreases in the central North Atlantic. Studies performed in the Holocene sediments raised from the lower slope of the western Arctic (Chuckchi Sea) also show millennial oscillations. Moreover, the analyses combining proxies for surface water temperature, salinity and sea-ice cover (dinocysts) with tracers of sub-surface to intermediate and bottom waters (isotopes in planktic and benthic foraminifers) tend to illustrate changes in the rate of intermediate water inflow from the eastern North Atlantic that are opposite to those in surface waters, which contribute to sea-ice and freshwater export to the western North Atlantic. For example, the early Holocene prior to 8 ka has been characterized by maximum inflow of warm North Atlantic water which is consistent with optimal conditions in the eastern North Atlantic, whereas it was marked by maximum sea-ice in the western Arctic, which is consistent with the maximum freshwater export (surface salinity minimum) recorded in the western North Atlantic. In this scenario, the increased heat flux to the Arctic, accompanied by more abundant precipitation over the Russian Arctic and freshwater flow over the Arctic shelves would have fostered formation of sea ice exported with the TransPolar Drift into the northwest North Atlantic. The above observations permit comparison with the modern Arctic Oscillation (AO-NAO) interdecadal variability, the AO+ situation being marked by stronger meridional transfer of heat to the east and increased southward sea-ice and freshwater export to the west. In turn, the freshwater-ice export from the Arctic plays a role on surface salinity and pressure gradients in the North Atlantic, and may thus be determinant on the strength and location of the meridional overturning. In conclusion, it seems that through self-oscillating mechanisms notably, the Arctic sea ice cover would play a major role on sea-surface variations over the northern North Atlantic during interglacial stages.
PP11C-08 09:45h
The Expression of the 8.2 ka and Younger Dryas Events in the Eastern Canadian Arctic
The two largest climate coolings following the end of the last glaciation are the Younger Dryas and the 8.2 ka events. Evidence for these cold excursions is widespread around the North Atlantic and in more distant regions. Both events are well expressed in Greenland ice cores; glacier readvances occurred across much of NW Europe during the Younger Dryas and cold surface waters returned to the North Atlantic, with depressed summer temperatures in eastern North America. The 8.2 ka event has a similar pattern, but the magnitude is substantially lower, with a much shorter duration. However, surprisingly little evidence has been presented for either event from the North Atlantic Arctic. Recently acquired lake sediment records from the Eastern Canadian Arctic contain evidence for both excursions. The 8.2 ka event is recorded at two sites as a significant glacier readvance of cirque and outlet glaciers of local ice caps at 8.2±0.1 ka. In some non-glacially-dominated lakes, a reduction in primary productivity is apparent at the same time. These records suggest colder summers without a dramatic reduction in precipitation, producing positive mass balances and glacier readvances. For most local glaciers, this was the last significant readvance before they receded behind their Little Ice Age margins. Only a few lakes contain records that extend through the Younger Dryas chron. The best-dated lake record, Donard Lake, extends back to 15 ka. Lacustrine sedimentation is currently dominated by a meltwater from an outlet glacier that terminates a few hundred meters from the lake. The glacier has been within the drainage basin of the lake for the past 5.5 ka, although the contribution of glacial sediment has been larger since about 2.5 ka. Prior to 5.5 ka, there is no evidence of a glacier in the catchment of Donard Lake, suggesting that throughout the entire Neoglacial period, the local glacier has been more advanced than at any time since 15 ka. During the Younger Dryas chron, lacustrine primary productivity was greatly reduced, whether measured as the flux of organic carbon to the lake floor or as the percentage of organic matter in lake sediment. We interpret this change to reflect a substantial reduction in summer temperatures for more than 1 ka. However, this temperature drop was not accompanied by a significant glacier readvance, suggesting precipitation must have been very low. This differs from the 8.2 ka event when precipitation must have remained relatively high. These records indicate that in the Eastern Canadian Arctic, summers during the Younger Dryas were much colder than present, but precipitation was dramatically lower too, so glaciers did not advance, whereas during the briefer, and less severe summer cooling associated with the 8.2 ka event, precipitation was not dramatically reduced and glaciers readvanced.