Description: The Amundsen Sea continental shelf is one of the remotest areas of coastal Antarctica, and was relatively unexplored until the late 1980s. Over the last two decades, however, it has increasingly become the focus of a strong interest in potential West Antarctic Ice Sheet (WAIS) sensitivity to warming. One concern is that a change in the ocean circulation and/or properties, aided by deep glacial troughs in the sea floor, could be responsible for the reported thinning ice shelves and accelerating glaciers in this sector. Oceanographic and geological interest has led to several cruises, one result of which is that there is now sufficient bathymetric data to compile a fairly detailed regional map of the Amundsen continental shelf. We have combined the available multibeam and single beam bathymetry data from this region and created a new regional bathymetric map of the Amundsen Sea continental shelf and slope. We also integrated data from the Bedmap project and the Airborne Geophysical Survey of the Amundsen Sea Embayment, Antarctica (AGASEA) project for the surrounding land and ice-covered area. After cleaning the individual data sets we used a natural neighbor algorithm to interpolate between the existing data and create a grid at 5 km raster resolution. The most prominent regional feature is a series of separate trough systems along the inner shelf, which are aligned with present glaciers, separated by shallower ridges, and shoaling seaward. These troughs identify pathways of glacial flow during glaciations and now serve as conduits and reservoirs for warm Circumpolar Deep Water that contributes to high melting rates near ice shelf grounding lines.

Description: Marine geoscience data indicate that during the Last Glacial Maximum (LGM) grounded ice extended to the shelf edge along most, if not all, of the 2500 km-long continental margin from the northern Antarctic Peninsula to the Amundsen Sea. Past extent of grounded ice is indicated by swath bathymetry data from the outer parts of cross-shelf troughs, which reveal relict elongated subglacial bedforms. The bedforms show that the troughs were paths of fast-flowing (streaming) ice. Geomorphological evidence regarding the nature of ice flow over intervening outer shelf banks has been erased through pervasive post-glacial ploughing by icebergs. However, seismic profiles across the banks reveal widespread shelf edge progradation and numerous glacial unconformities that indicate grounded ice has extended across them many times during the Pleistocene, and before. Subglacial tills in the outer parts of shelf troughs are overlain by up to 2 m of postglacial sediments, which are no older than the LGM in any core yet dated. A layer of soft, intermediate shear strength (12¬25 kPa) till, interpreted as deformation till, underlies the postglacial sediments in cores in the troughs. These observations are consistent with the interpretation that streaming ice extended along the troughs during the LGM, but the duration of such flow, and whether or not it spanned the entire period when ice extended to the outer shelf remains undetermined.To determine when, and how rapidly, ice retreated from the continental shelf, ages of core samples from near the base of postglacial sediments in several troughs have been determined by AMS radiocarbon dating. Samples to constrain glacial retreat have been taken from either the base of muds deposited in seasonally open-marine conditions similar to today, or underlying sandy muds interpreted as having been deposited close to the grounding line. Modern sea-floor sediments on some parts of the margin contain sufficient calcareous microfossils for dating to constrain the local marine 14C reservoir correction. However, even where they occur, contents of planktonic foraminifera decrease downcore, and most deglaciation ages have been obtained from acid insoluble organic material (AIOM). In some areas these ages are significantly affected by reworked fossil carbon, as shown by apparent ages from AIOM in modern sea-floor sediments that range up to ~6000 years. Thus radiocarbon results from this margin must be treated with caution and there is a clear need for development of alternative dating methods.Notwithstanding these uncertainties, deglaciation ages obtained thus far suggest variable retreat histories along the margin. Results from the Antarctic Peninsula shelf and Amundsen Sea embayment suggest relatively rapid post-LGM ice retreat from the outer and middle shelf, followed by slower Holocene retreat to the present day ice margin. However, initial results from the Bellingshausen Sea (Belgica Trough) suggest a slower, progressive retreat commencing about 25 ka (corrected radiocarbon years). These results show that local factors are important in controlling the rate of ice retreat, and this needs to be taken into account in numerical models that attempt to predict the dynamic behaviour of large ice sheets.

Description: The Amundsen Sea sector of the West Antarctic Ice Sheet (WAIS) is the most rapidly changing part of the Antarctic ice sheet and could have a significant impact on future sea level rise. However, sea level rise predictions in the recent Intergovernmental Panel on Climate Change Summary for Policymakers excluded the possible effects of future rapid dynamic changes in ice flow because ice sheet model predictions were not considered to be sufficiently reliable.One approach to testing and refining ice sheet models would be to examine their ability to reproduce ice sheet changes since the Last Glacial Maximum (LGM). Records of ice margin retreat and ice surface elevation change since the LGM also provide a context for recent changes. Until recently, however, knowledge of the chronology of change in the Amundsen Sea sector of WAIS since the LGM has been based on just seven radiocarbon dates from continental shelf sediment cores collected in 1999 on RV Nathaniel B. Palmer1. These dates indicated that there was already seasonally open water over the middle part of the shelf by 15,800 +/- 3900 radiocarbon years ago, and open water extended to within 100 km of the modern ice margin in Pine Island Bay before 10,150 +/- 370 radiocarbon years ago. Over the past 18 months we have obtained 34 new AMS radiocarbon dates on samples from 25 Amundsen Sea shelf sediment cores collected during research cruises in early 2006 on RRS James Clark Ross and RV Polarstern. Some dates are on carbonate (foraminifera) but most are on the acid insoluble organic fraction. Several dates are on modern surface sediment samples to evaluate the marine reservoir correction and the effect of reworked fossil carbon. We have also obtained the first surface exposure ages from the region by analysing cosmogenic isotopes in samples collected from sites accessed using helicopters operating from RV Polarstern. These ages provide the first data on long-term changes in surface elevation of ice in the Amundsen Sea sector of the WAIS.Our new radiocarbon dates, together with swath bathymetry data collected on the same cruises and some previous cruises2, confirm that the ice grounding line advanced to the continental shelf edge in the Amundsen Sea at the LGM. The retreat of the ice margin to its present position represents a loss of more than 150,000 km2 of ice sheet, i.e. more than 35% of the area that remains in the Amundsen Sea sector of the WAIS. Our new data are generally consistent with the timing of ice margin retreat suggested previously, and in the western part of the embayment most of the retreat to the present ice margin position was certainly complete by early Holocene time. Average rates of retreat and surface elevation change are more than an order of magnitude slower than those observed over recent decades, but we cannot discount the possibility that there might have been previous short-lived episodes of rapid change since the LGM.

Description: Understanding the past glacial history of regions undergoing potential rapid deglaciation is essential in order to estimate the possible threat of sea level rise. Recently acquired data have given new images of mega-scale glacial lineations on the sea floor of the Amundsen Sea, which provide us a new understanding of the direction of glacial flow on the continental shelf of the Amundsen Sea region. Two adjacent areas of seafloor on the outer shelf of the Amundsen Sea embayment exhibit remarkably different styles of glacial lineations, and allow the interpretation of a divergent glacial trough for the Pine Island Glacier during the last glacial maximum, whereas ice flow from the Abbot Ice Shelf probably converged with that from the Pine Island Glacier (PIG) to the north of a grounding zone wedge.

Description: The Amundsen Sea embayment lies between the Palaeozoic crustal blocks of Marie Byrd Land, Ellsworth Land and Thurston Island. Its continental margin is conjugate to the passive margin of the eastern New Zealand submarine continental plateaux and Bounty Trough, which underwent major extension during Cretaceous rifting between New Zealand and West Antarctica. Later, the embayment seems to have played a role as a plate boundary while the Bellingshausen Plate acted as an independent microplate until the early Tertiary. It is likely that the tectonic architecture, through the formation of deep basins and erosional troughs, laid the foundation for major glacier outflow from the West Antarctic Ice Sheet into Pine Island Bay and the South Pacific since early West Antarctic glaciation.During successive cruises on RRS James Clark Ross (cruise JR141) and RV Polarstern (expedition ANT-XXIII/4) in early 2006, we collected seismic, bathymetric, sub-bottom profiler and helicopter-magnetic data from the inner shelf, outer shelf, slope and deep sea of the Amundsen Sea embayment and Pine Island Bay, to address tectonic as well as sedimentary objectives. We will present preliminary results indicating the disposition of structural basement units and the role the tectonic basement and pre-glacial sediments have played as controlling parameters for ice-sheet expansion and retreat in the Amundsen Sea Embayment.

Description: Remarkably little is known about the Cretaceous rifting process between New Zealand and Antarctica as well as within the submarine parts of the microcontinent of New Zealand itself. The Bounty Trough offers good insights into these break-up processes. Here we present results from a combined gravity, multichannel seismic and wide-angle reflection/refraction seismic transect across the Bounty Trough and interpret this on the basis of velocity distribution and crustal composition derived from Poisson's ratio and P-wave velocity. The lower crust exhibits a high-velocity (vp "ca." 7 - 7.7 km/s, vs "ca." 3.9 - 4.5 km/s), high-density body ("rho" "equals" 3.02 kg/cm³) at the location of the most thinned crust of the Bounty Trough. In this part, crustal thickness is reduced from 22 - 24 km beneath Chatham Rise and Campbell Plateau to about 9 km. We interpret this high-velocity/density body as a magmatic intrusion into a thinned continental crust. Our results show that the Cretaceous rifting of the Bounty Trough is very likely not the result of back-arc extension caused by the Hikurangi Plateau subduction in the Gondwana margin, but of continental break-up processes related to the separation of New Zealand from Antarctica. Rifting ceased at the onset of seafloor spreading, so that only little oceanic crust was produced in the Middle Bounty Trough. Comparisons with the Oslo Rift and the Ethiopian/Kenya Rift indicate analogue systems and imply a stretching model that combines uniform stretching and simple shear extension.

Description: The Campbell Plateau and Chatham Rise are large submarine plateaux of continental origin forming parts of the submarine New Zealand continent. Prior to the break-up of this part of Gondwana, New Zealand was situated at the proto-Pacific plate boundary of Gondwana, connected to Marie Byrd Land. It is expected that the development of the continental fragments forming Campbell Plateau and Chatham Rise played a key role in the development of Gondwanas plate boundary from a convergent margin to continental rifting. Our new crustal models of Bounty Trough and Great South Basin infer thinned crust beneath both basins. The crust beneath the Bounty Trough is extremely thinned up to nascent seafloor spreading. Seismic information implies that several extensional phases and styles (pure shear and simple shear) have occurred. Beneath the Great South Basin, the crust is less thinned and underplating can be observed in some areas. Our models as well as geologic information suggest, that an initial extension of the Campbell Plateau predates the Great South Basin opening in Cretaceous time. This information related to the magnitude and style of rifting along Bounty Trough and Great South Basin, influence models of the break-up process between New Zealand and Antarctica that will be presented here.