The data and sample gathering phase of the Windless Bight project has now been successfully completed. Some results are immediately apparent and are summarised below. Others require further data reduction and analysis. Roles and responsibilities for the field phase just completed are shown in Table 1.

Camp was set up on the first site at the intersection of seismic lines MIS-1 and MIS-2 on January 3. However, because the site was too close to one of the approaches to a runway on Williams Field it had to be moved 1.75 northward along the seismic line to 77° 53 308′S; 167° 17.753′E and was designated HWD03-1 or Site 1 (Fig. 1). At the same time a Broadband ADCP current meter was installed in sea ice at the edge of the ice shelf south of Scott Base (77° 52.773″S; 166° 50.042′E) to record currents to 400 m depth continuously over the following 3-4 weeks. The hole at Site 1 was drilled on January 11 and after 4 attempts successfully reamed finally on January 12 (midnight) to a diameter of > 0.56 m throughout. Measurements and sampling through the hole took place from January 13 to 22. On January 23 and 24 the camp was shifted 7 km northeast to the second site to be occupied (HWD03-2 or Site 2). The access hole was drilled on January 26, reamed finally on January 27 (22:00) and kept open until February 2 for measurements and sampling. Camp and equipment were returned to Scott Base and the field operation completed by February 4.

OCEANOGRAPHIC AND SEA FLOOR MEASUREMENTS

The first measurements at Site 1 were to establish water depth and nature of the sea floor. A 3.5 kHz transducer was lowered and set 2 m below the base of the ice shelf to obtain a high quality acoustic record of the sea floor and subsurface sediment. Water depth was determined to be 855 m below the base of the ice shelf (908 m bsl, 926 m below the ice shelf surface, a convenient reference point for all subsequent oceanographic measurements). The sea floor reflector was sharp and stratification recognized down to a depth of over 300 m. This indicates that the sediment for this interval is largely fine-grained and unconsolidated. The water depth was determined by weighted line to be 938 m, 12 m deeper than the acoustic estimation. This depth was used for operational purposes on oceanographic casts. A similar procedure was followed for site 2, and an outline of the scientific data gathered for both sites is shown in table 3.

Fig. 4. Seismic record through Site 1 (MIS-1 line, Bannister et al., 2002), with image from 3.5kHz profiler inserted. Reflectors can be seen down to ~300 m below the sea floor.

Cast procedure and water/particulate chemistry samples: After a trial cast with the S4 1 m above the weighted end of the line and the CTD 5 m above that, a second cast was run with 1 litre NIO bottles at depths at 6 levels through the water column for water sampling. Several bottles proved difficult to bring up through the hole because currents in the water column deflected the rope, and were damaged. It was then decided to attach only two bottles 5 m apart and 5 m above the CTD on each cast but to trip them each cast at successively higher levels to cover the 6 planned sampling levels. This worked well, and the 14 casts were completed late on January 16. Filtration was done on samples from each level for 3 casts with a fourth close to the sea floor, and 3 further samples from water in the gravity corer just above the sediment water interface. Material could be seen on all filters, with those close to the sea floor showing an obvious brown coating. Samples were also taken for water chemistry.

CTD measurements: Oceanographic data were collected at both sites primarily with a Conductivity-Temperature-Depth profiler during water sampling casts, and later by a moored array of Acoustic Doppler Current Profilers (ADCPs). The setup and some results are shown in figure 5.

Zones B, C and D resemble High Salinity Shelf Water (HSSW) and Deep Ice Shelf Water (DISW) of Jacobs et al. (1985), whereas Zone A resembles Shallow Ice Shelf Water (SISW). Site J9, 450 km south of the Ross Ice shelf, has a similar Temperature/Salinity structure suggesting continuity of water masses with Windless Bight. However, the shallowest zones are much colder at J9 and lack the relatively warm intrusion (Zone A) seen in the Bight.

Fig 5. Oceanography beneath the McMurdo-Ross Ice Shelf. Site locations are shown in figure 1.

Current velocity measurements: Current measurements from the ADCP arrays are heavily influenced by the diurnal tides (Fig 5). Currents are relatively slow in the floor of the basin at both sub-ice shelf sites (averaging around 7 cm/s with a maximum of 17 cms), but are considerably faster (up to 60 cm/s) at the ice shelf edge off Hut Point Peninsula where the water is shallower (~600 m). Flood tides flow to the E and swing S and W below the McMurdo Ice shelf. This is also the direction of the mean flow, suggesting that, like the western and central Ross Ice Shelf edge, McMurdo Sound may also be a point of sub-ice shelf inflow.

Sea floor sediments: Initial gravity coring attempts yielded cores only a few cm long but by applying grease to the inside of the core liner cores in excess of 50 cm were consistently recovered page 8 from both sites. Grab samples of the top 2-3 cm were also recovered from both sites, and showed that the sea floor sediment at each comprises a thin (5 mm) layer of sandy mud with scattered basaltic pebbles up to 15 mm across. The sediments beneath are diatomaceous sandy mud with diverse diatoms, occasional foraminifera and rare shells. The mud gives way at 31 cm at Site 1 and 60 cm at Site 2 to a pebbly sandy mud or diamicton beneath (Fig. 6). Cores from Site 1 have from 23 to 30 cm an unusual laminated well sorted sand with a sharp base and top, tentatively interpreted as a sediment gravity flow. The diamicton in the lower part of each core was firm but not over-consolidated, indicating that the basal ice that deposited the sediment did not load or erode the sea floor.

The sand mineralogy is basaltic glass and rock fragments throughout both cores, but they also have a significant proportion of quartz, some of it rounded like grains in the Beacon sandstone from the Transantarctic Mountains 100 km to the west. Smear slides also show a trend of increasing biogenic silica up the core. While further work is needed these results are support the view that the cores represent a period of glacial retreat from a time of more extensive grounded ice, effectively recording ice retreat since the Last Glacial Maximum, a pattern consistent with cores elsewhere in the Ross Sea. A combination of field evidence and ice sheet modelling indicates that the last stage of the retreat to the present position of the ice margin on either side of Ross Island took place between 8000 and 4000 years ago (Kellogg et al., 1996)

SUMMARY

The lack of sediment compaction indicated by 3.5 kHz penetration to 300 m bsf shows that water was deep enough not to have grounded ice during the Last Glacial Maximum (or for previous such events for perhaps as much as a million years or more). These data indicate that the basin contains a long and continuous record of the presence, absence and (near) grounding of the Ross Ice Shelf over the last million years or more. One of ANDRILL's goals is to core this record.

Further work on the oceanographic data will help understand the origins and maintenance of oceanic circulation under the present global climate regime. It will also help in designing ANDRILL's deep coring system, with deployment planned for 2005-06.

REFERENCES

Bannister, S. and Naish, T.R. 2002. ANDRILL Site Investigations, New Harbour and McMurdo Ice Shelf, Southern McMurdo Sound, Antarctica. Institute of Geological and Nuclear Sciences Report 2002/01, 24 p.

Domack, E.W., Jacobson, E.A., Shipp, S. and Anderson, J.B. 1999. Late Pleistocene-Holocene retreat of the West Antarctic Ice Sheet system in the Ross Sea: Part 2 – sedimentologic and stratigraphic signature. Geological Soc. America Bull., 111, 1517-1536.

Horgan, H., Naish, T., Bannister, S., Balfour, N., Wilson, G., Finnemore, M. and Pyne, A. 2003. Seismic stratigraphy of the Ross Island flexural moat under the McMurdo-Ross Ice Shelf, Antarctica. Immediate Scientific Report to AntarcticaNZ

Jacobs, S.S., Fairbanks, R.G. and Horibe, Y., 1985. Origin and evolution of water masses near the Antarctic continental margin: evidence from H₂¹⁸O/H₂1⁶O ratios in seawater, American Geophysical Union Antarctic Research Series, vol. 43, 59-85.

Joughin, I. and Tulaczyck, S. 2001. Positive mass balance of Ross Ice Streams, West Antarctica. Science, 295, 476-480

Kellogg, T.B., Hughes, T. and Kellogg, D.E. 1996. Late Pleistocene interactions of East and West Antarctic ice-flow regimes: evidence from the McMurdo Ice Shelf. Journal of Glaciology, 42, 486-499.

Lacy, L, Harwood, D. and Levy, R. (eds.), 2002. Future Antarctic Margin Drilling: Developing a Science Program Plan for McMurdo Sound. Andrill Contribution 1. University of Nebraska-Lincoln, Lincoln, NE xxx pp.

Meluish, A, Henrys, S.A., Bannister, S. and Davey, F.J., 1995. Seismic profiling adjacent to Ross Island: constraints on Late Cenozoic stratigraphy and tectonics. Terra Antartica, 2, 127-136.

Nixdorf, U., Oeter, H. and Miller, H., 1994. First access to the ocean beneath Ekströmisen, Antarctica, by means of hot-water drilling. Annals of Glaciology, 20, 110-114.

Fig. 6. Sediment deposited beneath the McMurdo Ice Shelf in the last 8000 years or so.