PLIO-PLEISTOCENE GLACIAL SEQUENCE CORED AT CIROS 2, FERRAR FJORD, WESTERN McMURDO SOUND
CIROS (K041) - P.J. Barrett and Scientific Staff.
CIROS in 1984 drilled one hole near the middle of Ferrar Fjord, western McMurdo Sound, in 211 m of water. A sequence of sand and glacial debris was cored (67% recovery) to basement gneiss at 166 m. A preliminary estimate of the age of the sequence, based on diatoms and the abundance of basaltic debris, has it ranging from Early Pliocene (about 4 m.y.) to the present, and equivalent to the upper 183 m of DVDP 10 and the upper 240 m of DVDP 11 in adjacent Taylor Valley. A good chronology is expected from the paleo-magnetic stratigraphy, diatom assemblages and radiometric dating of basaltic material, including a vitric tuff from 124 m sub-bottom.
The core has been subdivided into 13 lithologic units, representing alternations of "interglacial" and "glacial" conditions. The older interglacial units (13, 11 and 9) are diatomaceous mudstones, but the younger ones (7, 5, 3 and 1) are largely black basaltic sand, like that accumulating on the sea floor today. The oldest 2 glacial units (12 and 10) are basal lodgement tills with internal horizontal shearing and clasts of basalt and basement rocks, some of them striated. The younger glacial units (8, 6, 4 and 2) also contain scattered clasts, some striated, but have stratification features suggesting considerable redeposition and settling through water. Nevertheless, all of these units are considered to represent periods when ice was much more extensive than today. The abundance of basaltic debris in most glacial units suggests that ice flowed into Ferrar Fjord mainly from the east, eroding and transporting debris from the volcanic piles south and east of McMurdo Sound.
The CIROS 2 core should lead to a substantial improvement in both chronology and interpretation of glacial history in western McMurdo Sound over the last 4 m.y.
The aim of the CIROS project is to obtain a record of the Cenozoic (and possibly Cretaceous) history of the southwest corner of the Ross Sea by coring the sedimentary strata offshore (Barrett, 1982). There is no record exposed on land to represent the time between the Jurassic basalts 180 m.y. ago and the rocks of the McMurdo Volcanic Group, erupted over the last 15 m.y., though this interval includes two poorly documented events of wide interest and which may be related - the growth of the Antarctic Ice Sheet and the rise of the Transantarctic Mountains.
Seismic surveys over the last decade (Northey et al., 1975; Davey and Bennett, 1981; Wong and Christoffel, 1981),have revealed along the Victoria Land coast a sedimentary basin (Fig. 1) of probable post-Jurassic age, the margins of which are accessible to shallow offshore drilling. The structure of the western margin of the basin in South Victoria Land is shown in figure 2. The MSSTS 1 core has shown that the uppermost strata (to 226 m sub-bottom) are of marine glacial character (Barrett and McKelvey, 1981), and extend back to 30 m.y. (Harwood, 1984).
Figure 1. The Ross Sea region showing the location of the Victoria Land basin and the area of CIROS drilling (Fig. 2)
The second hole in the first phase (CIROS 2) was to be drilled as far landward as the ice in Ferrar Valley would permit. This hole was to core the sediment deposited on the valley floor to basement to work out the glacial history of the valley and also to obtain a minimum age on the cutting of the valley from the sediment just above the basement. In addition, the core was expected to contain reference planes to correlate with CIROS 1 and hence allow us to gauge the timing and extent of vertical movement within the fault zone between the two sites. The basement core was also of interest for it would be the lowest sample obtained to date for apatite fission-track dating, and hence provide the youngest possible point on the uplift curve for this section of the Transantarctic Mountains.
Figure 2. A. Map of McMurdo Sound area, showing the main physiographic features, the location of MSSTS-1, and deep DVDP drill holes (numbers). X-Y locates the section shown in figure 2B.
B. Geologic section across McMurdo Sound (line X-Y in figure 2A). Offshore structure extrapolated from Iles and Dibble (1981) and Wilson et al. (1981). Faulting has been inferred from topography (Webb and Wrenn 1982) but has in places been confirmed by field observations (P.G.Fitzgerald, pers comm).
Figure 3. Growth is sea ice thickness at CIROS 1 compared with that at nearby DVDP 15 in 1975, a normal year.
The Butter Point camp operated successfully throughout the operation (mid-September to mid-November, 1984). The camp and its operation are described in some detail in the Manager's report (Stephenson, 1985).
NOTE: Regular surveys of the CIROS 1 site from late September to mid-November show that it did indeed freeze 'solid' after the 6 m of movement in early October and remained in place through several major storms (Fig. 7). This shows that the site can be safely occupied at least until mid-November even in a 'thin ice' year. Had this data been available early in the season we would have probably elected to proceed with the drilling of CIROS 1, though the unexpected failure of the floats would have almost certainly led to termination of the hole before target depth.
CIROS 2 was successfully drilled near the middle of Ferrar Fjord in 210.7 m of water, through 165.5 m of sediment into basement gneiss (Table 2). Basement was found to be slightly deeper than the preferred interpretation of the available seismic data, and somewhat shallower than the estimate based on geomorphic inference (Fig. 8).page 6
Figure 4. Active crack in sea ice at the tip of the McMurdo Ice Shelf. The crack is 10m wide and the ice at this time (September 4, 1984) is 30cm thick near the margin and 5cm thick in the middle.
The sedimentary sequence is subdivided into 10 units (Fig. 9), representing alternations of "interglacial" and "glacial" conditions. The oldest interglacial units (13, 11 and 9) are thin (1 to 5 m) diatomaceous mudstones, like the muds accumulating in Granite Harbour today. Units 7, 5, 3 and 1 consist mainly of black sorted fine to medium-grained sand (Fig. 9B) similar to that on the sea floor around CIROS 2 today, where 3/4 of the sand grains are basaltic (Barrett et al., 1984). The sorting and the indistinct horizontal stratification with the occasional mud laminae indicate sedimentation by settling. However, the sand was probably derived ultimately from the McMurdo Volcanics to the east, glacially transported and deposited on the walls of Ferrar Valley and then blown by wind offshore. Both mud and sand units probably represent times when glacial ice was no more extensive than today.page 8
Figure 8. Cross-section of Ferrar Valley through the CIROS 2 site, showing water depth and 3 estimates of the geometry of the valley fill. 1 is from Burdelik (1981) and based on 3 seismic refraction profiles (shown by arrow heads) parallel to the valley axis. 2 is a reinterpretation of Burdelik's data by F.J.Davey (letter to P.J.Barrett, October 1982). 3 is a sketch by Barrett based on the valley slopes above sea level and on profiles across the east-trending Dry Valleys, all of which have their lowest point just north of the middle of the valley. Basement was encountered at 377m below sea level, between estimates 2 (330m) and 3 (430m).
The 6 even numbered units are extremely poorly sorted mixtures of mud, sand and gravel (diamict, Fig. 9A) with occasional striated stones, deposited directly or indirectly from glacial ice. They represent periods when ice cover around Ferrar Fjord was more extensive than at present. The oldest diamicts (Units 12 and 10) contain horizontal striated surfaces that are interpreted as subglacial shear planes, indicating that these units at least are lodgement tills. The other gravelly units, however, show in a number of places signs of redeposition by gravity flows or sedimentation through the water column, but probably accumulated close to the ice front.
The ice that transported the debris forming the diamict units came from one of two directions - west through the Transantarctic Mountains, or east past the volcanic piles of McMurdo Sound, and the debris should reflect this. Basaltic clasts are abundant (30 to 60%) in all diamicts but unit 8, indicating an easterly source. Several small basaltic cones of the order of 100 m across are known from upper Ferrar Valley, and are a potential source for some basaltic debris, but are tiny compared with the large area of exposed basement rocks. The proportion of volcanic debris in the sand fraction should help resolve this question.
Unusually well developed stratification occurs at several levels in the core between 7 and 80 m. It consists of sets of parallel mud laminae 1 to 3 mm thick in a well sorted fine sand (Fig. 9A). They superficially resemble glacial varves but the mud laminae are discrete, rather than part of sand-mud couplets. Also they lack outsized clasts. As yet we have no explanation for them.
The chronology of events recorded in the CIROS 2 core will depend on the results of current paleomagnetic, micropaleontologic and radiometric studies (Table 3). However, a preliminary examination of well preserved diatom assemblages from the lower 30 m of the hole indicates an Early Pliocene age (around 4 m.y.) (Table 4), The abundance of basaltic debris throughout most of the core and to the bottom of the hole is another indication of Plio-Pleistocene age, for basaltic debris appears in cores from DVDP 10 and 11 in the fill of adjacent Taylor Valley only above the late Miocene-Early Pliocene unconformity (Elston and Bressler, 1981; Porter and Beget, 1981). This unconformity and the sediments just above it are taken by Elston and Bressler to represent a significant glacial advance from the Ross Sea. The basalt-bearing lodgement till at the base of CIROS 2 may also represent this event.page 9
Figure 9. Stratigraphic column showing the major lithologic units in CIROS 2. Insets show the main facies: core width is about 45mm, top to left.
|A.||Mud laminae in well sorted fine sand from 7.06 to 7.19m.|
|B.||Black sand from 61 to 66m.|
|c.||Diamict from 158.67 to 162.18m.|
The sediment resting on basement rock at CIROS 2 was expected to be older, for it was the lowest point (377 m below sea level) from which valley fill has been recovered in the Dry Valleys region. The deepest prior to CIROS 2, DVDP 11 in Taylor Valley, ended in diamict 7 m.y. old (Elston and Bressler, 1981) at 268 m below sea level, though still 150 m above basement (Hicks and Bennett, 1981). The CIROS 2 sequence is also young compared with the MSSTS 1 core 30 km to the northeast (30 m.y. at 420 m below sea level; Harwood, 1984). Nevertheless, the CIROS 2 core will be of special value for the Pliocene history of the region because it is the only sequence of this period in which paleomagnetic zones and diatom assemblages can be radiometrically dated.
The basement gneiss cored at CIROS 2 377 m below sea level has been sampled for apatite fission-track dating by P.G. Fitzgerald at the University of Melbourne to extend the vertical sections he has sampled up nearby Mount Barnes and Trig Herb. The age obtained by this sample should provide the youngest rate of uplift obtained thus far for this part of the Transantarctic Mountains.page 11
A sea ice deformation survey was carried out during drilling at CIROS 2. Results of this survey and the other CIROS surveying programmes will be presented in the NZARP Surveyors Report 1904-85. Tidal movement was also recorded at CIROS 2 and is summarised in Table 5.
This report will appear, in condensed form, in the New Zealand Antarctic Record. The detailed core descriptions will be published along with core photographs and grain size data in the University's Antarctic Data Series in a few months. An article on the stratigraphy and sedimentology will be prepared over the next year for the New Zealand Journal of Geology and Geophysics. An article for Science or a similar journal on the chronology of the core is planned when the results of the present round of work are known.
We are now preparing for CIROS 1 in late 1986, and will present a more detailed proposal using multichannel seismic data from the S.P. LEE for drilling CIROS 3 and 4 from Cape Roberts in 1988 to the Ross Dependency Committee later this year.
The success of CIROS 1984 depended on many people, but we must first thank the team at Butter Point for their cheerfulness, enthusiasm and hard work. We are also grateful to Antarctic Division and the team of 1983 for setting up such a fine base at Butter Point. Throughout CIROS 1984 we depended heavily on the support of the Scott Base winter-over team led by Eric Saxby, and the 1984-85 team led by Peter Cresswell.
Barrett, P.J. 1982. Proposal for cenozoic Investigations in the Western Ross Sea (CIROS). NZ Antarctic Record, 4 (2): 32-39.
Barrett, P.J. and B.C. McKelvey, 1981. Cenozoic glacial and tectonic history of the Transantarctic Mountains in the McMurdo Sound region: recent progress from drilling and related studies. Polar Record, 20 (129): 543-548.
Barrett, P.J., Stoffers, P., Glasby, G.P. and W.L. Pluger, 1984. Texture, mineralogy and composition of four sediment cores from Granite and New Harbours, southern Victoria Land. NZ Journal of Geology and Geophysics.
Burdelik, W. 1981. Crustal model beneath McMurdo Sound from seismic refraction and gravity data. Unpublished MS thesis, Northern Illinois University, DeKalb, 114 pp.
Davey, F.J., Bennett, D.J. and R.E. Houtz, 1982. Sedimentary basins of the Ross Sea, Antarctica. NZ Journal of Geology and Geophysics. 25: 245-255.
Eittreim, S.L. and A.K. Cooper, 1984. Marine geological and geophysical investigations of the Antarctic continental margin, 1984. U.S. Geological Survey Circular, 935: 12 pp.
Elston, D.P. and S.S. Bressler, 1981. Magnetic stratigraphy of DVDP drill cores and Late Cenozoic history of Taylor Valley, Transantarctic Mountains, Antarctica. In: L.D. McGinnis (ed.). Dry Valley Drilling Project, American Geophysical Union Antarctic Research Series, 33: 413-426.
Harwood, D. 1984. Reinterpretation of Paleogene-Neogene glacial-marine sediments in MSSTS-1 drill hole, southwestern Ross Sea, Antarctica: the diatom record. Geological Society of America Abstracts, 1984.
Hicks, S.R. and D.J. Bennett, 1981. Gravity models of the lower Taylor Valley, Antarctica. NZ Journal of Geology and Geophysics, 24: 555-562.
Iles, D. and R.R. Dibble. A seismic refraction survey on sea ice near Butter Point, McMurdo Sound. In: A.R. Pyne (comp.), Immediate Report of VUWAE 25 1980-81, Antarctic Research Centre, Victoria University of Wellington, New Zealand.page 13
McCollum, D., 1975. Diatom stratigraphy of the Southern Ocean, In: D.E. Hayes and L.A. Frakes (eds.), Initial Report of the Deep Sea Drilling Project, Vol. 28, U.S. Government Printing Office, Washington: 525-572.
Northey, D.J., Brown, C., Christoffel, D.A., Wong, H.K. and P.J. Barrett, 1975. A continuous seismic profiling survey in McMurdo Sound, Antarctica. Dry Valley Drilling Project Bulletin No. 5, Northern Illinois University, DeKalb: 167-179.
Porter, S.C. and J.E. Beget, 1981. Provenance and depositional environments of Late Cenozoic sediments in permafrost cores from lower Taylor Valley, Antarctica. In: L.D. McCinnis (ed.), Dry Valley Drilling Project, American Geophysical Union Antarctic Research Series, 33: 351-364.
Stephenson, N. 1985. CIROS Drilling Project 1984. Butter Point Camp Manager's Report. Antarctic Division, DSIR, 29 pp.
Wilson, D.D., McGinnis, L.D., Burdelik, W.J. and T.L. Fasnacht, 1981. McMurdo Sound upper crustal geophysics. Antarctic Journal of the United States, 16 (5): 31-33.
Wong, H.K. and D.A. Christoffel, 1981. A reconnaissance seismic survey of McMurdo Sound and Terra Nova Bay, Ross Sea. In: L.D. McGinnis (ed.), Dry Valley Drilling Project, American Geophysical Union Antarctic Research Series, Washington: 37-62.
Wrenn, J.H. and P.N. Webb, 1982. Physiographic analysis and interpretation of the Ferrar Glacier-Victoria Valley area, Antarctica. In: C. Craddock (ed.), Antarctic Geoscience, University of Wisconsin Press, Madison: 1091-1100.