Depositional Environment of Sirius Group Sediments

DEPOSITIONAL ENVIRONMENT OF SIRIUS GROUP SEDIMENTS

D DEPOSITIONAL ENVIRONMENT OF SIRIUS E GROUP SEDIMENTS, TABLE MOUNTAIN, DRY VALLEYS AREA, ANTARCTICA

JAM BY ES R. JAMES R. GOFF1, IAN W. JENNINGS2 AND WARREN W. DICKINSON2 GOF F, IAN 1GeoEnvironmental Consultants, Lyttelton, New Zealand W. 2Research School of Earth Sciences, Victoria University of Wellington, JEN Wellington, New Zealand

Goff, J.R., Jennings, I.W. and Dickinson, W.W., 2002: Deposition- In the past few years, Sirius Group deposits have al environment of Sirius Group sediments, Table Mountain, Dry been at the centre of a debate concerning the rela- Valleys area, Antarctica. Geogr. Ann., 84 A (1): 11–24. tive stability of the East Antarctic Ice Sheet (EAIS) ABSTRACT. Outcrops and cores of the Sirius Group sediments since the Miocene (Barrett 1997; Miller and Mabin were studied at Table Mountain, Dry Valleys area, Antarctica. 1998). The debate has largely focussed on the two These sediments form a surficial veneer at least 9.5 m thick. Three opposing views of the ‘stabilist’ approach (March- facies – a gravelly sandstone, a sandstone, and a sandy conglomerate – are mapped and described from 13 outcrops and three ant et al. 1993; Denton et al. 1993), and the ‘dy- cores. The gravelly sandstone, constituting 13% of all cored ma- namicist’ argument (Webb et al. 1984). Consider- terial, is bimodal with matrix-supported clasts comprising 5–33% able importance has been placed on the presence or of the facies. Fabric analysis indicates that it was deposited pri- lack of Pliocene diatoms in near-surface Sirius marily by lodgment from glacial ice but with minor elements of meltout and flow. The sandstone facies, constituting 77% of all Group deposits and their mode of deposition (Bar- cored material, is a well-sorted, fine- to medium-grained sand, rett et al. 1992; Webb and Harwood 1991; Burkle which commonly has laminated bedding. It is predominantly a and Potter 1996; Barrett 1997). If marine diatoms, glaciofluvial deposit but has some glaciolacustrine elements. The found in the Sirius Group, are attributed to aeolian sandy conglomerate, constituting 10% of all cored material, is a minor facies. It is massive and clast-supported. It was deposited deposition and subsequent recycling (e.g. Stroeven in a high-energy environment suggestive of subglacial meltwater and Prentice 1997), they are unlikely to represent channels. the true age of the deposits. Furthermore, both Sirius Group sediments at Table Mountain are the result of wet- non-marine and marine diatom assemblages found based ice advancing and retreating over waterlain deposits. This is consistent with an advancing ice mass in climatic conditions in the Sirius Group may represent numerous that were warmer than present. The majority of the sediments palaeoenvironmental conditions. were deposited by alpine ice following a similar pathway to the The age of the Sirius Group deposits on Table present-day Ferrar Glacier and as such the depositional environ- Mountain has been estimated by several means. ment is one that concurs with evidence of a stable East Antarctic Ice Sheet approach. At Table Mountain, the predominantly gla- Barrett and Powell (1982) identiﬁed three units of ciofluvial and glaciolacustrine facies is inferred to represent a glacigenic sediment: diamictite, conglomerate, and more distal part of the Sirius Group environment than that seen at sandstone. Based upon regional geology, they sug- other outcrops in the Dry Valleys. gested these were probably older than late Mi- ocene. Minimum ages of 2.6–2.9 Ma and 6 Ma were established for exposed surfaces at Table Introduction Mountain using 10Be and 3He/21Ne respectively, The Sirius Group is a series of glacigenic sediments and it is likely that the deposits are indeed much that are found mostly at elevations greater than older (Ivy-Ochs et al. 1995; Bruno et al. 1997). 1600 m in over 40 locations throughout the There are few comprehensive lithostratigraphic Transantarctic Mountains (TAM) (Mercer 1972; studies of Sirius Group deposits in the Dry Valleys Mayewski 1975; Barrett and Powell 1982; area and those that are available relate either to McKelvey et al. 1991). They have been generally provenance (Faure and Taylor 1981; Taylor and described as a compact glacial drift that uncon- Faure 1983) or to a broad-brush interpretation of formably covers pre-Tertiary rocks (Mercer 1972). glacial geology (Mayewski 1975; Barrett and Pow-

Geograﬁska Annaler · 84 A (2002) · 1 11 JAMES R. GOFF, IAN W. JENNINGS AND WARREN W. DICKINSON

Fig. 1. (A). Study area: Table Mountain, Transantarctic Mountains, Antarctica (tg = Tedrow Glacier, eg = Emmanuel Glacier). (B). Geology of Table Mountain: detail shows locations of Sirius Group deposits and study sites (transect A–B is shown in Fig. 5, mm f = mass movement feature). Parallel cores were taken from sites TM-1 (1, 1B, 1C) and TM-7 (7A, 7B). ell 1982). The most comprehensive work on the been relatively stable during warm periods in the glacial geology of the Sirius Group was reported by Pliocene (Stroeven and Prentice 1997). Stroeven and Prentice (1997) who addressed the is- At Table Mountain the Sirius Group forms a ve- sue of alpine versus ice sheet glaciation, an issue neer of sediment, deposited on a gentle slope that which is linked to the stabilist–dynamicist debate. formed the bottom of the ancestral Ferrar Valley. Stroeven and Prentice (1997) refute the dynamicist Today the Sirius Group at Table Mountain is theory that the EAIS was much reduced in area and perched about 1000 m above the present surface of volume during the Pliocene by producing an alpine the Ferrar Glacier, at an elevation of 2000 m. In this ice interpretation for Sirius Group deposits at paper we investigate the depositional environment Mount Fleming. They suggest that the Sirius Group of the Sirius Group at Table Mountain using surface deposits at Mount Fleming were not laid down by exposures and cores. expansion of the EAIS, which is assumed to have

12 Geograﬁska Annaler · 84 A (2002) · 1 DEPOSITIONAL ENVIRONMENT OF SIRIUS GROUP SEDIMENTS

(A)

(B)

(C)

Fig. 2. (A) Site 11.1. Sharp and undulated lower contact between gravelly sandstone facies and Terra Cot- ta Siltstone (hammer for scale). (B) Sandstone outcrop at site 6.34 (per- son for scale). (C) Sandy conglomerate at site 6.28 (notebook for scale).

Physical setting Mountain is early Palaeozoic granite which is un- Table Mountain is located on the southern side of conformably overlain by Devonian sediments of the Ferrar Glacier and is bounded to the southwest the Beacon Supergroup comprising the continental by the Tedrow Glacier, and to the east by the Em- sequence of Windy Gully Sandstone, Terra Cotta manuel Glacier (Fig. 1A). Basement rock at Table Siltstone, and New Mountain Sandstone (Fig. 1B).

Geograﬁska Annaler · 84 A (2002) · 1 13 JAMES R. GOFF, IAN W. JENNINGS AND WARREN W. DICKINSON

GRAVEL

Conglomerate

80% SANDY CONGLOMERATE FACIES

TM-1C Sandy conglomerate TM-6

TM-7B

OUTCROPS McMurdo Ice Shelf 30% Taylor Glacier

CIROS-1 Gravelly sandstone GRAVELLY SANDSTONE FACIES 5% 1% Sandstone Fig. 3. Ternary plot of lithostrati- 9:1 1:1 1:9 graphic facies including core and MUD SANDSTONE SAND FACIES outcrop samples.

Table 1a. Fabric data from major outcrops: summary of field data and directional information

Lithology: (no./no. striae top/no. striae base/no. keels) No. Eigenvector Mean Site clasts (degrees) S1 Roundness Sphericity a-axis D SS Q QA G

1.0 54 217 0.75 0.55 0.59 17.8 29/0/0/9 16/8/8/4 4/1/1/0 5/0/0/0 0 TM-1a 51 290 0.46 0.57 0.63 11.6 21/0/2/0 16/5/9/3 4/0/1/1 10/1/1/2 0 TM-1c 52 272 0.58 0.58 0.64 29.2 37/0/0/0 8/1/1/1 3/0/0/0 4/0/0/0 0 TM-1e 51 279 0.69 0.58 0.63 29.8 29/6/7/5 8/2/3/2 3/0/2/2 11/2/4/4 0 J2 51 290 0.60 0.57 0.62 20.6 36/1/1/14 9/3/4/3 3/0/0/0 3/1/1/0 0 J3 51 270 0.52 0.58 0.57 34.6 39/0/0/3 8/6/6/6 0 4/0/0/1 0 TM-6a 51 282 0.59 0.58 0.56 20.5 34/6/7/4 13/4/5/1 2/0/0/0 2/0/0/0 0 TM-6b 51 282 0.91 0.59 0.58 25.5 23/5/5/5 20/8/7/6 7/4/0/0 1/1/0/0 0 TM-6c 51 280 0.90 0.59 0.53 21.2 27/0/4/8 14/0/4/7 6/0/2/0 4/0/1/2 0 TM-6d 51 280 0.90 0.58 0.55 21.2 26/0/0/9 14/0/0/4 6/0/0/0 5/0/0/0 0 TM-7a 51 284 0.80 0.59 0.56 18.8 37/8/11/9 9/3/3/4 3/0/0/1 1/0/0/0 1/0/0/0 TM-7b 51 285 0.89 0.59 0.56 15.8 41/13/17/6 5/4/3/1 3/1/1/0 2/0/0/0 0 J6 51 281 0.79 0.58 0.55 17.0 32/13/15/13 12/6/6/6 1/0/0/0 2/0/2/1 4/2/3/0 6.34 51 277 0.73 0.60 0.54 13.8 44/8/8/7 1/1/1/0 2/0/0/0 4/1/1/2 0 11.1a 51 259 0.94 0.56 0.56 14.2 6/0/0/2 45/0/0/4 0 0 0 11.1b 51 281 0.80 0.59 0.54 12.2 39/17/18/3 8/4/5/0 3/3/3/1 1/1/1/0 0 12.17 51 280 0.74 0.59 0.55 19.8 32/15/15/6 7/3/3/2 4/2/2/0 3/2/2/2 5/1/1/1 J11 43 221 0.75 0.52 0.57 18.2 0 0 2/0/0/0 41/1/4/3 0 J12 43 233 0.77 0.52 0.64 16.9 0 17/0/4/2 3/0/0/1 23/0/0/1 0

Refer to Fig. 1 for site locations and Fig. 4 for stereonet diagrams. Units a, b, c, d, e were measured from the base of the section upwards, a = lowest unit in a section and so on. D = dolerite, SS = Terra Cotta Siltstone, Q = quartz, QA = quartz arenite, G = granite. Keel = keel-like form indicating weight distributed towards lower part of clast.

This sequence of sandstone and siltstone was sub- the present-day Ferrar Glacier, and is covered by a sequently intruded by the Ferrar Dolerite in the residual regolith comprising predominantly dolerite Jurassic (Barrett and Powell 1982; Fig. 1B). clasts, ranging in size from granules to large boul- The study area is a broad, gently dipping slope ders derived from Sirius Group deposits. Sirius between 1750 and 2000 m elevation, on the north- Group deposits at Table Mountain generally crop west side of Table Mountain. It is about 750 m above out as either low ridges covered by residual regolith,

14 Geograﬁska Annaler · 84 A (2002) · 1 DEPOSITIONAL ENVIRONMENT OF SIRIUS GROUP SEDIMENTS

Table 1b. Summary interpretation of fabric data.

Site Interpretation

1.0 Predominantly lodgment tillite – deposited from SW (Tedrow direction), reworked by ice from W (Ferrar direction) TM-1 Three distinct units of predominantly lodgment tillite separated by sand-filled shear planes (units b & d)–shows reorientation of flow from SW to W TM-1a Lodgment – deposited from SW, with some reworking in upper part of unit from W TM-1c Predominantly lodgment superposed on flow (shear cluster)–deposited from W, although initial deposition probably from SW TM-1e Probably lodgment from W reoriented by more lodgment from SW direction. Some clast rolling during deposition J2 Predominantly flow tillite, iceberg-rafted material related to ice movement from W. Some ice marginal reworking–possible remnant bimodal lodgment fabric J3 Strong lodgment of boulder into underlying lodgment tillite–reorientation of clasts indicate ice from W overidden material emplaced by weaker lodgment from SW TM-6 Transition from lodgment to flow tillite– ice advance with subsequent stalling/ ice marginal reworking. Overridden ice marginal lake sediments at base TM-6a Lodgment from W superposed on lodgment from SW–both shears and tension fractures present TM-6b Strong lodgment from W with some spread perhaps by subglacial meltout TM-6c Predominantly meltout–palaeodirection from W–possibly some earlier lodgment (weak transverse mode) TM-6d Flow till, reworking post-depositionally, no striae. Flow from till ridge? TM-7 Bottom section represents lake/ponded sediments separated by an erosional contact from overlying conglomerates–Over- riding ice sheared conglomerate into overlying tillite–overloading and dewatering structures present TM-7a Predominantly lodgment with some meltout (spreadout unimodal fabric, but not a girdle generated by flows)–some clasts rotated in viscous flow at base TM-7b Strong lodgment–associated flow in underlying sediments J6 Strong lodgment by ice moving from W. Appears to have lodged into proglacial deposits of waterlain sediment. Brecciated sands and water escape features 6.34 Strong lodgment of ice moving from W. Has lodged/sheared up at the top of the main ridge adjacent to hollow. Might mark confluence with cleaner ice to east. Some clasts have rolled/slid and some vertical rotation which might be due to overtopping of a moraine ridge/settling. Strong shear indicates active lodgment from west. 11.1 This section represents a weak stagnating SW ice flow (much meltwater) overridden by active ice from W 11.1a Probable meltout, ice from SW–possibly waning in comparison to ice from W? Very weak transverse mode suggests possible lodgment involved in lower part of unit 11.1b Strong lodgment from W–some settling indicated by reverse modality, perhaps caused by underlying instability in meltout? 12.17 Strong lodgment of ice moving from W. Lodgment into a moraine-dammed lake, the oversteepened tillite has settled slightly, some subsequent meltout has occurred J11 Weak meltout of a thin veneer of subglacial material. Striae data are equivocal–ice from S., bimodal fabric from some rolling/ sliding suggests general SSW movement J12 Weak meltout of a thin veneer of subglacial material. Unconsolidated material, fabric generally strong but deflation and settling appears to have generated bimodality

Refer to Fig. l for site locations. Refer to Table 1a and Fig. 4 for fabric data. Units a, b, c, d, e were measured from the base of the section upwards, a = lowest unit in a section and so on. or beneath mounds of large boulders of residual regolith. One outcrop is a prominent ridge that forms Methods the highest elevation of the Sirius Group at Table Field mapping was carried out by ‘pace and com- Mountain. This ridge is adjacent to a prominent pass’ with elevations determined using a combina- elongated depression that strikes northeast and con- tion of a THOMMEN surveying altimeter, and tains outcrops of the Terra Cotta Siltstone. A series GPS sites which were fixed to the trigonometrical of less well deﬁned, subparallel NE-trending ridges station at Table Mountain. A magnetic correction is found downslope from this main ridge (Fig. 1B). of 154o was used to give true grid readings. Fabric Patterned ground and several large mass movement was recorded at 12 outcrops by measuring the trend features indicate some degree of post-depositional and plunge of the long axis of about 50 prolate reworking in some areas. clasts in each subsample. Data were plotted on Schmidt equal-area projections, and additional

Geograﬁska Annaler · 84 A (2002) · 1 15 JAMES R. GOFF, IAN W. JENNINGS AND WARREN W. DICKINSON

Fig. 4. Fabric data for all major outcrops studied–see Fig. 1 for site locations. Stereonets include clast imbrication, striae (top surface striae: black fill rose diagram, bottom surface striae: grey fill rose diagram), poles to shear planes (black fill squares), and stoss/lee features (black fill circles indicate orientation of stoss end, adjacent number = number of clasts with this orientation); see Table 1 for a summary of data, directional and depositional interpretations. These are lower hemisphere plots of a Lambert/Schmidt equal-area pro- jection. Dip and dip direction are contoured at 2% intervals.

16 Geograﬁska Annaler · 84 A (2002) · 1 DEPOSITIONAL ENVIRONMENT OF SIRIUS GROUP SEDIMENTS

Fig. 5. Stratigraphy of Sirius Group deposits. (a) Transect A–B (see Fig. 1B) with representative composite sequence. (b) Stratigraphy for cores TM-7B, TM-6, and TM-1C. The horizontal direction of each unit gives relative texture, coarsening towards the right. data such as striae, stoss/lee features and keels were conglomerate units. For analysis of the conglom- noted and used to assist with interpretation. Per- erate, percentage sand, silt, and mud have been nor- centage gravel, and the ratio of sand to mud were malized to the estimated gravel percentage. plotted to determine primary textural names. Cores were taken with a portable drilling system using compressed air as a ﬂushing medium (Dick- Results and discussion inson et al. 1999). Cores were stored frozen at a Outcrops constant temperature of –18oC and grain size sam- Three distinct lithological facies were mapped and ples were taken at approximately 0.5 m intervals described from 13 outcrops at Table Mountain from three cores (TM-1C, TM-6, TM-7B). Sample (Figs 1B, 2–4). These facies are a gravelly sand- weight was generally about 20 g but the true clast stone, sandstone, and sandy conglomerate. To content of the conglomerate was not well repre- avoid possible misinterpretations associated with sented. To address this problem, the percentage names such as tillite and diamict, all lithological gravel measured from grain size analysis was re- descriptions in this text follow the textural classiﬁ- placed by a visual estimate of clast content of the cation of Folk et al. (1970).

Geograﬁska Annaler · 84 A (2002) · 1 17 JAMES R. GOFF, IAN W. JENNINGS AND WARREN W. DICKINSON

Gravelly sandstone facies. In surface exposures, from active ice which flowed from the west-north- this is the most extensive and variable of the three west. facies (Fig. 2a). Typically it is a massive, light pink- In contrast, fabric data from outcrops at sites ish brown to light brownish grey, bimodal, gravelly 11.1 and J3, indicate that there was a change in ice sandstone. The mainly subangular clasts constitute flow direction. These data indicate weak lodgment 5–33% of the material and range in size from gran- by ice flow from the southwest with subsequent ules to cobbles. Clast lithology is predominantly overriding by strong lodgment by ice flow from the dolerite, with some siltstone, quartz arenite, and west. This has been interpreted as representing the quartz. This facies is commonly cut by northwest- overriding of waning Tedrow Glacier ice by the dipping fractures imparted during deposition, some rapidly advancing Ferrar Glacier ice. Ice from the of which are intruded by veins of sandstone (see be- Tedrow Glacier direction is inferred to have low). In addition, there are numerous shear, defor- reached a maximum ice front towards the south- mation, water-escape structures and lenses of both west corner of the study site where it merged with sandstone and conglomerate. There is considerable Ferrar Glacier ice. variation in clast lithology and stratification, but Other sites confirm this strong lodgment signal because the clasts are mostly striated and strongly of ice advancing from the west. Site 6.34 is located oriented in the direction of palaeoflow, this facies on the uppermost moraine ridge (Fig. 1B) and has been interpreted as a subglacial diamicton. This shows evidence of active till emplacement with interpretation is refined using fabric analysis (Table lodged layers sheared over underlying material. 1, Fig. 4). Some vertical clast rotation has been caused by mo- The base of the gravelly sandstone facies was raine overtopping and subsequent settling out of visible at two sites, 11.1 and TM-7. At site 11.1, the material. Further downslope, sites J6 (west) and facies overlies Terra Cotta Siltstone with a sharp 12.17 (north) represent active ice advances into wa- and undulated lower contact (Fig. 2a). The base terlain sediments and water, respectively. Water- comprises a clast-rich (30–60%) sandy conglom- escape features and brecciated sands were recorded erate. The predominantly shale clasts are angular at J6, and the post-depositional settling of lodg- and range in size from granules to large cobbles. ment till was found at 12.17. Emplacement by ac- Shale concentrations decrease upsection and the tive ice in these two downslope locations is also unit grades into a typical gravelly sandstone with shown by the occurrence of granitic clasts sourced large dolerite boulders (<1.5 m a-axis). from below the dolerite sill. These have been At TM-7, a sharp erosional contact overlies the sheared up over the sill, but have not been carried sandstone facies of the Sirius Group and clasts have to the uppermost moraine ridges. a strong preferred orientation to the west-northwest Gravelly sandstone deposits at sites J11 and J12 with parallel striae and stoss/lee features. The fab- are anomalous. They do not indicate the almost ric is strong, and while lodgment is the predomi- ubiquitous deposition associated with active Ferrar nant process involved, there are elements of melt- Ice flow. The fabric indicates the presence of a lo- out, and possibly flow as indicated by a relatively cally derived ice mass flowing from the Table dispersed fabric (at site TM-7d). The overall pic- Mountain–Navajo Butte area to the south-south- ture is of an initial strong lodgment of material by west. These deposits are thin, unconsolidated, and active ice flowing from the west-northwest, fol- run subparallel to the elongate depression (hollow) lowed by subsequent stagnation and decay. It is adjacent to the uppermost moraine ridge. possible that overriding by a more recent ice advance, or periglacial or mass wasting processes Sandstone facies. This facies was rare in outcrop could have imparted a dispersed fabric. However, but was generally found in topographic depressions these alternative scenarios are unlikely because of beneath a veneer of surface debris; however, it was the lack of evidence for a subsequent ice advance. abundant in all cores (Fig. 2b). The sandstone Furthermore, the ubiquitous nature of the fabric at facies is also rare or non-existent in other Sirius these sites indicates that locally variable periglacial Group outcrops in the Dry Valleys area, probably or mass wasting processes have not substantially reflecting the absence of core data. Texturally, the altered the (macro)fabric. sandstone is the most homogeneous and friable of While fabric for sites TM-1 and 1.0 further east the three facies. Typically, it is a light brown, well have a slightly weaker cluster of points, they are sorted, medium to fine sandstone that contains few lodgment in origin. They also indicate deposition gravel-sized clasts (<2%). Outcrops and cores of

18 Geografiska Annaler · 84 A (2002) · 1 DEPOSITIONAL ENVIRONMENT OF SIRIUS GROUP SEDIMENTS the facies show fine laminations, and contain rare Cores dispersed dropstone clasts. TM-1C. A 7.9 m core was taken from a small out- At site 6.34 the facies was found in outcrop and crop of gravelly sandstone located at the southern contained two distinct units. The lower unit had end of the highest ridge (1948 m a.s.l.). The core fine (<5 mm), planar stratification while the up- was subdivided into seven units and contained ex- per unit was massive, and contained clasts with amples of all three facies (Fig. 5). associated soft sediment deformation. Horizontal Sandstone is the dominant facies in the lower stratification and the presence of dropstones in- half of TM-1C, with the lowest unit being highly dicate that these sediments were deposited in a deformed by both subhorizontal and vertical frac- proglacial environment often involving ice-raft- tures. Bedding within the sandstone includes pla- ed debris. To the west, at site TM-7, the sandstone nar low-angle, planar high-angle, and trough facies is overlain by gravelly sandstone, indicat- cross-bedding. The facies has two erosional con- ing that an advancing glacier overrode proglacial tacts, both of which are overlain by fining-upward sediments. sequences, the former from a gravelly sandstone to The morphological expression of the Sirius a fine sandstone, the latter from a sandy conglom- Group sediments in the study area indicates that erate to a coarse sandstone. most of the ridges are end-moraines composed of gravelly sandstone facies and the sandstone facies is TM-6. This 4.1 m core was taken at an elevation of located in hollows between the ridges. The sand- 1909 m a.s.l. on a low ridge, subparallel to the dip stone facies is interpreted as being a predominantly of the slope. The core contains four units of grav- glaciofluvial deposit with some glaciolacustrine el- elly sandstone and a sharp, erosional contact be- ements. The sandstone would most likely have been tween units 2 and 3 that is considered to be a shear deposited in streams feeding moraine-dammed plane. The lower contact with the underlying Bea- lakes that formed in a distal environment as the gla- con Supergroup was not cored, but clasts in the cier retreated. Such relatively coarse-grained lake lowest unit are dominated by the grey Terra Cotta sediments are unusual, but may indicate the high Siltstone, indicating that it may represent the top of energy regime associated with a predominantly gla- a basal unit, similar to that seen at site 11.1. ciofluvial environment. Grain size analysis shows that all units are bimodal, with an additional bimodality in the sand Sandy conglomerate facies. The least extensive of range that is most likely conferred by incorporation the three facies is mainly found in flat areas clear of of a combination of the Windy Gully and New surface debris (Fig. 2c). It is typically a light pink- Mountain Sandstones. ish grey, massive, bimodal, sandy conglomerate (e.g. site 6.28). The matrix is a well sorted, medium TM-7B. A 9.5 m core (elevation: 1863 m a.s.l.) was sand. Clast-supported material (50–80% clasts) is taken at a gravelly sandstone outcrop on a long, mostly well rounded, granule to cobble-sized, and eastward-trending ridge. Most of the 11 units found poorly sorted. Clast lithologies are mainly dolerite, in this core were sandstone facies. The gravelly with some quartz and quartz arenite. sandstone facies on the surface was not recovered This facies is generally associated with the grav- in the drill core, but was assumed to have extended elly sandstone. However, unlike the gravelly sand- about 0.4 m below the surface. Grain size data show stone it is clast-supported. The rounded clasts and that most units range from well sorted to poorly coarse texture indicate that this was deposited in a sorted. Coal clasts (<1%) are uniformly distributed high-energy environment. The form of the associ- throughout the sandstone facies, and in places they ation of the sandy conglomerate with the gravelly form discrete lenses up to 10 mm thick. sandstone at sites 11.1 and TM-6 indicates that the sandy conglomerate was deposited in subglacial meltwater channels and deposition must have been Core summary rapid because striae are preserved on several clasts. The gravelly sandstone facies comprises 13% of all However, where it is exposed in depressions and in cored material recovered, as opposed to 10% for association with the sandstone facies, it appears to the sandy conglomerate, and 77% for the wide- have been laid down either as a result of slumping spread sandstone facies. or channel fill. A comparison of the facies from the cores and outcrops shows that the gravelly sandstone found in

Geograﬁska Annaler · 84 A (2002) · 1 19 JAMES R. GOFF, IAN W. JENNINGS AND WARREN W. DICKINSON

Fig. 6. Schematic model of the depositional sequence on Table Mountain from emplacement of the uppermost moraine (it is assumed that bedrock underlies these deposits. Bedrock forms the base of the hollow). (A) Active lodgment of material, some overspill from the uppermost moraine into the abutting bedrock-lined hollow (possibly ice-filled), granite sheared up from beneath dolerite sill, reworking and overloading of ice marginal sediments. (B) Ice retreat/downwasting, aeolian and meltwater processes active, melting of any ice in bedrock-lined hollow. (C) Continued ice retreat/ downwasting, aeolian removal of material from bedrock-lined hollow, slumping and settling of oversteepened moraine ridges, continued meltwater processes. (D) Ice downwasted, ongoing periglacial processes (aeolian, mass movement and meltwater), most material transported below dolerite sill is removed by new Ferrar Glacier ice carving a deeper valley west of Table Mountain.

20 Geografiska Annaler · 84 A (2002) · 1 DEPOSITIONAL ENVIRONMENT OF SIRIUS GROUP SEDIMENTS the upper parts of cores TM-1C and TM-6, is the Provenance same as that studied in outcrop. Furthermore, there Provenance of the Sirius Group at Table Mountain is little variation in the gravelly sandstone with can be estimated from the proportion of different depth. However, the proportion of sandstone to clast types seen in outcrop (Table 1). About 67% of gravelly sandstone is much greater in the cores than all clasts in the Sirius Group are dolerite. Clasts in outcrop with the sandstone sequences at least constitute approximately 21% of the total outcrop, 9 m thick at one site. Stratigraphically, the sand- and therefore, as much as 15% of the Sirius Group stone is the lowest facies of the Sirius Group (Fig. at Table Mountain has probably been derived from 5). Ferrar Dolerite, which forms sills and dykes In the cores, the sandstone facies ranges from throughout the TAM. About 85% of outcrop mate- massive to cross-bedded, and contains erosional rial includes siltstone gravels, quartz arenite grav- contacts overlain by fining-upward sequences. els, sand, silt and mud that is most likely from the Pebble lenses are also common in many of the Beacon Supergroup. The gravels rich in quartz units. These features suggest that at least some of clasts (about 1% of the total outcrop), are thought the sandstone was deposited in a high energy envi- to have come from the Feather Conglomerate. This ronment comparable with hypothetical fluvial formation of the Beacon Supergroup crops out at models presented by Miall (1978). Although the Mount Feather, 35 km to the east, and up-glacier sandstone in the cores more closely resembles the from Table Mountain (McElroy and Rose 1987). Platte-type model, features of both the Donjek- and A similar exercise can be carried out for the Platte-types are seen (Miall 1978). However, the cores. From the average proportion of gravel-sized presence of ice-rafted dropstones throughout the clasts estimated for each core it is also possible to sandstone in the core shows that these sediments evaluate the provenance of the material of the Sirius were deposited in an ice marginal zone with a com- Group at Table Mountain. The estimated clast con- bination of glaciolacustrine and glaciofluvial envi- tents are: TM-1C 15%, TM-6 14%, TM-7B 3%. ronments. The proportion of dolerite, quartz arenite, and mud- Core TM-1C has a sequence of sandstone over- stone clasts is approximately the same as in the out- lain by gravelly sandstone, which is believed to re- crops which means that about 7% of the Sirius present glacial overriding of a glaciofluvial envi- Group at Table Mountain is from the Ferrar Doler- ronment, in other words, an ice advance. Although ite, and 93% is from the Beacon Supergroup. The this sequence suggests one glacial advance, there is small amounts (<1%) of coal found throughout the evidence in the core to suggest there may have been sandstone facies of the cores is most likely from the several advances. The lowest part of the sandstone Weller Coal Measures, a formation within the Vic- facies is highly deformed and has an upper erosion- toria Group, which forms the upper part of the Bea- al contact with the sandy conglomerate facies. This con Supergroup. The Weller Coal Measures crop contact might represent an earlier glacial advance out at the tops of Mount Feather and Mount Weller across, and deforming, the proglacial sediments. (McElroy and Rose 1987) which are about 35 km Core TM-6 was drilled through an outcrop of the to the east, and up-glacier of Table Mountain. gravelly sandstone facies at least 4.0 m thick. The sequence was deposited by a glacier advancing over glaciofluvial or glaciolacustrine sediments, similar Depositional model to the interpretation for core TM-1C. As the sand- In this paper the Sirius Group at Table Mountain is stone is not seen in the bottom of core TM-6, this described in terms of three distinct lithofacies. might indicate that either the fluvial sediments were Sandstone is the most common facies in core (77% completely removed by advancing ice or that core of core), but not in outcrop, and is stratigraphically TM-6 was drilled to the side of a fluvial channel. the lowest part of the Sirius Group. The sandstone The sediments described in core TM-7B are pre- facies is rare at all other Sirius Group sites in the Dry dominantly sandstone that has been overlain by Valleys area. While this may be a function of the ab- gravelly sandstone interpreted to be basal till. This sence of coring at other locations, it may also rep- is the same sequence as found in core TM-1C and resent the erosional expression of the deposits at Ta- is also interpreted as a glacial advance. However, ble Mountain. Our data show that the environment since core TM-7B does not extend into the under- of deposition for the sandstone facies is thought to lying rocks of the Beacon Supergroup, no earlier be both glaciofluvial and glaciolacustrine. Gravelly possible advances can be determined. sandstone is the next most common facies in core

Geografiska Annaler · 84 A (2002) · 1 21 JAMES R. GOFF, IAN W. JENNINGS AND WARREN W. DICKINSON and is interpreted as till or diamicton. This forms a Many of the sedimentary structures and grain thin layer above the sandstone facies, and accounts size analyses show deposition by glaciofluvial and for most of the outcrops seen at Table Mountain. glaciolacustrine conditions indicating wetter con- The third facies is a sandy conglomerate, which is ditions than exist today. The proposed depositional relatively sparse in both outcrop and core and is model suggests a rapid ice advance (e.g. Menzies found largely as lenses in the sandstone facies. Dep- 1995), although we are unable to assess the rapidity osition is believed to have taken place mainly in from the data available. subglacial meltwater channels. Aspects of the to- Climatic conditions were warmer than present pography at Table Mountain have a strong relation- and the deposits at Table Mountain appear to rep- ship to the distribution of the different facies. Facies resent the maximum extent of a Ferrar Glacier ice appear to control the present morphological expres- advance, although subsequent erosion has removed sion. Ridges are capped by gravelly sandstone over- some of the evidence. It is probable that the ad- lying the sandstone facies which is mainly exposed vance represented a change from cold to wet-based in the intervening depressions or beneath large ice. Clast provenance and the interaction between boulders. The presence of mass movement features small ice masses (Tedrow Glacier ice, Table Moun- and patterned ground indicates some degree of post- tain-Navajo Butte ice, and Ferrar Glacier ice) are depositional reworking, but this does not appear to interpreted as representing deposition by alpine ice have affected the macrofabric of prolate clasts. The as opposed to an ice sheet advance. This work sup- effect of patterned ground processes upon micro- ports that of Stroeven and Prentice (1997) at Mount fabric, and lateral and vertical movement within the Fleming in favouring a stable ice sheet scenario. sediments was not determined. The Sirius Group deposits on Table Mountain The presence of coal and quartz clasts in the Sir- are a small part of the entire group (McKelvey et al. ius Group indicate an easterly flow bringing errat- 1991) and further glacial geological work is needed ics from about 35 km to the east near Mount Feath- to see if they are representative of depositional en- er. However, the form of the channel in which the vironments at other sites. However, the presence of glaciofluvial sediments were deposited may have small pockets of Sirius Group deposits at other lo- been very different from the present Ferrar valley. cations is inferred to represent similar small alpine Barrett and Powell (1982) proposed that a broad, ice advances in response to a warming climate. At relatively shallow valley existed between Table Table Mountain, the predominantly glaciofluvial Mountain and Knobhead (Fig. 1A). While there is and glaciolacustrine facies is inferred to represent no direct evidence in the cores to support this mod- a more distal part of the Sirius Group environment el, the valley floor on which the glaciofluvial sedi- than that seen at other outcrops in the Dry Valleys. ments were deposited must have been at an elevation similar to the base of these sediments. There- fore, the valley would have been similar to the pro- Conclusions file indicated by Barrett and Powell (1982). Sirius Group deposits at Table Mountain are be- We propose a model of the depositional se- lieved to be the result of wet-based ice advancing quence which starts from the emplacement of the and retreating over waterlain material. The topo- uppermost moraine ridge (Fig. 6). For simplicity, graphy is largely deglacial, and has in localised ar- the model only considers depositional and eas been modified by periglacial processes and post-depositional events with respect to a retreat mass movement, which seems to have had little ef- and downwasting from this point. Alpine ice ad- fect upon the macrofabric of outcrop and cores. The vanced over glaciofluvial material and then as it re- maximum extent of Ferrar ice advance at this site treated a series of end moraines formed with melt- is marked by the uppermost moraine ridge that water drainage through inter-moraine streams and abuts a bedrock-floored elongate depression. Sub- lakes. Final ice retreat left a topography of gla- sequent ice retreat and downwasting from this ciofluvial sediments overlain by a thin veneer of point is marked by a series of subparallel ridges subglacial diamictons. Contemporary and subse- with meltwater (glaciofluvial and glaciolacustrine) quent aeolian processes have created a residual re- deposits laid down between them. golith which serves to armour the underlying sed- When Ferrar ice advanced from the west, the iments. Localised mass movement and patterned southwestern part of Table Mountain appears to ground processes have also contributed to the have been occupied by a weaker ice mass emanat- present geomorphology. ing from the direction of the Tedrow Glacier. This

22 Geograﬁska Annaler · 84 A (2002) · 1 DEPOSITIONAL ENVIRONMENT OF SIRIUS GROUP SEDIMENTS is indicated by variations in fabric and composi- tion. This ice mass was subsequently overridden References and much of its glacigenic sediments were re- Barrett, P.J., 1997: Antarctic paleoenvironment through Cenozo- worked. At the time of maximum advance it is pos- ic times–a review. Terra Ant., 3: 103–119. Barrett, P.J. and Powell, R.D., 1982: Middle Cenozoic glacial sible that the uppermost bedrock-floored hollow beds at Table Mountain, southern Victoria land. In: Craddock, was filled with ice flowing from the Table Moun- C. (ed.): Antarctic geoscience. University of Winconsin tain–Navajo Butte area. However, there is no direct Press, Madison: 1029–1067. evidence for the occupation of the hollow by ice Barrett, P.J., Adams, C.J., McIntosh, W.C., Swisher, C.C. III. and Wilson, G.S., 1992: Geochronological evidence supporting from any source. Antarctic deglaciation three million years ago. Nature, 359: Most subglacial diamicton was emplaced by 816–818. lodgment, and on the lower slopes of the study area Bruno, L.A., Heinrich, B., Thomas, G., Schluchter, C., Signer, P. there is evidence for active ice introducing granitic and Wieler, R., 1997: Dating of Sirius Group tillites in the Antarctic Dry Valleys with cosmogenic 3He and 21Ne Earth material from beneath the dolerite sill. Active Planet. Sci. Lett., 147: 37–54. shearing was also evident from within the outcrops Burkle, L.H. and Potter, N., Jr., 1996: Pliocene-Pleistocene dia- and cores studied where lodgment overlies gla- toms I Paleozoic and Mesozoic sedimentary and igneous cioﬂuvial material. Deposition is consistent with rocks from Antarctica: A Sirius problem solved. Geology, 24: 235–238. advancing ice, although the rapidity of the advance Denton, G.H., Sugden, D.E., Marchant, D.R., Hall, B.L. and cannot be determined from the available data. The Wilch, T.I., 1993: East Antarctic Ice Sheet sensitivity to interactions between three alpine ice sources indi- Pliocene climate change from a Dry Valley perspective. cate that these sediments were not laid down by ex- Geogr. Ann., 75A: 155–204. Dickinson, W.W., Cooper, P., Webster, B. and Ashby, J., 1999: A pansion of the EAIS and this supports the stabilist portable drilling rig for coring permafrosted sediments. J. Sed. conclusions of Stroeven and Prentice (1997). Res., 69: 518–527. Faure, G. and Taylor, K.S., 1981: Provenance of some glacial deposits in the Transantarctic Mountains based on Rb-Sr dating Acknowledgements of feldspars. Chem. Geol., 32: 271–290. Folk, R.L., Andrews, P.B. and Lewis, D.W., 1970: Detrital sedi- This paper relates to work carried out as part of a mentary rock classification and nomenclature for use in New FRST funded programme, No. DIC 601 to W.W.D. Zealand. NZ J. Geol. Geophys., 13: 937–968. In part it represents the results of a BSc Honours Ivy-Ochs, S., Schluchter, C., Kubik, B., Dittrich-Hannen, B. and Beer, J., 1995: Minimum 10Be exposure ages of early Pliocene thesis completed by I.W.J. carried out as part of the for the Table Mountain plateau and the Sirius Group at Mount programme. Thanks are due to Pat Cooper and Bain Fleming, Dry Valleys, Antarctica. Geology, 23: 1007–1010. Webster for drilling services and producing the Marchant, D.R., Swisher, C.C. III, Lux, D.R., West, D.P. Jr. and core we needed. Jon De Vries is thanked for solving Denton, G.H., 1993: Pliocene paleoclimate and East Antarctic ice sheet history from surface ash deposits. Science, 260: 667– almost insurmountable problems of both machin- 670. ery and logistics. Antarctica New Zealand and all Mayewski, P.A., 1975: Glacial geology and late Cenozoic history its staff both in Christchurch and Scott Base are of the Transantarctic Mountains, Antarctica. Institute of Polar thanked for their help and understanding. Studies Report 56. Ohio State University, Columbus. 75 p. McElroy, C.T. and Rose, G., 1987: Geology of Beacon Heights, Southern Victoria Land, Antarctica. NZ Geol. Survey Misc. Dr James R. Goff, GeoEnvironmental Consultants, Series Map 15, scale 1:50000, 1 sheet and notes. 11 The Terrace, Governors Bay, Lyttelton RD1, McKelvey, B.C., Webb, P.N., Harwood, D.M. and Mabin, M.C.G., New Zealand 1991: The Dominion range Sirius Group: a record of the late Pliocene–early Pleistocene Beardmore Glacier. In: Thomson, E-mail: [email protected] M.R.A., Crame, J.A. and Thomson, J.W. (eds): Geological evolution of Antarctica. Cambridge University Press. Cam- Ian W. Jennings, Research School of Earth Scienc- bridge. 675–682. es, Victoria University of Wellington, PO Box 600, Menzies, J., 1995: The dynamics of ice flow. In: Menzies, J. (ed.): Modern glacial environments. Processes, dynamics and sed- Wellington, New Zealand iments. Glacial environments: Volume 1. Butterworth Hein- emann. Oxford. 193–196. Dr Warren W. Dickinson, Research School of Earth Mercer, J.H., 1972: Some observations on the glacial geology of Sciences, Victoria University of Wellington, PO the Beardmore Glacier area. In: Adie, A. (ed.): Antarctic geology and geophysics. Universitetsforlaget. Oslo. 427–433. Box 600, Wellington, New Zealand Miall, A.D., 1978: Lithofacies types and vertical profile models in E-mail: [email protected] braided river deposits: a summary. In: Miall, A.D. (ed.): Flu- vial sedimentology. Canadian Society of Petroleum Geo- logists. Memoir 5. Calgary. 597–604. Miller, M.F. and Mabin, M.C.G., 1998: Antarctic Neogene land-

Geograﬁska Annaler · 84 A (2002) · 1 23 JAMES R. GOFF, IAN W. JENNINGS AND WARREN W. DICKINSON

scapes–In the refrigerator or in a deep freeze? GSA Today, 8: Webb, P.-N. and Harwood, D.M., 1991: Late Cenozoic glacial 1–4. history of the Ross Embayment, Antarctica. Quat. Sci. Rev., Stroeven, A.P. and Prentice, M.L., 1997: A case for Sirius Group 10: 215–223. alpine glaciation at Mount Fleming, South Victoria Land, Webb, P.-N., Harwood, D.M., McKelvey, B.C., Mercer, J.H. and Antarctica: A case against Pliocene East Antarctic Ice Sheet Stott, L.D., 1984: Cenozoic marine sedimentation and reduction. Geol. Soc. Am. Bull., 109: 825–840. ice-volume variation on the East Antarctic craton. Geology, Taylor, K.S. and Faure, G., 1983: Provenance dates of feldspar in 12: 287–291. glacial deposits, southern Victoria Land, Antarctica. In: Ol- iver, R.L., James, P.R. and Jago, J.B. (eds): Antarctic Earth Manuscript received January 2001, revised and accepted Octo- science. Australian Academy of Science. Canberra. 453–456. ber 2001.

24 Geograﬁska Annaler · 84 A (2002) · 1