Nature of the Antarctic Peninsula Ice Sheet during the Pliocene: Geological evidence and modelling results compared

Earth-Science Reviews 94 (2009) 79–94

Contents lists available at ScienceDirect

Earth-Science Reviews j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e a r s c i r ev

Nature of the Antarctic Peninsula Ice Sheet during the Pliocene: Geological evidence and modelling results compared John L. Smellie a,⁎, Alan M. Haywood a,c, Claus-Dieter Hillenbrand a, Daniel J. Lunt a,b, Paul J. Valdes b a b c

Geological Sciences Division, British Antarctic Survey, High Cross, Madingley Road, Cambridge, CB3 0ET, UK School of Geographical Sciences, The University of Bristol, University Road, Bristol, BS8 1SS, UK School of Earth & Environment, University of Leeds, Leeds LS2 9JT, UK

a r t i c l e

i n f o

Article history: Received 28 July 2008 Accepted 20 March 2009 Available online 28 March 2009 Keywords: Pliocene GCM Antarctic Peninsula Ice Sheet terrestrial marine glaciovolcanic

a b s t r a c t In this paper, we examine the nature of the Pliocene Antarctic Peninsula Ice Sheet by comparing the terrestrial and marine geological records of the Antarctic Peninsula and surrounding sea floor with estimated net snow accumulation in the region derived from numerical palaeoclimate model experiments. Pliocene geological data and our new modelling results are consistent and mutually supportive in suggesting that an ice sheet was present even during the warmest episodes of the Pliocene. The combined results suggest that the ice sheet in the Antarctic Peninsula is more robust to globally warmer conditions than is generally assumed, at least up to the climatic limits examined in our study. Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved.

Contents 1. 2. 3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Antarctic Peninsula and its ice sheet . . . . . . . . . . . . . . . . . . . . . . . . Geological evidence for Antarctic Peninsula Ice Sheet configuration . . . . . . . . . . . . 3.1. Terrestrial evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1. Northern Antarctic Peninsula: large polygenetic volcanic centres . . . . . . 3.1.2. Southern Antarctic Peninsula: multiple small monogenetic volcanic centres 3.1.3. Neogene glacial sedimentary rocks . . . . . . . . . . . . . . . . . . . . 3.2. Marine evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Modelling Pliocene climates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Palaeoclimate modelling for the Pliocene . . . . . . . . . . . . . . . . . . . . . 4.2. Pliocene boundary conditions for palaeoclimate modelling . . . . . . . . . . . . 4.3. Model predictions of Pliocene snow accumulation on Antarctica . . . . . . . . . . 5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Antarctica and the high southern latitudes have played a critical role in determining global climate change over much of the last ⁎ Corresponding author. E-mail address: [email protected] (J.L. Smellie).

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

79 80 81 81 81 84 85 87 87 88 88 89 89 92 92 92 92

100 million years (Barker et al., 1998; Gersonde et al., 1999). Paradoxically, the intricate relationships and responses of Antarctic ice volume, eustatic sea level and ocean circulation to climatic change over geological timescales are poorly resolved. Due to anthropogenic modification of climate, determination of the scale and rapidity of changes affecting the Antarctic ice mass and the links with global climate change is vital (IPCC, 2007). Numerical models of the global

0012-8252/$ – see front matter. Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.earscirev.2009.03.005

80

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

atmosphere–ocean–biosphere and cryosphere are now sufficiently mature to test hypotheses derived from geological data, as well as providing mechanistic explanations for environmental change recognised in Antarctic geological records (e.g. Florindo and Siegert, 2008, and papers therein; Francis et al., 2008). In order to evaluate such models, we must examine archives of climatic change preserved in the geological record (on the million-year timescale) to determine the relationship between ice-sheet fluctuations and climatic change. When planning to investigate past climates, regions in which large climate changes take place are generally more informative because the large signal to noise ratio means that the geological proxies used can more easily detect such changes. New proxy data acquired from these regions can enable us to carry out more rigorous tests of model simulations, resulting in better constraints on the various boundary conditions used to force the models (Lunt et al., 2008). The Antarctic Peninsula region and its ice sheet are believed to be particularly sensitive to climatic change (Barker et al., 2002). Observational records clearly demonstrate that over the last 50 years the Antarctic Peninsula has warmed substantially faster than the rest of Antarctica (Vaughan et al., 2001; Turner et al., 2002a; Vaughan et al., 2003). Thus, it seems likely that a similar climatic sensitivity probably affected the region in times past and, consequently, we might expect a more easily detectable signal in the geological proxies compared to regions characterised by greater climatic inertia. With this in mind, we have targeted the Antarctic Peninsula and, specifically, its climatically sensitive ice sheet, as an important test of modelling behaviour. Although much is known about the modern Antarctic Peninsula Ice Sheet (APIS) and its behaviour during the Quaternary, particularly since the Last Glacial Maximum (LGM) (e.g. Lucchi et al., 2002; Heroy and Anderson, 2005; Bentley et al., 2006; Domack et al., 2006; Pudsey et al., 2006; Sugden et al., 2006 and references therein), what was the nature of the APIS during pre-Quaternary warm periods? In particular what was the nature of the APIS during the last period of significantly greater sustained global warmth, i.e. the Pliocene? Existing palaeoclimate modelling experiments for the Pliocene have not been prescribed with an Antarctic Peninsula Ice Sheet. Yet the melting APIS may contribute disproportionately to global sea levels (Nakada et al., 2000; Vaughan, 2006) and we may anticipate a similar response as global warmth climbs steadily in future decades.

The premise behind the configuration of land ice cover in published Pliocene climate models is that any sea-level rise sufficiently large to require a substantial reduction in the relatively ‘stable’ East Antarctic Ice Sheet must have had a more dramatic effect on the smaller and ‘less stable’ Greenland, West Antarctic and Antarctic Peninsula ice sheets. But how realistic is that configuration in reality? Haywood and Valdes (2004) identified that the principal forcing mechanism driving globally-warmer climates during the mid Pliocene may have been reduced terrestrial ice cover combined with feedbacks from clouds and a modest increase in concentration of CO2 in the atmosphere. However, that result hinges on a realistic representation of global ice cover within the models. In addition, because of important variations in configuration of the individual interior drainage basins (e.g. Vaughan et al., 1999), the different parts of the Antarctic Ice Sheet might respond very differently to climate changes (e.g. Smellie et al., 2006b; Hill et al., 2007). In this paper, we examine the nature of the APIS during the Pliocene via examination of the geological record of the Antarctic Peninsula and offshore areas, together with snow accumulation estimates derived from numerical palaeoclimate model experiments. A particular focus is on the geological data derived from recent studies of the terrestrial outcrops, which we review in depth since they are not yet widely known but are uniquely capable of yielding otherwise unobtainable information on the thickness of past ice sheets. From our study, we are able to show that the geological proxy data and our modelling results are consistent and converge in suggesting that an APIS existed during the Pliocene warm period. 2. The Antarctic Peninsula and its ice sheet The Antarctic Peninsula is a north–south-orientated, narrow sliver of continental crust extending over approximately 12° of latitude (1700 km) and a height that decreases from about 2500 to 1000 m a.s. l. in a northerly direction (Figs. 1 and 3). The local climate is relatively warm and wet in the west but is colder and drier in the east (Martin and Peel, 1978; Morris and Vaughan, 2003). Precipitation mainly by snow is particularly high along the crest and west side of the Peninsula, where it is about 3–4 times higher than other parts of Antarctica (Turner et al., 2002b; van Lipzig et al., 2004).

Fig. 1. Synthetic oblique aerial view of the northern tip of the Antarctic Peninsula, looking northeast, showing the locations of some of the terrestrial outcrops described (Brabant Island, James Ross Island and Seal Nunataks). The tall narrow spine of the Antarctic Peninsula and its draping ice sheet are evident, as is the prominent ice-capped Graham Land plateau (see Section 5). Four times vertical exaggeration.

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

81

An ice sheet, the smallest and most northerly component of the ice sheet in Antarctica, covers the Antarctic Peninsula. It is confluent with the West Antarctic Ice Sheet at its southern extremity and has a mean thickness of about 600 m. The APIS spans an area of approximately 521,790 km2 and it differs considerably from the rest of the continental Antarctic ice sheet (Drewry, 1983). Topographically and climatically, the APIS has a closer affinity with areas such as coastal Greenland, Svalbard, Patagonia, and arguably Alaska, than with the remainder of the Antarctic Ice Sheet (Vaughan, 2006). In contrast to the homogenous ice sheet covering most of Antarctica, the APIS comprises a narrow central ice cover, broken in places by through-valleys and flanked by N400 largely independent mountain glaciers draining into ice shelves or marine tidewater glaciers, with just a few glaciers terminating on land. The assertion that the Antarctic Peninsula region is particularly sensitive to climatic change has been supported by numerous recent observations of retreating glaciers (e.g. Splettstoesser, 1992; Smith et al., 1998; Cook et al., 2005); reduction of permanent snow cover (Fox and Cooper, 1998); thickening of the ice sheet at high altitude (Morris and Mulvaney, 1996; Wingham et al., 2006; Thomas et al., 2008), and lengthening melt season (Torinesi et al., 2003). Furthermore, the retreat of ice shelves, a long-predicted consequence of warming (Mercer, 1978), is well underway (Vaughan and Doake, 1996; Scambos et al., 2001, 2003) and results in flow acceleration and thinning of nourishing glaciers (e.g. Rignot et al., 2004; Scambos et al., 2004).

lithofacies signifies that the eruptions ultimately penetrated the full ice sheet thickness; a very thick cover (i.e. thousands of metres) of ice is unlikely to melt through completely during eruptions. As the thickness of subaerial lithofacies is often relatively minor compared to the associated subaqueous or water-chilled lithofacies, it is probably acceptable in those cases to simply use the overall thickness of the erupted sequence as a proxy for ice thickness, with some adjustment for sagging. This is probably satisfactory except in cases where substantial subaerial effusion and accumulation of lava flows has continued long after an ice sheet is penetrated. True ice sheet thicknesses cannot be inferred for sequences lacking subaerial capping lithofacies and other means must be used (e.g. dissolved volatile concentrations in glass; Dixon et al., 2002). Conversely, glaciovolcanic sequences are not able to give any indication of the geographical limits of ice sheets, nor need they record ice sheet characteristics at maximum or minimum configurations, except fortuitously, and they are comparatively coarse-resolution records. The Antarctic Peninsula contains numerous late Neogene and younger alkaline and tholeiitic volcanic outcrops, most of which appear to have been erupted through a coeval ice sheet, and most are interbedded with glacial sedimentary rocks (Fig. 3; LeMasurier and Thomson, 1990; Smellie, 1999; Smellie et al., 2006b; Hambrey et al., 2008). Here, we present the first major review of glaciovolcanic sequences in the Antarctic Peninsula region, focussing on results of palaeoenvironmental importance for assessing APIS configuration.

3. Geological evidence for Antarctic Peninsula Ice Sheet configuration

3.1.1. Northern Antarctic Peninsula: large polygenetic volcanic centres Northern Antarctic Peninsula contains several large long-lived basaltic volcanic stratovolcanoes that erupted during the past 10 m.y. They include the James Ross Island Volcanic Group on the eastern flank, comprising the very large Mt Haddington stratovolcano and several smaller satellite centres with in situ outcrops dating back to 6.2 Ma but whose eruptions began as early as 9.9 Ma (Sykes, 1988; Jonkers et al., 2002; Smellie et al., 2008), and Brabant Island on the western flank of the Peninsula, which is draped by products of at least three large overlapping basaltic centres b200 ka in age (Smellie et al., 2006a). The James Ross Island outcrops are dominated by spectacular voluminous (tens of km3) far-travelled lava-fed deltas that mainly erupted in a glacial (i.e. ice sheet or ice cap) setting, although at least one delta and several tuff cones were marine-emplaced (Smellie et al., 2006b, 2008). By contrast, the individual eruptive units on Brabant Island are smaller-volume glaciovolcanic sheet-like sequences of Mt Pinafore type (sensu Smellie, 2008). For the large northern stratovolcanoes, late Miocene–Pleistocene ice sheet thicknesses of 200–350 m were typical during eruptions of the James Ross Island group for much of the period but with rare episodes of thicker ice also represented, up to c. 750 m thick (Fig. 4; Table 1; Smellie et al., 2008). Most of the dataset corresponds to the Pliocene epoch, and the trend of inferred ice thicknesses increases throughout the period represented. Smellie et al. (2008) concluded that the APIS was never more than 750 m thick in the James Ross Island region. That suggestion also relied on the presence of rare high mesa surfaces (Nc. 600 m a.s.l.), corresponding to the top of a single voluminous lava-fed delta formed at 4.69 Ma (corrected weightedmean age recalculated from that published by Smellie et al., 2008). Those surfaces were apparently unmodified by erosive glacial activity and thought never to have been overridden. New observations obtained in January 2008 indicate that glacial erosion is patchily present on the delta top, comprising striated surfaces in topographic hollows that were snow filled during the previous visit. The mesa must therefore have sustained an ice cover of uncertain thickness subsequent to 4.69 Ma. However, relict primary volcanic features are widespread and preserved in near-pristine condition on the mesa top (e.g. an exposed unmodified delta front slope, tumuli and delicate flow-top features), the extent of the mesa-top glacial erosion is minor despite having formed nearly 5 m.y. ago and, although the entire mesa

3.1. Terrestrial evidence Although the marine (shelf) limits of the APIS at LGM are now relatively well known (e.g. Anderson et al., 2002; Heroy and Anderson, 2005; Sugden et al., 2006), cosmogenic exposure ages have provided few constraints for its thickness at terrestrial sites (Bentley et al., 2006), and very little is known about the former extent and the thickness of the APIS prior to LGM (e.g. Bart, 2001; Barker et al., 2002). A glacial trimline has been documented at a single locality in Palmer Land. Although undated, it is assumed to relate to the APIS at LGM (Bentley and Anderson, 1998). However, trimlines simply record the maximum elevation of warm-based ice within an ice sheet (Goodfellow, 2007) and the assumption of an LGM age is difficult to sustain for the ice cover there, which currently, and possibly for several m.y., has a polar thermal regime and was therefore essentially non-erosive. The feature may be a relict related to a much older glacial period. Thus, the interior configuration of the APIS is insufficiently constrained for the LGM and effectively unknown for any pre-LGM period. Acquiring estimates of past ice thickness has proved particularly elusive. However, volcanoes that interact with ice sheets frequently preserve a record that can be used to deduce critical parameters of former ice sheets, including ice sheet thickness. How those parameters are acquired, using analysis of the lithofacies and sequence architecture, is summarised by Smellie (2000, 2001, 2006, 2008; see also examples by Tuffen et al., 2002; Schopka et al., 2006; McGarvie et al., 2007). Once a glaciovolcanic origin is determined for a sequence, it is a relatively simple matter to determine former ice sheet thickness (Fig. 2). The measured total thickness of subaqueous or water-chilled lithofacies overlying a glaciated substrate and/or basal glacial sediments in a monogenetic sequence, i.e. formed during a single volcanic eruption, is a known proxy for the thickness of the coeval ice sheet, but it must be adjusted upward by about 100 m to allow for syneruption sagging of the ice surface toward the volcanic units (Smellie and Skilling, 1994; Smellie, 2006; Schopka et al., 2006; Fig. 2). Even after adjusting, such estimates strictly only yield minimum thicknesses for the coeval ice cover. However, the true total ice thickness cannot be much greater since the presence of capping subaerial

82

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

Fig. 2. Diagram showing field photographs, cross sections and reconstructed 3D profile views of two fundamentally different glaciovolcanic sequence types that are characteristic of glaciovolcanic outcrops in the Antarctic Peninsula (e.g. Smellie et al., 1993, 2006a, 2008), and their inferred environmental settings. A — Longitudinal section through a far-travelled lava-fed delta rooted in a monogenetic eruption of tuya type (cf. Smellie, 2000, 2007), characteristic of many basaltic eruptions in relatively thick ice; features that distinguish lavafed deltas in a glacial setting are described by Smellie (2006). B — Transverse section through a sheet-like sequence of Mt Pinafore type (Smellie, 2008) characteristic of basaltic eruptions in relatively thin ice. Sagging of the ice towards the volcanic lithofacies is exaggerated here for clarity. Note the scale differences between the two sequences, the different major constituent lithofacies in each and their distinctive architectures.

surface has not been fully examined, no glacial erratics have been observed. The observations suggest that the former ice cover was only just thick enough to erode its substrate, and Johnson et al. (submitted

for publication) calculated empirical maximum thicknesses of c. 45– 200 m overlying the surfaces, yielding a total potential thickness of c. 850 m for the associated ice cover, measured down to the local pre-

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

83

Fig. 3. Map of the Antarctic Peninsula region showing the locations of late Neogene terrestrial outcrops (after Smellie,1999) and DSDP and ODP marine drillsites (after Barker et al., 2002).

delta bedrock surface. Moreover, cosmogenic nuclide dating of the surfaces has determined a maximum cumulative exposure history of 15.3 ka, consistent with an essentially persistent snow or ice cover on the surfaces since they formed (Johnson et al., submitted for publication). Thus, it seems that the APIS was generally persistent and thin, i.e. never substantially thicker than the limit inferred originally by Smellie et al. (2008). Additional support that former ice never exceeded 850 m in thickness comes from the presence of a crater that has recently (since 2001) melted out at the summit of Corry Island (west of James Ross Island, in Prince Gustav Channel). The crater is a pyroclastic cone that marks the vent for eruptions of the small monogenetic tuya volcano that ultimately formed the island (Smellie et al., 2006c). Despite a collapse that has breached the crater on its north side, it appears to be a pristine primary volcanic landform that has not been eroded by overriding ice. The island is undated but unlikely to differ significantly in age from nearby islands, which are also small independent monogenetic eruptive centres with ages 1.6–2.7 Ma (Smellie et al., 2006c). Corry Island reaches 510 m in elevation and sits on seafloor at least c. 500 m deep that was affected by Quaternary outlet glacier activity in Prince Gustav Channel (Camerlenghi et al., 2001) responsible for the cliff coastline of the island. Although the eruptive environment of the Corry Island volcano is currently unknown, any erosive ice present since the formation of the island cannot have exceeded c. 1000 m in thickness, consistent with inferences based on James Ross Island nearby.

The late Pleistocene (b200 ka) glacial cover associated with the Brabant Island centres was even thinner than that which affected James Ross Island. Measured thicknesses of the erupted units are up to 140 m, which probably indicates a maximum ice sheet thickness of about 200 m when ice sheet sagging caused by volcanic heat is taken into account (Table 2; Smellie et al., 2006a). Thus, both the eastern (James Ross) and western (Brabant) volcanic centres would have been covered by ice domes or local ice caps that, at times, would have been confluent with the remainder of the APIS draping the spine of the Peninsula. Confluence of Antarctic Peninsula and James Ross Island ice in the Prince Gustav Channel area is supported by the distribution of Antarctic Peninsula-derived erratics on James Ross Island, by orientations of glacial striations, and by swath imagery of the Channel itself which shows ice stream features associated with the APIS at LGM (Camerlenghi et al., 2001; Ingólfsson, 2004; Evans et al., 2005; Smellie et al., 2006b; Hambrey et al., 2008). The domes or ice caps on Brabant and James Ross islands, and along the central spine of the Peninsula, would have dominated the local glacial morphodynamics (Smellie et al., 2006a, 2008). This scenario also resembles the LGM conditions in the Palmer Deep region (directly to the SW of Anvers Island), for which Domack et al. (2006) reconstructed growth and coalescence of local ice domes rather than ice overriding by a single, huge APIS. There is thus no clear evidence for a “giant” APIS that drowned the topography and straddled the Peninsula from shelf edge to shelf edge (see also Section 5).

84

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

Fig. 4. Summary of ice thickness data derived from glaciovolcanic outcrops in the Antarctic Peninsula (data in Tables 1 and 2). Open symbols denote outcrops in which subaerial capping lithofacies are absent and for which only minimum ice thicknesses can be inferred. Measured thicknesses of the monogenetic sequences are indicated on the left axis. They are converted to inferred maximum thicknesses of the associated ice sheet by adding 100 m (right axis). The longest and most complete record by far is that for the James Ross Island region (emphasized here by shaded fields and regression lines), in which bedrock surface gradients were essentially similar and low, and for which a dominant climatic influence (rather than steep bedrock gradients) on ice thickness is likeliest. Although most of the data plotted for Seal Nunataks reflect minimum ice thicknesses, data for other nunataks in the group, in which subaerial capping lithofacies are preserved but which are currently undated, indicate that the range of thicknesses is probably a reasonably accurate reflection of the original ice cover. Data for Brabant Island and Elgar Uplands are accurate but may be strongly affected by relatively steep bedrock gradients as much as climatic conditions (cf. Smellie et al., 2006a). Two estimates for ice thickness are shown for Sims Island because of uncertainty about the full sequence thickness (see text). The empirical division into a field of “normal ice thicknesses” identified by Smellie et al. (2008), which comprise most of the data and which probably include some periods of ice-poor conditions (so-called interglacials), and uncommon “exceptional ice thicknesses”, is also shown by the new expanded dataset. Regressions are shown for the James Ross Island dataset (after Smellie et al., 2008) but do not change significantly for the full dataset. The regressions suggest that, overall, ice sheet thicknesses increased progressively from Miocene times. The full dataset also clearly suggests that eruptions typically took place when the APIS was only a few hundred metres thick (b400 m), and less commonly when it was substantially thicker (i.e. 750–850 m). Although the maximum thickness attained is not accurately defined by the available data, there are indications that it may not have exceeded that limit by much (see Section 3.1.1).

3.1.2. Southern Antarctic Peninsula: multiple small monogenetic volcanic centres Isolated outcrops of volcanic rocks formed by eruptions of multiple small monogenetic centres in several volcanic fields are scattered across the southern half of the Antarctic Peninsula (Seal Nunataks, southern Palmer Land, Alexander Island, and eastern Ellsworth Land; Fig. 3). They are of similar age (b8 Ma) to the large northern stratovolcanoes (LeMasurier and Thomson, 1990; Smellie, 1999). The outcrops comprise a mixture of lava-fed deltas, tuff cones and pillow volcanoes (corresponding to centres of tuya, tindar and pillow volcano types, respectively), sheet-like sequences and rare subaerial cinder cones (Table 2; Smellie et al., 1993; Smellie and Skilling, 1994; Smellie and Hole, 1997; Smellie, 1999; terminology explained by Smellie, 2007). Features of the southern outcrops indicate the former presence of ice no greater than 200 m thick (and probably significantly less) in latest Miocene times (c. 7.7–5.4 Ma) in northern Alexander Island (Elgar Uplands; Table 2; Fig. 4). They are sheet-like sequences of Mt Pinafore type (sensu Smellie, 2008), whose thinness might be due to a combination of high bedrock elevation and moderately steep prevolcanic bedrock slopes influencing snow accumulation and ice flow as much as climatic factors. Coeval ice thickness at lower elevations nearby may have been somewhat thicker (i.e. more than 200 m), probably because bedrock gradients were much lower, but the outcrops are too eroded to make a meaningful estimate of original ice thickness. This is suggested by the presence of two small outcrops on Rothschild Island, each comprising likely subaqueous tuff cone relicts, which probably formed originally within two separate water-filled englacial vaults at 5.4 Ma (Smellie, 1999; cf. Smellie and Hole, 1997). A late Pliocene outcrop in northern Alexander Island, at Hornpipe Heights, was erupted entirely subaerially and must have been situated above any contemporary ice sheet. The local underlying pre-volcanic

bedrock slopes are high (exceeding 55°) and were probably too steep to sustain a draping ice cover. Even today, the bedrock ridge is snowand ice-free, although a large drainage glacier, Sullivan Glacier, is present close to the base of the volcanic outcrop. The Hornpipe Heights outcrop has an age of 2.6 Ma coincident with at least one dated much lower-elevation centre on Beethoven Peninsula, southern Alexander Island, at Mussorgsky Peaks. The Beethoven Peninsula outcrops are relicts of several relatively large monogenetic edifices, individually at least 2 km in original diameter. Few are dated but they are probably Plio-Pleistocene in age and were erupted in association with a substantial ice sheet that may have been 550 to possibly N800 m thick (Table 2; Fig. 4; cf. Smellie and Hole, 1997). Sims Island is another Pliocene centre situated to the south of Alexander Island (Fig. 3). It has a likely age of 2.30 Ma (by 40Ar/39Ar) and appears to be a subglacially erupted volcano that stalled at the early pillow-dominated stage of volcano construction (cf. Smellie and Hole, 1997), when the coeval ice sheet thickness was N400 m (based on the height of the island; Hathway, 2001). However, an ice thickness exceeding 700 m can possibly be inferred based on estimated depth to pre-volcanic bedrock assumed to crop out on the sea floor adjacent (Smellie, 1999). Snow Nunataks comprise several small tuya and tindar volcanic centres similar to those on Beethoven Peninsula. The ages of the centres and sequence thicknesses are poorly known, but they are probably Plio-Pleistocene (i.e. c. 4.6–1.6 Ma; Table 2), and coeval ice thicknesses may have varied between 400 and N700 m. Only one of the several volcanic centres in the Seal Nunataks is known to be of Pliocene age: Oceana Nunatak, c. 2.9–3.0 Ma (Smellie and Hole, 1997). A glaciovolcanic origin is unproven but thought to be likely and, if true, an ice sheet several hundred metres thick (c. 470 m?) is plausible. Some support for this estimate is provided by volcanic units of Pliocene age on James Ross Island b200 km NE, which suggest broadly comparable ice sheet thicknesses (Fig. 4). Most of the volcanic

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94 Table 1 Estimates of the thickness of the Antarctic Peninsula Ice Sheet, using measurements of glacially-emplaced lava-fed deltas in the James Ross Island region. Based on Smellie et al. (2008), updated; all ages by 40Ar/39Ar. Locality and delta name (where different)

Age (Ma)

Delta foresets thickness (m)

Maximum measured whole-delta thickness (m)

Inferred maximum ice sheet thickness (m)

Dobson Dome Dundee Island (Cape Purvis) Vega Island (Keltie Head) Tabarin Peninsula (Brown Bluff) Tabarin Peninsula (7 Buttresses) Jefford Point Vega Island (Vertigo Cliffs, upper) Sungold Hill (Sungold) Cape Gage, upper delta St Rita Point Forster Cliffs Flett Crags Rhino Corner, local delta #4 Cape Gage middle delta, N of St Rita Point Cape Gage middle delta Rhino Corner, local delta 2 Cape Gage basal delta, N of St Rita Point Cape Gage basal delta, Cape Gage area Hamilton Point Lachman Crags, upper delta Patalamon Mesa upper (main) delta, at Patalamon Mesa Patalamon Mesa upper (main) delta, Back Mesa Patalamon Mesa upper (main) delta, Crisscross Crags Patalamon Mesa upper (main) delta, N&NW Davis Dome Tumbledown Cliffs delta; Crisscross Crags (basal delta) Tumbledown Cliffs delta; Back Mesa-Holluschickie Bay area Tumbledown Cliffs delta; Cape Obelisk Tumbledown Cliffs delta; at Tumbledown Cliffs Rink Point main delta, at NW Davis Dome Rink Point main delta, at Kipling Mesa Vega Island (Sandwich Bluff) Stickle Ridge lower delta, at Stickle Ridge Stickle Ridge lower delta, at Smellie Peak

0.08 0.13 0.99 1.22

N 601 N 230 N 180 N 350

650 250 200 400

750 350 300 500

1.69

N 150

200

300

1.89 2.03

N 150–250 0–180

400 250

500 350

2.03 2.23 2.50 2.50 2.68 3.01 3.08

193 153 191 8–352 N 515 215 121

240 223 221 360 600 255 146

340 323 321 460 700 355 246

3.08 3.51 3.52

7–86 38–66 N 36–103

126 75 190

226 175 290

3.52

N 70–165

202

302

3.69 3.95 4.61

80 5–81 0–N 200

120 81 253

220 181 353

4.61

250

270

370

4.61

56

107

207

4.61

2–25

43

143

5.15

110

110

210

5.15

N 104–N149

149

249

5.15

140

155

255

5.15

125

125

225

5.36

50–100

125

225

5.36

68

109

209

5.42 6.16

N 325 150

475 173

575 273

6.16

N 87

160

260

centres in Seal Nunataks are Pleistocene in age (1.5– b 0.2 Ma). They may have formed in association with an ice sheet variably 300–500 m thick (maximum N 570 m; Table 2), again similar to equivalent-age deltas in the James Ross Island region. However, the maximum surface elevation of the Pleistocene ice sheet at Seal Nunataks was seldom greater than 200 m above present datum in Pliocene and Pleistocene times, reflecting the very low pre-volcanic bedrock surface on which most of the volcanic centres were constructed (probably mainly c. 80– 100 m below sea level; Table 2). The presence of erratics on some nunataks indicates that a thickened APIS overrode them at times (Smellie and Hole, 1997). A Pleistocene ice sheet on Beethoven Peninsula is represented by the Gluck Peak volcanic centre, which erupted at b1 Ma (Smellie, 1999). It may have been N435 m thick.

85

Similar to interpretations of the James Ross Island sequences, comparison of inferred coeval ice sheet thickness derived from the southern volcanic outcrops suggests that ice sheet thickness increased from older to younger periods (Fig. 4). However, the interpretation is not so clear-cut since the outcrops are of two main types: monogenetic glaciovolcanic centres of tindar and tuya type (Seal Nunataks, Beethoven Peninsula, Rothschild Island, Snow Nunataks, Sims Island), and less common sheet-like sequences (northern Alexander Island). The latter are restricted to high elevations with steep pre-volcanic bedrock gradients that might influence the accumulation and thickness of any glacial cover, and its hydraulics, as strongly as any climatic effects. 3.1.3. Neogene glacial sedimentary rocks Sedimentary beds, mainly glacial diamicts and conglomerates but also better bedded glaciomarine deposits, are interbedded with the Neogene volcanic rocks on James Ross and Brabant islands, and in northern Alexander Island. They provide additional information, particularly on the extent, thermal regime and basal coupling of the associated ice sheet. More than seventy outcrops have been discovered in the James Ross Island region alone (Dingle et al., 1997; Pirrie et al., 1997; Jonkers et al., 2002; Hambrey and Smellie, 2006; Smellie et al., 2006b; Hambrey et al., 2008). The coeval volcanic rocks have provided an unusually good chronological record mainly by 40Ar/39Ar dating (Smellie et al., 2008), and Sr isotopic ages have been obtained on shell material in the deposits on James Ross Island. The shells and other fossil fauna probably required the sea surrounding James Ross Island to be at least seasonally ice-free, thus corresponding to “interglacial” periods. Early Pliocene ice-poor periods (c. 5–4 Ma) are particularly well represented in the Sr isotopic dataset, but with relatively large analytical uncertainties on the ages (Jonkers and Kelley, 1998; Jonkers et al., 2002; Smellie et al., 2006b; McArthur et al., 2007). There is also much scarcer evidence in the associated volcanic lithofacies for local marine conditions, thought to relate to ice-poor periods, preserved in several marine-emplaced tuff cones and possibly two of the lava-fed deltas, with ages (where known) of 5.90, 5.06, 3.69 and 0.66 Ma (Smellie et al., 2006b; Williams et al., 2006; Smellie et al., 2008; Nelson et al., 2008). The commonest diamict lithofacies contains bullet-shaped, facetted and striated clasts and, together with striated bedrock surfaces and distinctive environmentally diagnostic features of some of the volcanic lithofacies, suggest erosive conditions and a wet-based glacier thermal regime that persisted at least until 1 m.y. ago (Smellie, 2006; Smellie et al., 2008; Hambrey et al., 2008). Even some presentday glaciers show signs of being wet-based, at least locally (Lundqvist et al., 1995), although most are probably polythermal (Chinn and Dillon, 1987). This is consistent with the observation that almost all of the Neogene volcanic outcrops in the James Ross Island region are conspicuously glacially eroded (i.e. by wet based ice), even those as young as 132 ka (Smellie et al., 2006c). Rare exceptions are apparently pristine volcanic landforms, including a large satellite centre at Dobson Dome (b80 ka), and small cinder cones on the eastern margins of the Mount Haddington ice cap and on Paulet Island (b1 ka?; Smellie et al., 2006c, 2008). In summary, the burgeoning geological evidence from terrestrial outcrops scattered along the length of the Antarctic Peninsula suggests that it sustained a wet-based erosive ice sheet during the late Miocene–Pliocene period, and possibly continuing into the Quaternary. Moreover (and notwithstanding orbitally-influenced fluctuations), the trend of ice thicknesses apparently indicates a progressive increase from late Miocene times (regression lines in Fig. 4). There is evidence for several ice-poor periods in the James Ross Island group, particularly clustered in the early Pliocene. However, practically all of the late Neogene volcanic units in the Antarctic Peninsula indicate interactions with a coeval ice sheet or ice cap. Moreover, the complete absence of lavas erupted in a wholly subaerial environment (rather

86

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

Table 2 Estimates of the thickness of the Antarctic Peninsula Ice Sheet, using measurements of glaciovolcanic sequences in the Antarctic Peninsula region, excluding James Ross Island. Outcrop exposed thickness (m)a

Outcrop summit elevation (m)b

Bedrock elevation (m)

Ages (Ma)c

Preferred age (Ma)

Whole sequence thickness (m)d

Inferred maximum ice sheet thickness

Outcrop type/dominant lithofaciese

140 75 100

n/a n/a n/a

n/a n/a n/a

0.19 Undateable (b40 ka?) Undateable (b40 ka?)

0.19 0.04 0.04

140 75 100

240 175 200

Mt Pinafore-type sheet-like sequence Mt Pinafore-type sheet-like sequence Lava-fed delta

90

90

− 100

0.7; b 0.1

0.7

190

290

Arctowski

150

235

− 100

1.4

1.4

335

435

Bruce

200

320

− 100

1.5

1.5

420

520

Bull Castor

200 60–100

200 155

− 100 − 100

None None

300 255

400 355

Christensen Dallmann

b 100 Minor

299 210

− 100 − 100

None None

400 310

500 410

Donald Evensen Gray Hertha

130 150 b 150 b 50

130 173 150 226

− 100 − 100 − 100 − 100

b 0.2 (4); 1.4 b 0.2 none

0.2 1.4 0.2

230 273 250 326

330 373 350 426

Larsen

130

151

− 100

1.5; b0.1

1.5

251

351

Lindenberg Murdoch

130 130

130 368

− 100 − 100

None None

230 468

330 568

Oceana

200

271

− 100

3.0; 2.8

2.9

371

471

All scree N 100

34 Unknown

− 100 Unknown

None 5.4

5.4

134 100

234 200

Pillow volcano: mainly feeder dyke and pillow lava⁎ Pillow volcano: mainly feeder dyke and pillow lava Tindar: feeder dykes, subaqueous tuff cone, pillow lava Feeder dyke and subaerial(?) lava⁎ Subaerial lavas capping tuff cone (subaerial and probably subaqueous)⁎ Subaerial(?) tuff cone⁎ Pillow volcano: mainly feeder dyke and pillow lava⁎ Cinder cone deposits⁎ Dyke; pillow lava Dyke/sill; tuff cone; pillow lava Subaerial lavas & minor pyroclastic rocks⁎ Pillow volcano: mainly feeder dyke and pillow lava Dyke & minor tuff cone deposits⁎ Dykes, pillow lava, minor tuff cone deposits⁎ Dykes, subaerial pyroclastics, subaqueous tuff cone⁎ Dyke(s)? Tindar?: subaqueous tuff cone

c. 25 35 40 90 100

n/a n/a n/a n/a n/a

n/a n/a n/a n/a n/a

6.6; 6.3 5.4; (3.9) 6.2; 6.0 7.1; 6.9 7.7; 7.6; 7.3

6.15 5.4 6.1 7 7.6

25 35 40 90 100

125 135 140 190 200

Mt Mt Mt Mt Mt

250–300

500

50

2.5

2.5

450

550

50 300–400

500 800

20 50

None None

480 750

580 850

b 50 b 50 N 380

817 335 380

50 10 Unknown

None b1 3.46; 2.30

1 2.3

767 335 400

867 435 500

Tuya: lava-fed delta outcrop on subaqueous tuff cone Tindar: subaqueous tuff cone Tuya: lava-fed delta outcrop on subaqueous tuff cone Tindar: subaqueous tuff cone Tindar: subaqueous tuff cone Pillow volcano

300

300

0?

20 (unreliable?)

–

300

400

Mt Benkert

350

600

0?

4.7; 4.6

4.6

600

700

Mt Thornton

425

425

0?

1.7; 1.6

1.6

425

525

Locality

Brabant Island “Cairn Point” Driencourt Point Metchnikoff Point Seal Nunataks Akerlundh

Pollux Rothschild Island Elgar Uplands KG.2431 KG.3616 KG.3617 KG.2223 KG.3606 Beethoven Peninsula Mussorgsky Peaks Mt Liszt Mt Greig Mt Strauss Gluck Peak Sims Island Snow Nunataks Mt McCann

Pinafore-type sheet-like Pinafore-type sheet-like Pinafore-type sheet-like Pinafore-type sheet-like Pinafore-type sheet-like

sequence sequence sequence sequence sequence

Tuya?: subaerial lava on subaqueous tuff cone and pillow lava Tindar: subaqueous tuff cone & minor pillow lava Tindar: subaqueous tuff cone & minor pillow lava

Notes: a Exposed thicknesses from Smellie et al. (1993), Smellie and Hole (1997) and unpublished. b Outcrop summit elevations from BAS Mapping & Geographic Information Centre (various published and unpublished sources); bedrock elevations from BEDMAP (Lythe and Vaughan, 2001). c Isotopic ages by K–Ar (from Smellie, 1999), except Brabant Island (by 40Ar/39Ar; Smellie et al., 2006a) and Sims Island (also by 40Ar/39Ar; Hathway, 2001); Snow Nunataks ages (by K–Ar) are unpublished (personal communication from JW Thomson, 2000). d Sequence thicknesses calculated from difference between summit and bedrock elevations, assuming continuity of outcrop, except for sheet-like sequences (Brabant Island; Elgar Uplands), which are measured thicknesses. e Most of the Seal Nunataks outcrops also have minor capping subaerial lithofacies (see Smellie and Hole, 1997), marked here by asterisk in “outcrop type” column; most of those nunataks are currently undated.

than simply as subaerial capping units of lava-fed deltas) on James Ross Island during a period of 6.2 m.y. that included N50 eruptions (Smellie et al., 2008) also suggests that there was never a time that was entirely ice-free. This is perhaps unsurprising for the past 3 m.y., when δ18O values of benthic foraminifera were typically higher than today, cor-

responding to lower global temperatures and/or greater ice volumes (e.g. Lisiecki and Raymo, 2005, Fig. 4). However, it also implies that an ice cover was present in the Antarctic Peninsula region back to 6.2 Ma, at least, even when δ18O was lower and the climate warmer and/or ice volume less. Therefore, we suggest that the glaciovolcanic evidence,

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

together with results of the cosmogenic surface dating study by Johnson et al. (submitted for publication), imply that an ice cover was present and remarkably persistent in the region, although fluctuating in size, since latest Miocene times and that even the warmer so-called interglacials were probably ice-poor rather than ice-free (cf. Smellie et al., 2006b; Smellie et al., 2008). The notion of a persistent ice sheet, despite climatic fluctuations (see also Cowan, 2002; Hillenbrand and Ehrmann, 2002; Hepp et al., 2006; Cowan et al., 2008; Section 3.2), conflicts with the boundary condition used in most conventional climate modelling reconstructions, in which an ice-free Antarctic Peninsula is prescribed for models of the Pliocene warm period. However, it is fully consistent with our new modelling results (Section 4), which persistently show that an ice cover would have been able to form during the warm mid-Pliocene. During the ice-poor “interglacials”, the APIS may have shrunk back toward the present coastline and open-water marine conditions may have prevailed at times, at least seasonally. 3.2. Marine evidence The results of marine geological and geophysical studies on the Pacific margin of the Antarctic Peninsula, e.g. Ocean Drilling Program (ODP) Leg 178 (Barker et al., 1999; Barker et al., 2002), have also contributed to the reconstruction of Pliocene environmental conditions in the Antarctic Peninsula region. Reflection seismic profiles have revealed that the outer shelf is underlain by prograding and aggrading sequences, which were interpreted to consist of subglacial and glacial-marine sediments younger than c. 6.5 Ma (Larter et al., 1997). Sedimentary successions drilled at ODP Leg 178 Site 1097 on the outer shelf confirmed a glacial (-marine) origin and a late Miocene to Pliocene age of these seismic units (Fig. 3; Barker et al., 1999; Eyles et al., 2001; Barker et al., 2002). Moreover, the progradational and aggradational geometry of seismic sequences, which were dated on the basis of the age model for Site 1097, shows that grounded ice may have advanced onto the outer shelf throughout the Pliocene (Bart and Anderson, 2000; Bart, 2001). This conclusion is corroborated by results of mineralogical and granulometric investigations on Pliocene sediments recovered from large drift deposits located on the adjacent continental rise (Leg 178 sites 1095, 1096, and 1101), which indicate repeated advance and retreat of grounded ice masses across the shelf and supply of iceberg-rafted debris (IRD) eroded on the Antarctic Peninsula (Cowan, 2002; Hassler and Cowan, 2002; Hillenbrand and Ehrmann, 2002; Pudsey, 2002a,b; Hillenbrand and Ehrmann, 2005; Hepp et al., 2006). New studies on sediments from ODP Leg 178 Site 1101 have also revealed that IRD deposited during the last c. 1.4 m.y. is less weathered, but more crushed and abraded than IRD deposited during the Pliocene, which might be attributed to a cooler climate and/or reduced supraglacial transport of clasts, possibly as a consequence of a thicker APIS (Cowan et al., 2008). Analysing a seismic profile from the Pacific margin of the Antarctic Peninsula, Bart and Anderson (2000) and Bart (2001) found a significantly higher number of late Neogene ice sheet grounding events compared to the Ross Sea and Prydz Bay. The authors speculated that the grounding events were better preserved on the western Antarctic Peninsula shelf because of its higher subsidence rate (cf. Larter et al., 1997). Alternatively, the APIS may have been glaciologically more dynamic than the East Antarctic Ice Sheet (EAIS) and West Antarctic Ice Sheet (WAIS; Bart and Anderson, 2000; Smellie et al., 2006b). The latter idea is consistent with the suggestion of Barker (1995) that subglacial sediment transport to the shelf edge might have actually increased around Antarctica during the warm Pliocene period compared to present, because of greater ice accumulation and faster outward flow facilitated by more warm-based ice. Barker's hypothesis was based on the observation that sedimentation rates at seven Deep Sea Drilling Project (DSDP) and ODP sites on the continental rise around Antarctica were higher during the early to middle Pliocene

87

than before or thereafter. However, sedimentation rates at the more recently drilled ODP Site 1095 west of the Antarctic Peninsula and Leg 188 Site 1165 on the continental rise offshore from Prydz Bay proved to be higher during the late Miocene and the early to late Miocene, respectively, than during the supposedly warmer Pliocene (Barker et al., 2002; Cooper et al., 2004). Nevertheless, the idea that Antarctic ice sheets may have grown under generally warmer climatic conditions in response to higher precipitation caused by intensified evaporation over the surrounding Southern Ocean may be correct. This socalled “snow gun hypothesis”, which was initially proposed by Prentice and Matthews (1991), was corroborated by a glaciological model revealing ice sheet growth, for global warming of up to 5 K (Huybrechts, 1993). Increased precipitation in Antarctica (particularly the Antarctic Peninsula) in response to projected warming during the 21st century is also reproduced in the most recent climate change models (e.g. Bracegirdle et al., 2008, and references therein). Relatively high precipitation may have been fuelled by a reduced or even absent seasonal sea-ice coverage around the Antarctic Peninsula, enlarging open ocean areas and thus favouring more evaporation. Indeed, sea ice reduction or absence during the early and middle Pliocene is suggested by relatively high contents of biogenic silica (Hillenbrand and Fütterer, 2002; Hillenbrand and Ehrmann, 2005) and diatoms (Pudsey, 2002a,b) in the corresponding sections at sites 1095, 1096, and 1101. Interestingly, some support for high precipitation in the Pliocene is provided by our model results (next section). Our model runs reveal that during the mid-Pliocene the snow accumulation rate over the APIS was similar to or higher than today (Section 4). Recently, Rebesco et al. (2006) interpreted seismic reflection profiles from the Pacific continental margin of the Antarctic Peninsula to indicate a change in the style of margin accretion at ca. 3 Ma. The authors claimed that a regional change in margin architecture was caused by the transition of the APIS from a sub-polar to a polar glacial regime, i.e. from wet-based to cold-based conditions. The transition, which was linked with global cooling, significantly reduced the volume and rate of discharge of sediments being eroded from the Antarctic Peninsula, and also affected the EAIS and WAIS during the late Pliocene (Rebesco et al., 2006; see also Barker, 1995). However, these interpretations and conclusions are controversial (cf. Larter, 2007; Rebesco et al., 2007). 4. Modelling Pliocene climates A significant number of palaeoclimate modelling studies have been published for the Pliocene, specifically for the period of mid Pliocene warmth (defined narrowly as ~3.3 to 3 Ma BP; e.g. Chandler et al., 1994; Sloan et al., 1996; Haywood et al., 2000a,b, 2001, 2002a,b,c; Haywood and Valdes, 2004; Haywood et al., 2005; Haywood and Valdes, 2006). These experiments represented significant steps forward in the sophistication of the modelling techniques employed to study the climate of the mid Pliocene. All of these modelling studies have similarities: they used boundary conditions prescribed from the PRISM Project (Pliocene Research Interpretations and Synoptic Mapping; see http://geology.er.usgs.gov/ eespteam/prism/index.html). The PRISM project has two principal aims. The first is to identify and characterise the nature and variability of climate during the mid Pliocene warm interval. The second goal is to develop a series of global-scale, quantitative datasets for use in experiments to model climate and environmental conditions during the mid Pliocene (Dowsett et al., 1999). The ambitious premise behind this work is to test the ability of climate models to simulate past warmer conditions on Earth and to provide insights into the mechanisms and effects of global warming (Dowsett et al., 1992; Chandler et al., 1994; Sloan et al., 1996). PRISM-based reconstructions of mid Pliocene boundary conditions are organised as a series of datasets representing different

88

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

environmental attributes. The datasets were originally designed for use with the Goddard Institute for Space Studies (GISS) Model II AGCM using an 8° × 10° resolution (Hansen et al., 1983). The first step for PRISM in documenting the Pliocene climate involved the assignment of an appropriate fraction of land versus ocean to each grid box. Following grid cell by grid cell, land versus ocean allocations, winter and summer sea-ice coverage, and then winter and summer seasurface temperature (SST) were assigned to open ocean areas. Average land ice cover was recorded for land areas and then land areas not covered by ice were assigned proportions of six vegetation or land cover categories modified from Hansen et al. (1983). Numerous records exist which suggest higher-than-present sea levels during the Early and mid Pliocene. Estimates range from +25 to +60 m (Haq et al., 1987a,b; Dowsett and Cronin, 1990; Zubakov and Borzenkova, 1990; Krantz, 1991; Brigham-Grette, 1994). In the most recent complete dataset, the PRISM2 reconstruction, a relatively conservative sea-level estimate of +25 m compared to present-day was used. PRISM2 uses model results from Prentice (pers. commun. to Harry Dowsett; cited in Dowsett et al., 1999) to guide the areal and topographic distribution of Antarctic and Greenland ice. To accomplish a +25 m sea level, the Greenland Ice Sheet was reduced in size by 50%, and all other Northern Hemisphere ice (e.g. Iceland and mountain glaciers) was completely removed. The volume of the EAIS was also reduced by ~33% (Dowsett et al., 1999). Finally, and more significantly in present context, the APIS was removed completely, as was the WAIS. 4.1. Palaeoclimate modelling for the Pliocene The particulars of the version of the UKMO GCM (hereafter referred to as HadCM3) used in this study are well documented (Gordon et al., 2000). The model is one of a new breed of coupled Ocean–Atmosphere GCMs (OAGCM) that requires no flux corrections to be made, even for simulations of a thousand years or more (Gregory and Mitchell, 1997). The GCM consists of a linked atmospheric model, ocean model, sea ice model and vegetation model. The horizontal

resolution of the atmospheric model is 2.5° in latitude by 3.75° in longitude. This gives a grid spacing at the equator of 278 km in the North–South direction and 417 km East–West and is approximately comparable to a T42 spectral model resolution. The atmospheric model consists of 19 layers. The spatial resolution over the ocean in HadCM3 is 1.25° × 1.25° and the model has 20 layers. For simulations of the present-day, HadCM3 has a much-improved SST and sea ice climatology compared to earlier generations of the model, whilst the atmospheric part of the model is capable of a realistic simulation of the surface heat-flux (Gordon et al., 2000). 4.2. Pliocene boundary conditions for palaeoclimate modelling The required boundary conditions were supplied by the United States Geological Survey (USGS) PRISM2 & PRISM3D 2° × 2° digital datasets. The particulars of the PRISM2 dataset have been well documented in previous papers (Dowsett et al., 1999; Haywood et al., 2000a and references therein). In brief, the prescribed boundary conditions cover the time slab between 3.29 and 2.97 Ma BP according to the geomagnetic polarity time scale (Berggren et al., 1995). Boundary conditions integrated into the model that are specific to the Pliocene include: (1) continental configuration, modified by a 25 m increase in global sea level, (2) modified present-day elevations, (3) reduced ice sheet size and height for Greenland (~50% reduction) and Antarctica (~33% reduction), (4) Pliocene vegetation distribution and (5) Pliocene SSTs and sea ice distributions. The geographical extents of the Greenland and Antarctic ice sheets within the PRISM2 dataset are based on global sea-level estimates derived for the Pliocene by Dowsett and Cronin (1990). For a more detailed description of the PRISM2 dataset and how it differs from earlier PRISM datasets see Dowsett et al. (1999: http://pubs.usgs.gov/ openfile/of99-535). The PRISM3D dataset, which is currently under development (see http://geology.er.usgs.gov/eespteam/prism/index. html), represents a significant enhancement over PRISM2. PRISM3D incorporates improvements to the SST and sea-ice datasets as well as increased coverage in the vegetation dataset. Estimates of Pliocene

Table 3 Table of palaeoclimate modelling experiments for the Pliocene used within this study. Experiment names, brief description of experiments, the run length and averaging period for each experiment along with prescribed trace gas concentrations are provided. Selected results are shown in Fig. 5. Results of all experiments are shown in Fig. 1 (Supplementary information). Name

Description

xajpa xaiuk xblmd

Present-day control run Pliocene control run using full PRISM2 boundary conditions As xaiuk but using trace gas concentrations conforming to Pliocene simulations using the GISS GCM run by Mark Chandler As xaiuk but using PRISM3D Minimum SST data set and trace gas concentrations conforming to Pliocene simulations using the GISS GCM run by Mark Chandler As xaiuk but using PRISM3D Maximum SST data set and trace gas concentrations conforming to Pliocene simulations using the GISS GCM run by Mark Chandler As xaiuk but using dynamic vegetation predicted by the TRIFFID vegetation model As xaiuk but except for using present-day rather than Pliocene orography. This increases the extent of the Greenland and Antarctic ice sheets to the maximum possible give the Pliocene land/sea mask equivalent to a +12 m sea-level rise compared to 25 m. As xaiuk but using smaller Antarctic ice sheet from the PRISM1 data set equivalent to a + 35 m change in sea level. As xaiuk but with reduced leads fraction As xaiuk but using a slab ocean model using oceanic heat transports calibrated from the PRISM2 SST reconstruction As xaiun but with higher CO2 and CH4 concentrations As xaiun but with modified horizontal ocean heat flux As xaiut but with modified orbital configuration equivalent to 2.97 Ma BP As xaiut but with modified orbital configuration equivalent to 3.13 Ma BP As xaiut but with modified orbital configuration equivalent to 3.29 Ma BP Fully coupled ocean–atmosphere GCM run for the Pliocene using PRISM2 boundary conditions (sea surface temperatures and sea-ice predicted not prescribed) and slightly modified trace gas concentrations from xaiuk As xatsb but with a higher CO2 concentration

xblmc xblma xatwd xaiuf

xaius xaiud xaiun xaiup xaiut xaiuu xaiuv xaiux xatsb

xbpsa

Averaging period (in years)

CO2 (in ppmv)

CH4 (in ppbv)

N2 O (in ppbv)

23 15 20

21 12 12

345 315 315

790 790 1250

284 284 287

20

12

315

1250

287

20

12

315

1250

287

30 13

10 12

400 315

790 790

284 284

18

12

315

790

284

12 17

11 12

315 315

790 790

284 284

25 26 26 26 26 350

20 20 20 20 20 60

400 315 315 315 315 400

1580 790 790 790 790 703

284 284 284 284 284 276

200

30

560

703

276

Total length of run (in years)

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

89

Fig. 5. Model-predicted snow accumulation (metres of water equivalent per decade) over Antarctica for experiments selected to illustrate a wide range of snow-mass configurations. Experiment xajpa is a present-day control run. All other results are from Pliocene sensitivity experiments. In xbpsa, the CO2 concentration was set at 560 ppmv, resulting in the highest mean global temperatures examined (4.5 °C above present). See Table 3 and supplementary information provided online for details of specific experiments.

surface ocean temperatures have already been enhanced by the addition of maximum and minimum SST warming datasets providing a range of variability in Pliocene SST (for further details see Dowsett, 2004; Dowsett et al., 2005). Summary details of all experiments presented in this paper can be found in Table 3 with full details provided in supplementary information available online. 4.3. Model predictions of Pliocene snow accumulation on Antarctica Due to the inherent difficulties in modelling fast flowing ice-streams, combined with the relatively coarse resolution of GCMs, we are unable to directly simulate the APIS with an ice-sheet model. Instead we use the GCM's predictions for snow accumulation as a proxy for the presence or absence of an APIS. Snow accumulation (in metres of water equivalent per decade) is a predicted variable within the model. The models' predictions are based on precipitation patterns and amounts combined with the surface temperature at the time that precipitation occurs in each model grid box. As a decadal mean, the model predicts if precipitation, in the form of snow, survives and is allowed to accumulate, thus theoretically leading to the development of an ice sheet. The experimental design of the simulations presented serve as a severe test of the models' ability to predict the presence of an APIS since all of the experiments, save one (xaius), were initiated without an APIS. Moreover, the ensemble used included the PRISM MAX SST experiment. It is designed to recognise the warmest condition possible for the mid Pliocene and orbital variability is also contained within it, although every orbital configuration cannot be tested. The models' results from each simulation are consistent. In every case snow is predicted to accumulate over the Antarctic Peninsula, including experiment xbpsa with its much higher atmospheric CO2 (560 ppmV; Fig. 5, and supplementary information, Fig. 1). This suggests that given enough time an APIS would develop regardless of the different initial and boundary conditions used within the GCM and, thus, that the APIS was a permanent and persistent feature of the Pliocene world. Wild et al. (2003) reached a similar conclusion using a

modern climatology configured with 2× modern CO2 concentrations. Variations in the amount of snow accumulation are predicted between different experiments. For example, snow accumulation is more widespread over the Peninsula and the West Antarctic region in general in experiments xaiuf, xaiuu, xaiuv, and is less extensive than normal in experiment xatsb. In experiment xaiuf, a larger EAIS was prescribed which had the effect of reducing Antarctic temperatures and hence increasing snow accumulation. In experiment xaiuu and xaiuv, insolation values were modified to conform to two specific time slices (2.97 and 3.12 Ma BP). In these experiments the incoming solar radiation (insolation) at the top of the atmosphere (TOA) is reduced during the Southern Hemisphere winter (see Haywood et al., 2002c). This means that Southern Hemisphere summer temperatures are cooler and hence snow accumulation is greater. This result indicates that the size of the APIS could vary in response to orbital variations over Milankovitch timescales. Experiment xbpsa (a fully coupled ocean– atmosphere GCM simulation) had the smallest snow accumulation of all the experiments. That run stimulated the largest global mean rise in Pliocene temperatures relative to present-day, i.e. ~4.5 °C, using a CO2 concentration of 560 ppmV. Overall the simulations indicate the potential existence of an APIS over a global mean temperature change spanning from + 1.9 to +4.5°C. 5. Discussion The role of the cryosphere in the Earth's climate system is important for understanding the mechanisms of climate change and for better constraining future climate predictions. It is important to validate ice sheet reconstructions for deep-time (i.e. pre-Quaternary) periods using direct geological evidence, in order to constrain any models and thereby increase our confidence in the modelling results. This is an objective that is addressed in this paper for the APIS. The Pliocene epoch, in particular, is crucial to our understanding of ice sheet responses to future climate change since it is the most recent period in geological time when global temperatures were generally

90

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

higher than modern (Haywood and Valdes, 2004). It is also a period for which, uniquely in Antarctica, we now have a chronologically wellconstrained geological dataset of ice sheet information for the Antarctic Peninsula, the importance of which is only just beginning to be assessed (Smellie et al., 2006a,b, 2008; Hambrey et al., 2008). The amplification of temperatures in the polar regions makes it imperative to assess the possible impacts on melting and general stability of ice sheets, and global sea levels. At present, ice sheet volume reconstructions used in modelling Pliocene climate are based on estimates of sea level and marine oxygen isotope data. Both have large error bars associated with them (e.g. Dowsett and Cronin, 1990; Krantz, 1991; Wardlaw and Quinn, 1991) and a wide range of significantly different ice sheet scenarios is possible. Because of the high degree of climate sensitivity of the Antarctic Peninsula, many workers have assumed that it was ice free during the warm Pliocene and, until now, this assumption has been the basis for all subsequent climate modelling (e.g. Abreu and Anderson, 1998; Dowsett et al., 1999). However, this assumption is conflicted by the results of marine studies (Section 3.2), in particular seismic and drilling campaigns that seem to show that an APIS was present and may have been intrinsically more dynamic than either the WAIS or EAIS for the period (e.g. Barker at al., 1999, 2002; Bart and Anderson, 2000; Hepp et al., 2006; Bart et al., 2007). Unfortunately, most offshore records are incapable of establishing the thickness and dominant glacial regime of the APIS since they are biased toward ice stream flow and deposition and may not be representative of the greater area of the ice sheet dominated by sheet-flow (so-called “inland ice”). Glacial– interglacial cyclicity has been inferred for the Pliocene APIS from physical, mineralogical and chemical records, X-ray images and IRD distribution and abundance (e.g. Cowan, 2002; Pudsey, 2002a,b; Hillenbrand and Ehrmann, 2005; Hepp et al., 2006; Cowan et al., 2008). The marine record contains gaps in the measured indices used to represent ice sheet input, and the claims for a continuously-present ice sheet throughout the Pliocene and younger are still unconfirmed. But, however it is viewed, the record unequivocally indicates that an APIS was present in the warm Pliocene, it experienced fluctuations in extent and it was prone to undergoing large-scale changes in volume. The terrestrial record of glaciations in the Antarctic Peninsula is typically regarded as of limited value compared with the more con-

tinuous and usually higher-resolution marine record (e.g. Hill et al., 2007). However, recent geological investigations onshore, particularly in the northern Antarctic Peninsula region, have created a well-dated, increasingly rich dataset of palaeoenvironmental data. The terrestrial record significantly enhances the marine record and both are consistent in supporting a more holistic picture for pre-Quaternary time comprising an areally dynamic Pliocene APIS that fluctuated in thickness between rather narrow limits. The terrestrial outcrop information suggests a generally thin low-profile APIS throughout the Pliocene and Quaternary with a thickness varying up to c. 850 m (Figs. 4 and 6). This scenario contrasts with thicker and higher-profile ice sheets predicted or inferred for the LGM by some modelling and marine studies (e.g. Denton et al., 1984; Kennedy and Anderson, 1989; Denton et al., 1991; Nakada et al., 2000; Huybrechts, 2002). However, other reconstructions for the LGM point to an APIS that was drained by ice streams advancing to near the shelf break and which might have therefore had a low profile (Anderson et al., 2002; Heroy and Anderson, 2005; Domack et al., 2006; Sugden et al., 2006, and references therein). Although maximum ice sheet thicknesses cannot be precisely defined from the existing terrestrial (glaciovolcanic and cosmogenic) data, they may not have significantly exceeded 750–850 m in the James Ross Island region and probably elsewhere, and there was probably never a “giant” APIS (sensu Denton et al., 1984, 1991). This suggestion is supported by poorly-dated geomorphological evidence from Palmer Land that suggests that the APIS was only up to 500 m higher than today at LGM (Bentley et al., 2006). Interpretation of the derivation of the coarse-grained terrigenous sediment fraction offshore also suggests that glaciers were thin and did not inundate the Peninsula topography until Pleistocene times, at least (Cowan et al., 2008). Thus, multiple lines of geological evidence are now converging in suggesting that the ice sheet would have draped, rather than deeply submerged, the topography of the Antarctic Peninsula. Local ice caps would have been prominent on its east and west flanks (e.g. James Ross and Brabant islands) and dominated the local morphodynamics of ice flow (Smellie et al., 2006a, 2008; Hambrey et al., 2008; Fig. 6). Similar maximum ice thicknesses are also suggested by glaciovolcanic outcrops in southern Antarctic Peninsula in Plio-Pleistocene time (Table 2). At times of thinner ice and possibly even during periods of thicker ice, the landscape would have resembled an icefield more than an ice sheet,

Fig. 6. Schematic cross section through northern Antarctic Peninsula in which the APIS is depicted at the maximum thickness inferred from this study (essentially equivalent to the LGM). The diagram also shows the substantially thicker APIS profile estimated from the modelling study of Denton et al. (1991). However, the new reconstructed ice sheet surface is essentially coincident with results of later modelling studies by Denton and Hughes (2002; not shown). The reconstruction shows how any part of the APIS might have looked during multiple “thick ice” periods (glacial maxima) from latest Miocene times.

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

91

Fig. 7. View of the Graham Land plateau from the Ineson Glacier on James Ross Island. The view shows the northern end of Detroit Plateau, with an elevation of about 1500 m a.s.l.

with the underlying bedrock topography, not the ice surface, determining the ice sheet flow dynamics (Fig. 6). There is also other independent evidence that suggests the APIS was probably always a relatively thin glacial cover. Northern Antarctic Peninsula contains a conspicuous yet enigmatic landscape feature, comprising a gently undulating, strikingly planar ice-covered plateau that can be traced for about 800 km in Graham Land (Fig. 7; see also Fig.1). It is an elevated bedrock erosion surface with a general elevation of 1750 m (varying between 900 m and 2000, rising progressively southwards). The surface formed either by marine erosion or prolonged subaerial denudation probably close to sea level, and was subsequently uplifted (Linton, 1964; Elliot, 1997). The age of the erosion surface is uncertain but uplift in the late Miocene and Pliocene is currently favoured (Elliot, 1997) and it was probably related to the progressive subduction of young oceanic crust under the Pacific margin of the Antarctic Peninsula from the Miocene to the Pliocene (e.g. Larter and Barker, 1991a,b; Larter et al., 1997). Significant uplift must have taken place by mid–late Miocene times, at least, since the presence of Antarctic Peninsula-derived erratics in the oldest glacial sediments on James Ross Island (Pirrie et al., 1997; Smellie et al., 2006b; Hambrey et al., 2008) suggests that the Peninsula was already an upstanding landmass supporting an erosive ice sheet by 6.2 Ma, at least (age of the oldest in situ volcanism, which erratic-bearing glacial sediments underlie). An existing high topography would have had a major influence on local precipitation and elevation-related temperature patterns, and thus on the seeding and growth of an ice mass as global temperatures fell during the Neogene (Barker, 1995; Elliot, 1997). The plateau is flanked by subvertical rock faces and cirques that feed fast-moving, typically tidewater glaciers. Submerging the topography in a very thick erosive ice sheet would rapidly reduce the Peninsula to a lower, more subdued landform by focused corner erosion (Fig. 8A). Conversely, a thin ice sheet would more likely erode principally by landward recession of the cirque headwalls, thus preserving the plateau and its steep flanks, features that are seen today (Fig. 8B; cf. Fig. 1; Elliot, 1997). The terrestrial record is unusually well dated and it suggests that the trend of APIS thicknesses increased between late Miocene time (b7 Ma) and present (Fig. 4). This might have occurred in response to increasing intensity of global cooling and its effect on the ice cover: i.e. colder ice, hence “harder”, with cooler basal conditions, higher basal friction and slower movement due to higher viscosity (Marshall, 2005), thus causing ice sheet thickening (cf. Hughes, 1981, Figs. 5–8; Barker, 1995). The terrestrial data show no clear change in thermal regime, although information becomes quite sparse for periods younger than c. 1 Ma. This conflicts with recent suggestions that a fundamental

continental-scale transition in Antarctic Ice Sheet regime (to a cold polar dry-based ice sheet) took place at c. 3 Ma, based on seismic reflection data from the continental margin (Rebesco et al., 2006; but see criticisms by Larter, 2007). Our results are closer to suggestions by Hassler and Cowan (2002) and Cowan et al. (2008) that the APIS changed from a sub-polar to a polar ice sheet between c. 1.4 and 0.76 Ma, when its thickness increased. However, an erosive subpolar(?) ice sheet is thought to have carved the margins of the Cape Purvis volcano (southern Dundee Island) at 132 ka and wet-based erosive ice also draped Brabant Island in late Pleistocene time, b200 ka (Smellie

Fig. 8. Schematic diagrams illustrating how erosional effects on a landscape caused by (A) a thick overriding ice sheet, or (B) thin draping ice sheet might differ greatly in the Antarctic Peninsula in Neogene time. In (A), gravity drives ice to flow downslope, and surface slope and ice thickness control the gravitational driving stress. Thus, as thickness increases so does ice flow. As a result, rapid corner erosion will rapidly reduce the Graham Land plateau to a more rounded subdued landscape (shown by thick dashed lines). In (B), by contrast, a thin ice sheet will cause rapid cirque headwall erosion, which will reduce the width of the plateau but largely preserve its morphology.

92

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

et al., 2006a,c), which, if true, suggests that any switch to a polar thermal regime in northern Antarctic Peninsula was not permanent but may have changed back at times. An absence of snow accumulation in the northernmost tip of the Antarctic Peninsula is observed in most of our experiments (Fig. 5). Such an APIS configuration is consistent with the model of Huybrechts (1993), even at small temperature increases. However, although the geological data appear to disagree, it may be an artefact of (1) the coarse modelling resolution used on a northerly-thinning narrow landmass, and (2) the mismatch between the Pliocene land–sea mask used and the true coastline configuration, which is particularly serious for northern Graham Land. The higher-resolution modelling study by Wild et al. (2003) avoided the latter problem by restricting their results for the Antarctic Peninsula to Palmer Land and southern Graham Land. Thus, whilst it cannot yet be demonstrated unambiguously that the APIS persisted in the interglacial periods, the geological datasets are most easily interpreted that way (e.g. Bart and Anderson, 2000; Cowan, 2002; Hillenbrand and Ehrmann, 2002; Pudsey, 2002a,b; Hillenbrand and Ehrmann, 2005; Bart et al., 2007; Smellie at al., 2008; Cowan et al., 2008; Johnson et al., submitted for publication) and it is consistent with our modelling results. Despite Pliocene warmth, the ability of the APIS to expand repeatedly across the continental shelf, combined with the likelihood that it never completely melted away, makes it questionable if significant terrestrial vegetation could have become established in the northernmost part of the Antarctic Peninsula during the period. Experiments with a vegetation model for East Antarctica do not exclude this possibility (Francis et al., 2007) nor does the biome-modelling study by Salzmann et al. (2008). However, the abundance and origin of pollen and spores in Pliocene sediments offshore of Antarctica give no strong support for a spreading of terrestrial flora (e.g. Fleming and Barron, 1996; Iwai et al., 2001). The palynology is particularly depauperate, in both numbers and species/genera and is equally consistent with an origin by reworking rather than an in situ flora. If the thin APIS did extend to the continental shelf edge, then it would also have had a lower profile than published modelling results suggested (cf. Denton et al., 1984, 1991; Huybrechts, 2002). With such a relatively thin glacial cover, the results also allow the possibility that low-altitude ice-free exposures persisted through multiple glaciations even at periods of glacial maxima. This result is likely to be highly important in the search for possible environmental refugia for animal and plant species isolated in Antarctica by continental separation and threatened with extinction during glacial periods (Convey et al., 2008). 6. Conclusions Assessing the stability of polar ice sheets under climates significantly warmer than present, and thereby their influence on global sea levels, are high priorities for palaeoenvironmental research. As a result of this review of all available geological (terrestrial and marine) datasets and our model predictions for snow accumulation on the Antarctic Peninsula, it is clear that the different datasets are convergent and consistently indicate that an APIS was present and probably persistent under the widest range of warm Pliocene climatic conditions tested so far. Our results appear to be robust for global mean temperature increases up to 4.5 °C. Acknowledgements The Natural Environment Research Council is acknowledged for supporting this work through the provision of High Performance Computing. This paper is published as part of the British Antarctic Survey's GEACEP programme (ISODYN Project) that seeks to investigate climate change over geological time scales, and it also contributes to the SCAR ACE programme (Antarctic Climate Evolution). The authors are also grateful to Ellen Cowan and Peter Barrett for their helpful reviews of this paper.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.earscirev.2009.03.005. References Abreu, V.S., Anderson, J.B., 1998. Glacial eustacy during the Cenozoic: sequence stratigraphic implications. American Association of Petroleum Geologist Bulletin 82, 1385–1400. Anderson, J.B., Shipp, S.S., Lowe, A.L., Wellner, J.S., Mosola, A.B., 2002. The Antarctic ice sheet during the last glacial maximum and its subsequent retreat history: a review. Quat. Sci. Rev. 22, 49–70. Barker, P.F., 1995. The proximal marine sediment record of Antarctic climate since the Late Miocene. In: Cooper, A.K., Barker, P.F., Brancolini, G. (Eds.), Geology and Seismic Stratigraphy of the Antarctic Margin. Antarct. Res. Ser., vol. 68. AGU, Washington, DC, pp. 25–57. Barker, P.F., Barrett, P.J., Camerlenghi, A., Cooper, A.K., Davey, F.J., Domack, E.W., Escutia, C., Kristoffersen, Y., O'Brien, P., 1998. Ice sheet history from Antarctic continental margin sediments: the ANTOSTRAT approach. Terra Antartica 5, 737–760. Barker, P.F., Camerlenghi, A., Acton, G.D., et al. 1999. Proc. ODP, Init. Repts. 178 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station, TX 77845-9547, U.S.A. Barker, P.F., Camerlenghi, A., Acton, G.D., et al. 2002. Proc. ODP, Sci. Results 178 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Bart, P.J., 2001. Did the Antarctic ice sheets expand during the early Pliocene? Geology 29, 67–70. Bart, P.J., Anderson, J.B., 2000. Relative temporal stability of the Antarctic Ice Sheets during the late Neogene based on the minimum frequency of outer shelf grounding events. Earth Planet. Sci. Lett. 182, 259–272. Bart, P.J., Hillenbrand, C.-D., Ehrmann, W., Iwai, M., Winter, D., Warny, S.A., 2007. Are Antarctic Peninsula Ice Sheet grounding events manifest in sedimentary cycles on the adjacent continental rise? Mar. Geol. 236, 1–13. Bentley, M.J., Anderson, J.B., 1998. Glacial and marine geological evidence for the ice sheet configuration in the Weddell Sea–Antarctic Peninsula region during the Last Glacial Maximum. Antarct. Sci. 10, 309–325. Bentley, M.J., Fogwill, C.J., Kubik, P.W., Sugden, D.E., 2006. Geomorphological evidence and cosmogenic 10Be/26Al exposure ages for the Last Glacial Maximum and deglaciation of the Antarctic Peninsula Ice Sheet. Geol. Soc. Amer. Bull. 118, 1149–1159. Berggren, W.A., Kent, D.V., Swisher, C.C., Aubry, M.P., 1995. A revised Cenozoic geochronology and chronostratigraphy. In: Berggren, W.A., Kent, M.P., Aubry, Hardenbol, J. (Eds.), Geochronology, Time Scales and Global Stratigraphic Correlation. Tulsa, Society for Sedimentary Geology Special Publication, vol. 54, pp. 129–212. Bracegirdle, T.J., Connolley, W.M., Turner, J., 2008. Antarctic climate change over the twenty first century. J. Geophys. Res. 113. doi:10.1029/2007JD008933. Brigham-Grette, J., 1994. Warm Pliocene high sea-level events in the Arctic Alaska. Transactions of the American Geophysical Union, Spring Meeting, vol. 75. EOS (Suppl.). Camerlenghi, A., Domack, E.W., Rebesco, M., Gilbert, R., Ishman, S., Leventer, A., Brachfeld, S., Drake, A., 2001. Glacial morphology and post-glacial contourites in northern Prince Gustav Channel (NW Weddell Sea, Antarctica). Mar. Geophys. Res. 22, 417–443. Chandler, M., Rind, D., Thompson, R., 1994. Joint investigations of the middle Pliocene climate II: GISS GCM Northern Hemisphere results. Glob. Planet. Change 9, 197–219. Chinn, T.J.H., Dillon, A., 1987. Observations on a debris-covered polar glacier, “Whisky Glacier, James Ross Island, Antarctic Peninsula. J. Glaciol. 33, 300–310. Convey, P., Gibson, J.A.E., Hillenbrand, C.-D., Hodgson, D.A., Pugh, P.J.A., Smellie, J.L., Stevens, M.J., 2008. Antarctic terrestrial life — challenging the history of the frozen continent? Biol. Rev. 83, 103–117. Cook, A.J., Fox, A.J., Vaughan, D.G., Ferrigno, J.G., 2005. Glacier fronts on the Antarctic Peninsula over the past half-century. Science 308, 541–544. Cooper, A.K., O'Brien, P.E., Richter, C. (Eds.), 2004. Proc. ODP, Sci. Results 188 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Cowan, E.A., 2002. Identification of the glacial signal from the Antarctic Peninsula since 3.0 Ma at Site 1011 in a continental rise sediment drift. In Barker, P.F., Camerlenghi, A., Acton, G.D., and Ramsay, A.T.S. (eds.), Proc. ODP, Sci. Results 178, 1–22 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Cowan, E.A., Hillenbrand, C.-D., Hassler, L.E., Ake, M.T., 2008. Coarse-grained terrigenous sediment deposition on continental rise drifts: a record of Plio-Pleistocene glaciation on the Antarctic Peninsula. Palaeogeogr. Palaeoclimatol. Palaeoecol. 265, 275–291. Denton, G.H., Prentice, M.L., Kellogg, D.E., Kellogg, T.B., 1984. Late Tertiary history of the Antarctic Ice Sheet: evidence from the Dry Valleys. Geology 12, 263–267. Denton, G.H., Prentice, M.L., Burckle, L.H., 1991. Cainozoic history of the Antarctic Ice Sheet. In: Tingey, R.J. (Ed.), The Geology of Antarctica. Oxford University Press, Oxford, pp. 365–433. Denton, G.H., Hughes, T.J., 2002. Reconstructing the Antarctic Ice Sheet at the Last Glacial Maximum. Quat. Sci. Rev. 21, 193–202. Dingle, R.V., McArthur, J.M., Vroon, P., 1997. Oligocene and Pliocene interglacial events in the Antarctic Peninsula dated using Strontium isotope stratigraphy. Journal of the Geological Society London 154, 257–264. Dixon, J.E., Filiberto, J.R., Moore, J.G., Hickson, C.J., 2002. Volatiles in basaltic glasses from a subglacial volcano in northern British Columbia (Canada): implications for ice sheet thickness and mantle volatiles. In: Smellie, J.L., Chapman, M.G. (Eds.), Volcano– Ice Interaction on Earth and Mars. Geol. Soc. Lond., Spec. Publ., vol. 202, pp. 255–271.

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94 Domack, E., Amblas, D., Gilbert, R., Brachfeld, S., Camerlenghi, A., Rebesco, M., Canals, M., Urgeles, R., 2006. Subglacial morphology and glacial evolution of the Palmer deep outlet system, Antarctic Peninsula. Geomorphology 75, 125–142. Dowsett, H.J., 2004. Bracketing mid Pliocene sea surface temperature: maximum and minimum possible warming. U.S. Geological Survey Data Series DS114. http://pubs. usgs.gov/ds/2004/114/. Dowsett, H.J., Cronin, T.M., 1990. High eustatic sea level during the Middle Pliocene: evidence from the southeastern U.S. Atlantic Coastal Plain. Geology 18, 435–438. Dowsett, H.J., Cronin, T.M., Poore, P.Z., Thompson, R.S., Whatley, R.C., Wood, A.M., 1992. Micropaleontological evidence for increased meridional heat-transport in the North Atlantic Ocean during the Pliocene. Science 258, 1133–1135. Dowsett, H.J., Barron, J.A., Poore, R.Z., Thompson, R.S., Cronin, T.M., Ishman, S.E., Willard, D.A., 1999. Middle Pliocene paleoenvironmental reconstruction: PRISM2. USGS Open File Report 99-535, http://pubs.usgs.gov/openfile/of99-535. Dowsett, H.J., Chandler, M.A., Cronin, T.M., Dwyer, G.S., 2005. Middle Pliocene sea surface temperature variability. Paleoceanography 20, PA2014. Drewry, D.J., 1983. Antarctica: Glaciological and Geophysical Folio. Scott Polar Research Institute, Cambridge. Elliot, D.H., 1997. The planar crest of Graham Land, northern Antarctic Peninsula: possible origins and timing of uplift. American Geophysical Union Antarctic Research Series 71, 51–73. Evans, J., Pudsey, C.J., ÓCofaigh, C., Morris, P., Domack, E., 2005. Late Quaternary glacial history, flow dynamics and sedimentation along the eastern margin of the Antarctic Peninsula Ice Sheet. Quat. Sci. Rev. 24, 741–774. Eyles, N., Daniels, J., Osterman, L.E., Januszczak, N., 2001. Ocean Drilling Program Leg 178 (Antarctic Peninsula): sedimentology of glacially influenced continental margin topsets and foresets. Mar. Geol. 178, 135–156. Fleming, R.F., Barron, J.A., 1996. Evidence of Pliocene Nothofagus in Antarctica from Pliocene marine sedimentary deposits (DSDP Site 274). Mar. Micropaleontol. 27, 227–236. Antarctic climate evolution. In: Florindo, F., Siegert, M. (Eds.), Developments in Earth and Environmental Sciences, vol. 8. Elsevier. Fox, A.J., Cooper, A.P.R., 1998. Climate-change indicators from archival aerial photography of the Antarctic Peninsula. Ann. Glaciol. 27, 636–642. Francis, J.E., Haywood, A.M., Ashworth, A., Valdes, P., 2007. Tundra environments in the Neogene Sirius Group, Antarctica: evidence from the geological record and coupled atmosphere–vegetation models. J. Geol. Soc. 164, 317–322. doi:10.1144/ 0016-76492005-191. Francis, J.E., Ashworth, A., Cantrill, D.J., Crame, J.A., Howe, J., Stephens, R., Tosolini, A.-M., Thorn, V., 2008. 100 million years of Antarctic climate evolution: evidence from fossil plants. In: Cooper, A.K., Barrett, P.J., Stagg, H., Storey, B.C., Stump, E., Wise, W., and the 10th ISAES editorial team, eds. Antarctica: A Keystone in a Changing World. Proceedings of the 10th International Symposium on Antarctic Earth Sciences. The National Academies Press, Washington, D.C., 19–27. Gersonde, R., Hodell, D.A., Blum, P., et al., 1999. Proc. ODP, Init. Repts. 177 (CD-ROM). Ocean Drilling Program, Texas A&M University, College Station, TX 77845-9547, U.S.A. Goodfellow, B.W., 2007. Relict non-glacial surfaces in formerly glaciated landscapes. Earth-Scie. Rev. 80, 47–73. Gordon, C., Cooper, C., Senior, C.A., Banks, H., Gregory, J.M., Johns, T.C., Mitchell, J.F.B., Wood, R.A., 2000. The simulation of SST, sea ice extents and ocean heat transports in a version of the Hadley Centre coupled model without flux adjustments. Clim. Dyn. 16, 147–168. Gregory, J.M., Mitchell, J.F.B., 1997. The climate response to CO2 of the Hadley Centre coupled AOGCM with and without flux adjustment. Geophys. Res. Lett. 24,1943–1946. Hambrey, M.J., Smellie, J.L., 2006. Distribution, lithofacies and environmental context of Neogene glacial sequences on James Ross and Vega islands, Antarctic Peninsula. In: Francis, J.E., Pirrie, D., Crame, J.A. (Eds.), Cretaceous–Tertiary High-latitude Palaeoenvironments, James Ross Basin, Antarctica. Geological Society, London, Special Publication, vol. 258, pp. 187–200. Hambrey, M.J., Smellie, J.L., Nelson, A.E., Johnson, J.S., 2008. Late Cenozoic glacier–volcano interaction on James Ross Island and adjacent areas, James Ross Island. Geol. Soc. Amer. Bull. 120, 709–731. Hansen, J., Russell, G., Rind, D., Stone, P., Lacis, A., Lebedeff, S., Ruedy, R., Travis, L., 1983. Efficient three-dimensional global models for climate studies: models I and II. Mon. Weather Rev. 111, 609–662. Haq, B.H., Hardenbol, J., Vail, P.R., 1987a. Chronology of fluctuating sea levels since the Triassic. Science 235, 1156–1167. Haq, B.H., Hardenbol, J., Vail, P.R., 1987b. The new chronostratigraphic basis of Cenozoic and Mesozoic sea level cycles. Cushman Found Foraminiferal Research Special Publication 24, 7–13. Hassler, L.E., Cowan, E.A., 2002. Characteristics of ice-rafted pebbles from the continental rise sediment drifts west of the Antarctic Peninsula (Sites 1095, 1096, and 1101). In Barker, P.F., Camerlenghi, A., Acton, G.D., and Ramsay, A.T.S. (eds.), Proc. ODP, Sci. Results, 178, 1–23 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Hathway, B., 2001. Sims Island: first data from a Pliocene alkaline volcanic centre in eastern Ellsworth Land. Antarct. Sci. 13, 87–88. Haywood, A.M., Valdes, P.J., 2004. Modelling Middle Pliocene warmth: contribution of atmosphere, oceans and cryosphere. Earth Planet. Sci. Lett. 218, 363–377. Haywood, A.M., Valdes, P.J., 2006. Vegetation cover in a warmer world simulated using a dynamic global vegetation model for the Mid-Pliocene. Palaeogeogr. Palaeoclimatol. Palaeoecol. 237, 412–427. Haywood, A.M., Valdes, P.J., Sellwood, B.W., 2000a. Global scale palaeoclimate reconstruction of the middle Pliocene climate using the UKMO GCM: initial results. Glob. Planet. Change 25, 239–256. Haywood, A.M., Sellwood, B.W., Valdes, P.J., 2000b. Regional warming: Pliocene (3 Ma) paleoclimate of Europe and the Mediterranean. Geology 28, 1063–1066.

93

Haywood, A.M., Valdes, P.J., Sellwood, B.W., Kaplan, J.O., Dowsett, H.J., 2001. Modelling Middle Pliocene warm climates of the USA. Palaeontologia Electronica 4 (1) art. 5: 21pp., 933 KB. http://palaeoelectronica.org/2001_1/climate/issue1_01.htm. Haywood, A.M., Valdes, P.J., Sellwood, B.W., Kaplan, J.O., 2002a. Antarctic climate during the middle Pliocene: model sensitivity to ice sheet variation. Palaeogeogr. Palaeoclimatol. Palaeoecol. 182, 93–115. Haywood, A.M., Valdes, P.J., Sellwood, B.W., 2002b. Magnitude of middle Pliocene climate variability: a palaeoclimate modelling study. Palaeogeogr. Palaeoclimatol. Palaeoecol. 188, 1–24. Haywood, A.M., Valdes, P.J., Francis, J.E., Sellwood, B.W., 2002c. Global middle Pliocene biome reconstruction: a data/model synthesis. Geochemistry, Geophysics, Geosystems 3 (1), 1072. doi:10.1029/2002GC000358. Haywood, A.M., Dekens, P., Ravelo, A.C., Williams, M., 2005. Warmer tropics during the mid Pliocene? Evidence from alkenone paleothermometry and a fully coupled ocean–atmosphere GCM. Geochemistry, Geophysics, Geosystems 6. doi:10.1029/ 2004GC000799. Hepp, D.A., Mörz, T., Grützner, J., 2006. Pliocene glacial cyclicity in a deep-sea sediment drift (Antarctic Peninsula margin). Palaeogeogr. Palaeoclimatol. Palaeoecol. 231, 181–198. Heroy, D.C., Anderson, J.B., 2005. Ice-sheet extent of the Antarctic Peninsula region during the Last Glacial Maximum (LGM) — insights from glacial geomorphology. Am. Geol. Soc. Bull. 117, 1497–1512. Hill, D.J., Haywood, A.M., Hindmarsh, R.C.A., Valdes, P.J., 2007. Characterizing ice sheets during the Pliocene: evidence from data and models. In: Williams, M., Haywood, A.M., Gregory, F.J., Schmidt, D.N. (Eds.), Deep-time Perspectives on Climate Change: Marrying the Signal from Computer Models and Biological Proxies. The Micropalaeontological Society Special Publications, pp. 517–538. Hillenbrand, C.-D., Ehrmann, W., 2002. Distribution of clay minerals in drift sediments on the continental rise west of the Antarctic Peninsula, ODP Leg 178, Sites 1095 and 1096. In Barker, P.F., Camerlenghi, A., Acton, G.D., Ramsay, A.T.S. (eds.), Proc. ODP, Sci. Results 178, 1–29 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Hillenbrand, C.-D., Fütterer, D.K., 2002. Neogene to Quaternary deposition of opal on the continental rise west of the Antarctic Peninsula, ODP Leg 178, Sites 1095, 1096, and 1101. In Barker, P.F., Camerlenghi, A., Acton, G.D., and Ramsay, A.T.S. (eds.), Proc. ODP, Sci. Results 178, 1–33 [CDROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Hillenbrand, C.-D., Ehrmann, W., 2005. Late Neogene to Quaternary environmental changes in the Antarctic Peninsula region: evidence from drift sediments. Glob. Planet. Change 45, 165–191. Hughes, T.J., 1981. Numerical reconstruction of paleo-ice sheets. In: Denton, G.H., Hughes, T.J. (Eds.), The Last Great Ice Sheets. John Wiley and Sons, New York, pp. 221–261. Huybrechts, P., 1993. Glaciological modelling of the Late Cenozoic East Antarctic Ice Sheet: stability or dynamism? Geogr. Ann. 75A, 221–238. Huybrechts, P., 2002. Sea-level changes at the LGM from ice-dynamic reconstructions of the Greenland and Antarctic ice sheets during the glacial cycles. Quat. Sci. Rev. 21, 203–231. Ingólfsson, Ó., 2004. The Quaternary glacial and climate history of Antarctica. In: Ehlers, J., Gibbard, P.L. (Eds.), Quaternary Glaciations of the World, Part III. Kluwer, Dordrecht, pp. 3–43. IPCC, 2007. Climate change 2007: the physical science basis. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, New York, NY, USA. 996 pp. Iwai, M., Kameo, K., Miyake, N., 2001. Calcareous nannofossils, pollen, and spores from Leg 178 Sites 1095, 1097, 1100, and 1103, western Antarctic Peninsula: age constraints and environmental implications. In Barker, P.F., Camerlenghi, A., Acton, G.D., Ramsay, A.T.S. (eds.), Proc. ODP, Sci. Results 178, 1–22 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Johnson, J.S., Smellie, J.L., Nelson, A.E., Stuart, F.M., submitted for publication. Stability of the Antarctic Peninsula Ice Sheet since the early Pliocene — evidence from cosmogenic dating of Pliocene lavas on James Ross Island, Antarctica. Glob. Planet. Change. Jonkers, H.A., Kelley, S.P., 1998. A reassessment of the age of the Cockburn Island Formation, northern Antarctic Peninsula, and its palaeoclimatic implications. Journal of the Geological Society London 155, 737–740. Jonkers, H.A., Lirio, J.M., del Valle, R.A., Kelley, S.P., 2002. Age and environment of Miocene– Pliocene glaciomarine deposits, James Ross Island, Antarctica. Geol. Mag. 139, 577–594. Kennedy, D.S., Anderson, J.B., 1989. Glacial-marine sedimentation and Quaternary glacial history of Marguerite Bay, Antarctic Peninsula. Quat. Res. 31, 255–276. Krantz, D.E., 1991. A chronology of Pliocene sea-level fluctuations: the U.S. middle Atlantic coastal plain record. Quat. Scie. Rev. 10, 163–174. Larter, R.D., 2007. Margin architecture reveals the transition to the modern Antarctic ice sheet ca. 3 Ma: comment. Geology e139. doi:10.1130/G23422C.1 online forum. Larter, R.D., Barker, P.F., 1991a. Effects of ridge crest–trench interaction on Antarctic– Phoenix spreading: forces on a young subducting plate. J. Geophys. Res. 96, 19583–19607. Larter, R.D., Barker, P.F., 1991b. Neogene interaction of tectonic and glacial processes at the Pacific margin of the Antarctic Peninsula. In: Macdonald, D.I.M. (Ed.), Sedimentation, Tectonics and Eustasy, International Association of Sedimentologists, vol. 12. Special Publication, pp. 165–186. Larter, R.D., Rebesco, M., Vanneste, L.E., Gamboa, L.A.P., Barker, P.F., 1997. Cenozoic tectonic, sedimentary and glacial history of the continental shelf west of Graham Land, Antarctic Peninsula. In: Barker, P.F., Cooper, A.K. (Eds.), Geology and Seismic Stratigraphy of the Antarctic Margin, Part 2. Antarctic Research Series, vol. 71. American Geophysical Union, Washington, DC, pp. 1–27. LeMasurier, W.E., Thomson, J.W. (Eds.),1990. Volcanoes of the Antarctic Plate and Southern Oceans. Antarctic Research Series, vol. 48. Am. Geophys. Union. 487 pp.

94

J.L. Smellie et al. / Earth-Science Reviews 94 (2009) 79–94

Lisiecki, L.E., Raymo, M.E., 2005. A Pliocene–Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography 20. doi:10.1029/2004PA001071. Linton, D.L., 1964. Landscape evolution. In: Priestley, R., Adie, R.J., Robin, G., de, Q. (Eds.), Antarctic Research. Butterworths, London, pp. 85–100. Lucchi, R.G., Rebesco, M., Camerlenghi, A., Busetti, M., Tomadin, L., Villa, G., Persico, D., Morigi, C., Bonci, M.C., Giorgetti, G., 2002. Mid-late Pleistocene glacimarine sedimentary processes of a high-latitude, deep-sea sediment drift (Antarctic Peninsula Pacific margin). Mar. Geol. 189, 343–s370. Lundqvist, J., Lilliesköld, M., Östmark, K., 1995. Glacial and periglacial deposits of the Tumbledown Cliffs area, James Ross Island, West Antarctica. Geomorphology 11, 205–214. Lunt, D.J., Flecker, R., Valdes, P.J., Salzmann, U., Gladstone, R., Haywood, A.M., 2008. A methodology for targeting palaeo proxy data acquisition: a case study for the Late Miocene. Earth Planet. Sci. Lett. 271, 53–62. Lythe, M.B., Vaughan, D.G., 2001. BEDMAP: a new ice thickness and subglacial topographic model of Antarctica. J. Geophys. Res. 106, 11335–11351. Marshall, S.J., 2005. Recent advances in understanding ice sheet dynamics. Earth Planet. Sci. Lett. 240, 191–204. Martin, P.J., Peel, D.A., 1978. The spatial distribution of 10 m temperatures in the Antarctic Peninsula. J. Glaciol. 20, 311–317. McArthur, J.M., Rio, D., Massari, F., Castradori, D., Bailey, T.R., Thirlwall, M., Houghton, S., Dingle, R.V., 2007. A revised Pliocene record for marine-87Sr/86Sr used to date an interglacial event, Cockburn Island Formation, northern Antarctic Peninsula. Palaeogeogr. Palaeoclimatol. Palaeoecol. 242, 126–136. McGarvie, D.W., Stevenson, J.A., Burgess, R., Tuffen, H., Tindle, A.G., 2007. Volcano–ice interactions at Prestahnúkur, Iceland: rhyolite eruption during the last interglacial– glacial transition. Ann. Glaciol. 45, 38–47. Mercer, J.H., 1978. West Antarctic Ice Sheet and CO2 greenhouse effect — threat of disaster. Nature 271 (5643), 321–325. Morris, E.M., Mulvaney, R., 1996. Recent changes in surface elevation of the Antarctic Peninsula ice sheet. Z. Gletsch.kd. Glazialgeol. 31, 7–15. Morris, E.M., Vaughan, D.G., 2003. Spatial and temporal variation of surface temperature on the Antarctic Peninsula and the limit of variability of ice shelves. American Geophysical Union Antarctic Research Series 79, 61–68. Nakada, M., Kimura, R., Okuno, J., Moriwaki, K., Miura, H., Maemoku, H., 2000. Late Pleistocene and Holocene melting history of the Antarctic ice sheet derived from sea-level variations. Mar. Geol. 167, 85–103. Nelson, A.E., Smellie, J.L., Williams, M., Zalasiewicz, J., 2008. Late Miocene marine trace fossils from James Ross Island. Antarct. Sci. 20, 591–592. Pirrie, D., Crame, J.A., Riding, J.B., Butcher, A.R., Taylor, P.D., 1997. Miocene glaciomarine sedimentation in the northern Antarctic Peninsula region: the stratigraphy and sedimentology of the Hobbs Glacier Formation, James Ross Island. Geol. Mag. 136, 745–762. Prentice, M.L., Matthews, R.K., 1991. Tertiary ice sheet dynamics: the snow gun hypothesis. J. Geophys. Res. 19, 6811–6827. Pudsey, C.J., 2002a. Data report: Grain-size data, Sites 1095, 1096, and 1101, Antarctic Peninsula continental rise. In: Barker, P.F., Camerlenghi, A., Acton, G.D., and Ramsay, A.T.S. (eds.), Proc. ODP, Sci. Results 178, 1–34 [CDROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Pudsey, C.J., 2002b. Neogene record of Antarctic Peninsula glaciation in continental rise sediments: ODP Leg 178, Site 1095. In: Barker, P.F., Camerlenghi, A., Acton, G.D., and Ramsay, A.T.S. (eds.), Proc. ODP, Sci. Results 178, 1–25 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Pudsey, C.J., Murray, J.W., Appleby, P., Evans, J., 2006. Ice shelf history from petrographic, and foraminiferal evidence, Northeast Antarctic Peninsula. Quat. Sci. Rev. 25, 2357–2379. Rebesco, M., Camerlenghi, A., Geletti, R., Canals, M., 2006. Margin architecture reveals the transition to the modern Antarctic ice sheet ca. 3 Ma. Geology 34, 301–304. Rebesco, M., Camerlenghi, A., Geletti, R., Canals, M., 2007. Reply: margin architecture reveals the transition to the modern Antarctic ice sheet ca. 3 Ma. Geology. doi:10.1130/ G23894Y.1 online forum. Rignot, E., Casassa, G., Gogineni, P., Krabill, W., Rivera, A., Thomas, R., 2004. Accelerated ice discharge from the Antarctic Peninsula following the collapse of Larsen B ice shelf. Geophys. Res. Lett. 31, L18401. doi:10.1029/2004GL020697. Salzmann, U., Haywood, A.M., Lunt, D.J., Valdes, P.J., Hill, D.J., 2008. A new global biome reconstruction and data-model comparison for the Middle Pliocene, Global Ecology and Biogeography. Scambos, T.A., Hulbe, C., Fahenstock, M., Bohlander, J., 2001. The link between climate warming and break-up of ice shelves in the Antarctic Peninsula. J. Glaciol. 154, 516–530. Scambos, T., Hulbe, C., Fahnestock, M., 2003. Climate-induced ice shelf disintegration in the Antarctic Peninsula. American Geophysical Union Antarctic Research Series 79, 79–92. Scambos, T.A., Bohlander, J.A., Shuman, C.A., Skvarca, P., 2004. Glacier acceleration and thinning after ice shelf collapse in the Larsen B embayment, Antarctica. Geophys. Res. Lett. 31, L18402. doi:10.1029/2004GL020670. Schopka, H.H., Gudmundsson, M.T., Tuffen, H., 2006. The formation of Helgafell, southwest Iceland, a monogenetic subglacial hyaloclastite ridge: sedimentology, hydrology and volcano–ice interaction. J. Volcanol. Geotherm. Res. 152, 359–377. Sloan, L.C., Crowley, T.J., Pollard, D., 1996. Modeling of middle Pliocene climate with the NCAR GENESIS general circulation model. Mar. Micropaleontol. 27, 51–61. Smellie, J.L., 1999. Lithostratigraphy of Miocene-Recent, alkaline volcanic fields in the Antarctic Peninsula and eastern Ellsworth Land. Antarct. Sci. 11, 362–378. Smellie, J.L., 2000. Subglacial eruptions. In: Sigurdsson, H. (Ed.), Encyclopaedia of Volcanoes. Academic Press, San Diego, pp. 403–418. Smellie, J.L., 2001. Lithofacies architecture and construction of volcanoes in englacial lakes: Icefall Nunatak, Mount Murphy, eastern Marie Byrd Land, Antarctica. In:

White, J.D.L., Riggs, N. (Eds.), Lacustrine Volcaniclastic Sedimentation. International Association of Sedimentologists Special Publication, vol. 30, pp. 73–98. Smellie, J.L., 2006. The relative importance of supraglacial versus subglacial meltwater escape in basaltic subglacial tuya eruptions: an important unresolved conundrum. Earth-Sci. Rev. 74, 241–268. Smellie, J.L., 2007. Quaternary vulcanism: subglacial landforms. In: Elias, S.A. (Ed.), Encyclopedia of Quaternary Sciences. Elsevier, Amsterdam, pp. 784–798. Smellie, J.L., 2008. Basaltic subglacial sheet-like sequences: evidence for two types with different implications for the inferred thickness of associated ice. Earth-Sci. Rev. 88 (1–2), 60–88. Smellie, J.L., Skilling, I.P., 1994. Products of subglacial eruptions under different ice thicknesses: two examples from Antarctica. Sediment. Geol. 91, 115–129. Smellie, J.L., Hole, M.J., 1997. Products and processes in Pliocene–Recent, subaqueous to emergent volcanism in the Antarctic Peninsula: examples of englacial Surtseyan volcano construction. Bull. Volcanol. 58, 628–646. Smellie, J.L., Hole, M.J., Nell, P.A.R.,1993. Late Miocene valley-confined subglacial volcanism in northern Alexander Island, Antarctic Peninsula. Bull. Volcanol. 55, 273–288. Smellie, J.L., McIntosh, W.C., Esser, R., 2006a. Eruptive environment of volcanism on Brabant Island: evidence for thin wet-based ice in northern Antarctic Peninsula during the late Quaternary. Palaeogeogr. Palaeoclimatol. Palaeoecol. 231, 233–252. Smellie, J.L., McArthur, J.M., McIntosh, W.C., Esser, R., 2006b. Late Neogene interglacial events in the James Ross Island region, northern Antarctic Peninsula, dated by Ar/ Ar and Sr-isotope stratigraphy. Palaeogeogr. Palaeoclimatol. Palaeoecol. 242, 169–187. Smellie, J.L., McIntosh, W.C., Esser, R., Fretwell, P., 2006c. The Cape Purvis volcano, Dundee Island (northern Antarctic Peninsula): late Pleistocene age, eruptive processes and implications for a glacial palaeoenvironment. ntarct. Sci. 18, 399–408. Smellie, J.L., Johnson, J.S., McIntosh, W.C., Esser, R., Gudmundsson, M.G., Hambrey, M.J., van Wyk De Vries, B., 2008. Six million years of glacial history recorded in the James Ross Island Volcanic Group, Antarctic Peninsula. Palaeogeogr. Palaeoclimatol. Palaeoecol. 260, 122–148. Smith, A.M., Vaughan, D.G., Doake, C.S.M., Johnson, A.C., 1998. Surface lowering of the ice ramp at Rothera Point, Antarctic Peninsula, in response to regional climate change. Ann. Glaciol. 27, 113–118. Splettstoesser, J.F., 1992. Antarctic global warming? Nature 355, 503. Sugden, D.E., Bentley, M.J., Ó Cofaigh, C., 2006. Geological and geomorphological insights into Antarctic ice sheet evolution. Philosophical Transactions of the Royal Society A 364 (1844), 1607–1625. Sykes, M.A., 1988. New K–Ar age determinations on the James Ross Island volcanic Group, north-east Graham Land, Antarctica. Br. Antarct. Surv. Bull. 80, 51–56. Thomas, E.R., Marshall, G.J., McConnell, J.R., 2008. A doubling in snow accumulation in the western Antarctic Peninsula since 1850. Geophys. Res. Lett. 35, L01706. doi:10.1029/ 2007GL032529. Torinesi, O., Fily, M., Genthon, C., 2003. Variability and trends of the summer melt period of Antarctic ice margins since 1980 from microwave sensors. J. Climate 16 (7), 1047–1060. Tuffen, H., McGarvie, D.W., Gilbert, J.S., Pinkerton, H., 2002. Physical volcanology of a subglacial-to-emergent rhyolite tuya at Raudafossafjoll, Torfajokull, Iceland. In: Smellie, J.L., Chapman, M.G. (Eds.), Volcano–ice Interaction on Earth and Mars. Geol. Soc. Lond., Spec. Publ., vol. 202, pp. 213–236. Turner, J., King, J.C., Lachlan-Cope, T.A., Jones, P.D., 2002a. Recent temperature trends in the Antarctic. Nature 418, 291–292. Turner, J., Lachlan-Cope, T.A., Marshall, G.J., Morris, E.M., Mulvaney, R., 2002b. Spatial variability of Antarctic Peninsula net surface mass balance. J. Geophys. Res. 107. doi:10.1029/2001JD000755. Van Lipzig, N.P.M., King, J.C., Lachlan-Cope, T.A., van den Broeke, M.R., 2004. Precipitation, sublimation and snow drift in the Antarctic Peninsula region from a regional atmospheric model. J. Geophys. Res. 109. doi:10.1029/2004JD004701. Vaughan, D.G., 2006. Recent trends in melting conditions on the Antarctic Peninsula and their implications for ice-sheet mass balance and sea level. Arctic Antarctic and Alpine Research 38, 147–152. Vaughan, D.G., Doake, C.S.M., 1996. Recent atmospheric warming and retreat of ice shelves on the Antarctic Peninsula. Nature 379 (6563), 328–331. Vaughan, D.G., Bamber, J.L., Giovinetto, M., Russell, J., Cooper, A.P.R., 1999. Reassessment of net surface mass balance in Antarctica. J. Climate 12, 933–946. Vaughan, D.G., Marshall, G.J., Connolley, W.M., King, J.C., Mulvaney, R., 2001. Climate change: devil in the detail. Science 293, 1777–1779. Vaughan, D.G., Marshall, G.J., Connolley, W.M., Parkinson, C., Mulvaney, R., Hodgson, D.A., King, J.C., Pudsey, C.J., Turner, J., 2003. Recent rapid regional climate warming on the Antarctic Peninsula. Clim. Change 60, 243–274. Wardlaw, B.R., Quinn, T.M., 1991. The record of sea-level change at Enewetak Atoll. Quat. Sci. Rev. 10, 247–258. Wild, M., Calanca, P., Scherer, S.C., Ohmura, A., 2003. Effects of polar ice sheets on global sea level in high-resolution greenhouse scenarios. J. Geophys. Res.108, 4165. doi:10.1029/ 2002JD002451. Williams, M., Smellie, J.L., Johnson, J.S., Blake, D.B., 2006. Late Miocene Asterozoans (Echinodermata) from the James Ross Island Volcanic Group. Antarct. Sci.18,117–122. Wingham, D.J., Shepherd, A., Muir, A., Marshall, G.J., 2006. Mass balance of the Antarctic ice sheet. Philosophical Transactions of the Royal Society A 364, 1627–1635. Zubakov, V.A., Borzenkova, I.I., 1990. Global Palaeoclimate of the late Cenozoic. Dev. Palaeontology & Stratigraphy 12. Elsevier, Amsterdam. 456 pp.