(PDF) The Last Glacial Maximum at 44°S documented by a 10Be moraine chronology at Lake Ohau, Southern Alps of New Zealand

ABSTRACT The termination of the last ice age featured a major reconfiguration of Earth&amp;#39;s climate and cryosphere, yet the underlying causes of these massive changes continue to be debated. Documenting the spatial and temporal variations of atmospheric temperature during deglaciation can help discriminate among potential drivers. Here, we present a 10Be surface-exposure chronology and glaciological reconstruction of ice recession following the Last Glacial Maximum (LGM) in the Rakaia valley, Southern Alps of New Zealand. Innermost LGM moraines at Big Ben have an age of 17,840 ± 240 yrs, whereas ice-marginal moraines or ice-molded bedrock surfaces at distances up-valley from Big Ben of 12.5 km (Lake Coleridge), ˜25 km (Castle Hill), ˜28 km (Double Hill), ˜43 km (Prospect Hill), and ˜58 km (Reischek knob) have ages of 17,020 ± 70 yrs, 17,100 ± 110 yrs, 16,960 ± 370 yrs, 16,250 ± 340 yrs, and 15,660 ± 160 yrs, respectively. These results indicate extensive recession of the Rakaia glacier, which we attribute primarily to the effects of climatic warming. In conjunction with geomorphological maps and a glaciological reconstruction for the Rakaia valley, we use our chronology to infer timing and magnitude of past atmospheric temperature changes. Compared to an overall temperature rise of ˜4.65 °C between the end of the LGM and the start of the Holocene, the glacier recession between ˜17,840 and ˜15,660 yrs ago is attributable to a net temperature increase of ˜4.0 °C (from ‑6.25 to ‑2.25 °C), accounting for ˜86% of the overall warming. Approximately 3.75 °C (˜70%) of the warming occurred between ˜17,840 and ˜16,250 yrs ago, with a further 0.75 °C (˜16%) increase between ˜16,250 and ˜15,660 yrs ago. A sustained southward shift of the Subtropical Front (STF) south of Australia between ˜17,800 and ˜16,000 yrs ago coincides with the warming over the Rakaia valley, and suggests a close link between Southern Ocean frontal boundary positions and southern mid-latitude climate. Most of the deglacial warming in the Southern Alps occurred during the early part of Heinrich Stadial 1 (HS1) of the North Atlantic region. Because the STF is associated with the position of the westerly wind belt, our findings support the concept that a southward shift of Earth&amp;#39;s wind belts accompanied the early part of HS1 cooling in the North Atlantic, leading to warming and deglaciation in southern middle latitudes.

Alice M. Doughty, Joerg M. Schaefer, Aaron E. Putnam, George H. Denton, Michael R. Kaplan, David J.A. Barrell, Bjørn G. Andersen, Samuel E. Kelley, Robert C. Finkel, and Roseanne Schwartz Here we address a long-standing puzzle of ice-age climate called the “fly in the ointment of the Milankovitch theory.” Using geomorphic mapping and 10Be surface-exposure dating, we show that five moraine belts were formed during maxima of the last ice age by the Pukaki glacier in New Zealand’s Southern Alps. They afford ages of 41.76 ± 1.09 ka, 35.50 ± 1.26 ka, 27.17 ± 0.68 ka, 20.27 ± 0.60 ka, and 18.29 ± 0.49 ka. These five maxima spanned an entire precessional cycle in summer insolation intensity at the latitude of the Southern Alps. A similar mismatch between summer insolation and glacier extent also characterized the Chilean Lake District in the mid-latitudes of South America. Thus, in apparent contrast to northern ice sheets linked by Milankovitch to summer insolation at 65°N latitude, the behavior of southern mid-latitude glaciers was not tied to local summer insolation intensity. Instead, glacier extent between 41.76 ka and 18.29 ka, as well as during the last termination, was aligned with Southern Ocean surface temperature and with atmospheric carbon dioxide.

Recent debate on records of southern midlatitude glaciation has focused on reconstructing glacier dynamics during the last glacial termination, with different results supporting both in-phase and out-of-phase correlations with Northern Hemisphere glacial signals. A continuing major weakness in this debate is the lack of robust data, particularly from the early and maximum phase of southern midlatitude glaciation (∼30-20 ka), to verify the competing models. Here we present a suite of 58 cosmogenic exposure ages from 17 last-glacial ice limits in the Rangitata Valley of New Zealand, capturing an extensive record of glacial oscillations between 28-16 ka. The sequence shows that the local last glacial maximum in this region occurred shortly before 28 ka, followed by several successively less extensive ice readvances between 26-19 ka. The onset of Termination 1 and the ensuing glacial retreat is preserved in exceptional detail through numerous recessional moraines, indicating that ice re...

(This is a sample cover image for this issue. The actual cover is not yet available at this time.) This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy Quaternary Science Reviews 62 (2013) 114e141 Contents lists available at SciVerse ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev The Last Glacial Maximum at 44! S documented by a Lake Ohau, Southern Alps of New Zealand 10 Be moraine chronology at Aaron E. Putnam a, b, *, Joerg M. Schaefer a, c, George H. Denton b, David J.A. Barrell d, Sean D. Birkel b, Bjørn G. Andersen e,1, Michael R. Kaplan a, Robert C. Finkel f, Roseanne Schwartz a, Alice M. Doughty g a Lamont-Doherty Earth Observatory, 61 Rt. 9W, Palisades, NY 10944, USA Department of Earth Sciences, Climate Change Institute, University of Maine, Orono, ME 04469, USA Department of Earth and Environmental Sciences, Columbia University, New York, NY 10027, USA d GNS Science, Private Bag 1930, Dunedin 9054, New Zealand e Department of Geosciences, University of Oslo, 0316 Oslo, Norway f Department of Earth and Planetary Sciences, University of California, Berkeley, CA 95064, USA g Antarctic Research Centre, School of Earth Sciences, Victoria University of Wellington, PO Box 600, Wellington, New Zealand b c a r t i c l e i n f o a b s t r a c t Article history: Received 17 March 2012 Received in revised form 24 October 2012 Accepted 29 October 2012 Available online Determining whether glaciers registered the classic Last Glacial Maximum (LGM; w26,500ew19,000 yrs ago) coevally between the hemispheres can help to discriminate among hypothesized drivers of ice-age climate. Here, we present a record of glacier behavior from the Southern Alps of New Zealand during the ‘local LGM’ (LLGM). We used 10Be surface-exposure dating methods and detailed glacial geomorphologic mapping to produce a robust chronology of well-preserved terminal moraines deposited during the LLGM near Lake Ohau on central South Island. We then used a glaciological model to estimate a LLGM glacier snowline and atmospheric temperature from the Ohau glacier record. Seventy-three 10Be surface-exposure ages place culminations of terminal moraine construction, and hence completions of glacier advances to positions outboard of present-day Lake Ohau, at 138,600 " 10,600 yrs, 32,520 " 970 yrs ago, 27,400 " 1300 yrs ago, 22,510 " 660 yrs ago, and 18,220 " 500 yrs ago. Recessional moraines document glacier recession into the Lake Ohau trough by 17,690 " 350 yrs ago. Exposure of an ice-molded bedrock bench located inboard of the innermost LLGM moraines by 17,380 " 510 yrs ago indicates that the ice tongue had receded about 40% of its overall length by that time. Comparing our chronology with distances of retreat suggests that the Ohau glacier terminus receded at a mean net rate of about 77 m yr#1 and its surface lowered by 200 m between 17,690 and 17,380 yrs ago. A long-term continuation of ice retreat in the Ohau glacier catchment is implied by moraine records at the head of Irishman Stream valley, a tributary of the Ohau glacier valley. The Irishman Stream cirque glacier advanced to produce a set of Lateglacial moraines at 13,000 " 500 yrs ago, implying that the cirque glacier was less extensive prior to that advance. We employed a glaciological model, ﬁt to these mapped and dated LLGM moraines, to derive snowline elevations and temperature parameters from the Ohau glacier record. The modeling experiments indicate that a snowline lowering of 920 " 50 m and temperature depression of 6.25 " 0.5 ! C below modern values allows for the Ohau glacier to grow to an equilibrium position within the LGM moraine belt. Taken together with a glaciological simulation reported from the Irishman Stream valley, snowlines and temperatures increased by at least w520 m and w3.6 ! C, respectively, between w18,000 and w13,000 yrs ago. Climate parameters derived from the Ohau glacier reconstruction are similar to those derived from glacier records from Patagonia, to air temperature indicators from Antarctica, as well as to sea-surface temperature and stratiﬁcation signatures of the Southern Ocean. We think that the best explanation for the observed southern LLGM is that southern winter duration modulated Southern Ocean sea ice, which in turn inﬂuenced Southern Ocean stratiﬁcation and made the surface ocean cooler. Orbitally induced cooling of the Southern Ocean provides an explanation for the LLGM in the Southern Alps having Keywords: Last Glacial Maximum Southern Ocean Glacier Geomorphology 10 Be Surface-exposure dating Cosmogenic nuclide Paleoclimatology Southern Alps New Zealand Antarctica Patagonia Glaciology Snowline Last Glacial termination * Corresponding author. Lamont-Doherty Earth Observatory, 61 Rt. 9W, Palisades, NY 10944, USA. E-mail address: aputnam@ldeo.columbia.edu (A.E. Putnam). 1 Deceased. 0277-3791/$ e see front matter ! 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.quascirev.2012.10.034 Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 115 been coincident with the northern LGM. We argue further that the global effect of North Atlantic stadials led to disturbance of Southern Ocean stratiﬁcation, southward shifts of the subtropical front, and retreat of Southern Alps glaciers. Collapse of Southern Ocean stratiﬁcation during Heinrich Stadial-1, along with attendant sea-surface warming, triggered the onset of the Last Glacial termination in the Southern Alps of New Zealand. ! 2012 Elsevier Ltd. All rights reserved. 1. Introduction 2. Setting Late Quaternary ice-age cycles have constituted the fundamental pulse beat of Earth’s climate system over the past one million years, but a complete explanation of these cycles has yet to emerge. Variations of ice volume, dominated by Northern Hemisphere ice sheets, resemble orbital frequencies in general (Hays et al., 1976) and the summer insolation intensity signal at 65! N in particular (Broecker, 1966; Broecker and van Donk, 1970), as was predicted by Milankovitch (1941). Additional support for Milankovitch (1941) comes from the close match between the rate of change of global ice volume and June insolation at 65! N (Roe, 2006). In particular, Northern Hemisphere ice sheets most recently attained maximum sizes during a prominent trough in the boreal summer insolation intensity curve at 26,500e 19,000 yrs ago, known classically as ‘the Last Glacial Maximum’ (LGM; Mix et al., 2001; Clark and Mix, 2002; Clark et al., 2009; MARGO Project Members, 2009). But there are problems with regard to the Southern Hemisphere. For example, Hays et al. (1976) showed that, over the last 300,000 yrs, biological indicators of Southern Ocean sea-surface temperature (SST) and water-column stratiﬁcation ﬂuctuated in phase with northern ice sheets, despite different signatures of precession-dominated insolation in each hemisphere. Moreover, Mercer (1984) identiﬁed simultaneous glacier expansion during the classic LGM in both southern and northern middle-to-high latitudes. Because summer insolation intensity is out-of-phase during the LGM at those latitudes, the apparent northesouth synchrony presents a conundrum that Mercer (1984) called “a ﬂy in the ointment of Milankovitch Theory” (p. 307). The solution to this problem requires an explanation for full-glacial climatic conditions in the Southern Hemisphere during the classic LGM. To address this issue, we start by documenting the initiation, duration, magnitude, and termination of the Local Last Glacial Maximum (LLGM) cold period in the Southern Alps of New Zealand. We focused on the geomorphology and chronology of an exceptionally well-preserved set of moraines deposited by an ice-age glacier, referred to here as the Ohau glacier, which occupied the Lake Ohau valley on the eastern side of the Southern Alps. This moraine chronology utilizes recent analytical improvements in 10 Be surface-exposure dating (e.g., Schaefer et al., 2009), as well as a local 10Be production-rate calibration site (Putnam et al., 2010b). Glaciers in the Southern Alps are regarded as sensitive recorders of atmospheric temperatures (e.g., Oerlemans, 1997; Anderson and Mackintosh, 2006; Anderson et al., 2010; Purdie et al., 2011). We then employ a version of the University of Maine Ice Sheet glaciological model (UMISM) adapted to mountain-glacier reconstruction to estimate the temperature conditions, relative to modern values, required to generate ice of LLGM extent in the Ohau valley. Using the moraine chronology and glacier-climate reconstruction, we assess potential drivers of Southern Hemisphere ice-age climate. In this regard, the Southern Alps are well positioned for evaluating hypothesized mechanisms of global climate change because they are at the antipode of the North Atlantic region, far distant from northern ice sheets as well as from sites of North Atlantic Deep Water formation. South Island, New Zealand (40e46! S, 164e170! W) is situated just north of the modern position of the subtropical front (STF) in the South Paciﬁc Ocean, in the southern westerly wind belt (Fig. 1). Produced by oblique convergence of the Indo-Australian and Paciﬁc plates, the Southern Alps form the spine of South Island and reach altitudes in excess of 2000 m on the main drainage divide (hereafter “Main Divide”). Heavily faulted, uplifted Mesozoic sedimentary rocks constitute the central Southern Alps east of the Main Divide. Highly indurated quartzofeldspathic greywacke sandstones and argillite mudstones are among the most common lithologies (Cox and Barrell, 2007). The Southern Alps impose a strong föhn effect on the prevailing westerly winds, inducing 10 m yr#1 or more precipitation (water equivalent) just northwest of the Main Divide. Precipitation declines with increasing distance southeast of the Main Divide (Henderson and Thompson, 1999). More than 3000 inventoried glaciers cover 1158 km2 of the Southern Alps (Chinn et al., 2005). The present-day snowline is w1500 m above sea level (a.s.l.) just west of the Main Divide at Franz Josef Glacier, and trends upwards to >2200 m a.s.l. east of the Main Divide at the Ben Ohau Range (Lamont et al., 1999). High precipitation totals in ice catchment basins of the central Southern Alps result in large ice ﬂuxes through outward-ﬂowing glaciers. Such rapid ice throughput elevates glacier sensitivity to atmospheric temperature variations, resulting in response times to climate change of decades or less for many glaciers of the Southern Alps (e.g., Anderson et al., 2010). Historical records of variations of glacier lengths show that non-calving glacier termini have ﬂuctuated in-step during the last century on both ﬂanks of the Southern Alps. Thus, glaciers on opposite sides of the Main Divide register a common climate signal on decadal-to-centurial timescales (Chinn, 1996; Burrows, 2005; Barrell et al., 2011). In foreland regions, moraine belts outline former ice tongues that ﬂowed outward from the Southern Alps during recent glaciations (Suggate, 1990; Barrell, 2011; Barrell et al., 2011), and associated glacioﬂuvial outwash plains extend coastwards. Northwest of the Main Divide, post-glacial coastal erosion has truncated some distal parts of the LLGM moraine belts in some valleys. However, well-preserved moraine systems rim the LLGM glacial troughs that contain Lake Tekapo, Lake Pukaki, and Lake Ohau at the western margin of the Mackenzie basin southeast of the highest sector of the Southern Alps. The glaciers that occupied these troughs deposited numerous lateral and terminal moraine ridges during and following retreat from the LLGM. Moraines formed subsequently during Lateglacial time lie farther toward the heads of mountain valleys (Kaplan et al., 2010; Putnam et al., 2010a). These post LLGM landforms provide a basis for assessing glacier and climate changes during the Last Glacial termination. 3. Geomorphology 3.1. Overview of the Lake Ohau study area The Ohau glacier tongue was produced by the coalescence of major tributary glaciers from the Hopkins River and Dobson River Author's personal copy 116 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Fig. 1. Right panel: New Zealand with schematic depiction of surrounding physical oceanography (adapted from Carter et al., 1998). Arrows indicate surface-ocean currents, with purple denoting subtropical currents, red representing ﬂow along the Subtropical Front (STF), and blue illustrating currents associated with the northern boundary of the Antarctic Circumpolar Current and the Subantarctic Front (SAF). Elevation color scale is inset. Left panel: Wind speed just west of New Zealand plotted versus latitude [derived from ECMWF ERA-Interim reanalysis data; averaged over the period AD1979 to AD2011 (Dee et al., 2011)]. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.) valleys, both of which drain from the Main Divide. The overall catchment for this system at the LLGM was w900 km2. The Ohau glacial geomorphological assemblage is shown in Fig. 2 and is illustrated in Fig. 3. Farthest outboard are remnants of moraines and outwash older than the LLGM, while the LLGM moraine belts are more continuous and lie closer to the glacier trough. The assemblage of moraines and outwash surfaces has been inﬂuenced by the effects of uplift on the northwest side of the tectonically active Ostler Fault (Fig. 2) (Cox and Barrell, 2007). Outwash plains spilled through four gaps across the Ostler Fault uplift zone (Amos et al., 2007; Ghisetti et al., 2007). The Quailburn, Willowbank, Clearburn, and Ohau River gaps are former glacier meltwater spillways, with post-glacial drainage from Lake Ohau conﬁned to the modern Ohau River (Fig. 2). Time-transgressive eastward meltwater deﬂection controlled by uplift along the Ostler Fault was instrumental in preserving outer terminal moraine belts (Amos et al., 2007, 2010). We divide glacial landforms near Lake Ohau into two groups on the basis of morphology and soil development (Cox and Barrell, 2007; Barrell et al., 2011); these groups have been attributed to the last two major glacialeinterglacial cycles (Amos et al., 2007, 2010; Cox and Barrell, 2007; Barrell et al., 2011). Detailed geomorphological interpretations are shown in Fig. 2. Our glacial geomorphology mapping distinguishes topographic ridges of moraine from all other hummocky or irregular morainal landforms. In accord with the approach of Sugden and John (1976), we interpret an individual moraine ridge as marking a former position of the glacier margin, representing either a stillstand or a culmination of an advance of the ice margin. The material forming the crest of a moraine ridge was deposited immediately prior to withdrawal of the ice margin from that ridge. Lake Ohau stands at 520 m above sea level (a.s.l.), and the Ohau moraine complex rises to as much as w700 m a.s.l. Pollen records from a site at similar altitude 20 km northeast of Lake Ohau (McGlone and Moar, 1998) indicates an early Holocene forest cover of shrub and small-tree species, but that from mid-Holocene onwards, natural ﬁre events led to a more open shrubland and grassland. Nothofagus forest spread into at least some parts of the area in the late Holocene. During the past w800 years since human arrival, ﬁre markedly reduced the native forest cover, resulting in widespread grassland (McGlone and Moar, 1998). Since European settlement w150 years ago, frequent burning to assist pastoral farming has produced the modern vegetation of various introduced grasses, native short tussock (Festuca novae-zelandiae) and scattered patches of native and introduced shrubs. Introduced mammals, particularly rabbits and sheep, have generated small areas of soil erosion on the landforms. In recent years, a move away from burning practices has resulted in numerous wild-spreading patches of introduced pines and, on bouldery moraine ridges near Lake Ohau, considerable recolonization by native shrubs and tussocks. Patches of native forest persist in some tributaries of Lake Ohau, giving way to subalpine shrub and grassland above the local natural treeline at w1200 m a.s.l. 3.2. Ohau I landforms The outermost and hence oldest glacial landform assemblage (Ohau I) lies south and southeast of Lake Ohau and is mapped in green and olive in Fig. 2. The Ohau I landscape is gentle and rolling. Individual moraine ridges are typically w2e5 m high. Channeled outwash surfaces have a generally subdued form with patches of loess draped on sheltered surfaces. Boulders are few, and they commonly have shattered surfaces with spalled material accumulated around their bases. The southeast sector of this landform assemblage is preserved at Table Hill, which has been uplifted, and tilted back to the northwest, by movement on the Ostler Fault (Amos et al., 2007; Ghisetti et al., 2007) (Figs. 2 and 4). At the crest of the Ostler Fault scarp, which forms the southeastern edge of Table Hill, vertical offset of the Ohau I landform assemblage is at least w180 m, based on the Author's personal copy Fig. 2. Glacial geomorphologic map of Lake Ohau study area. Panel A. Index map of Lake Ohau catchment. Study area location is given on regional New Zealand map (inset). Blue line delineates catchment. Panel B. Detailed geomorphologic map of the Lake Ohau valley and foreland area. Roman numerals correspond to glacial geomorphologic landform assemblages described in text. ORG: Ohau River gap. CG: Clearburn gap. WG: Willowbank gap. QG: Quailburn gap. Yellow dots are locations of boulders sampled for 10Be surfaceexposure dating. Geomorphologic legend is inset. Panel C. Map of Irishman Stream valley tributary of the Lake Ohau catchment. Panel D. Detailed geomorphologic map of glacier landforms at the head of Irishman Stream, adapted from Kaplan et al. (2010). Geomorphologic legend is inset. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.) Author's personal copy 118 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Fig. 3. Panorama of the Ohau terminal moraine complex. View is to the north. Terminal moraine ridges, ice-contact outwash plains, relict braided stream channels, and kettle lakes occupy the foreground. Lake Ohau is at center distance, with the Hopkins (left) and Dobson (right) river valleys in the center far distance. Ben Ohau Range is in the middle distance right of center, and Ohau Range is in distance on the left. height of the scarp and the observation that Ohau IIeIII outwash lies, and presumably buries Ohau I deposits, on the downthrown side of the fault. The Table Hill glaciogenic deposits have been mapped in two regional formations; Balmoral Formation and, for older deposits on the crest of the fault scarp, Wolds Formation, by Gair (1967), Cox and Barrell (2007) and Amos et al. (2007). The southern sector of the Ohau I landform assemblage comprises subdued moraines and outwash deposits north and west of Quailburn and Willowbank gaps. The trends of braided stream channels preserved on the outwash deposits indicate that meltwater ﬂowed through the Willowbank and Quailburn gaps during construction of the Ohau I moraines. 3.3. Ohau II landforms The moraines and outwash surfaces constituting the Ohau II landform assemblage (Figs. 2 and 5), situated inboard of Ohau I landforms, have a surface form that is distinctively sharper than the Ohau I landforms. For the most part, the Ohau II landforms are patchy remnants, mainly of moraine and moraine ridges, preserved as islands surrounded by extensive Ohau III outwash plains (Figs. 2, 3, and 6). Moraine ridges in the southern portion of the Ohau II landform complex are broad and discontinuous. They stand 2e5 m high above neighboring general moraine or outwash surfaces. In contrast, the outer moraine ridges in the southwestern portion of the Ohau II landform complex are relatively sharp and continuous, with ridges typically 5e10 m high. The landforms mostly have stony surfaces, with rounded gravel common on general moraine. Scattered on the landforms are greywacke and semischist boulders (i.e., textural zones I and IIA/IIB, based on the classiﬁcation used by Cox and Barrell, 2007). The Ohau II glaciogenic deposits, along with those of Ohau III, are correlated with the regional Mt. John Formation (Gair, 1967; Amos et al., 2007; Cox and Barrell, 2007). The sharp-crested outer moraine ridges in the southwestern portion of the moraine complex are associated with aggradational outwash deposits, implying that these moraines were formed at the culmination of ice-margin advances (see Fig. 2). Relatively subdued inboard Fig. 4. Table Hill and Ohau I moraine and outwash complex. View is to east, with Ostler Fault scarp, partly in shadow, delimiting the right (southeast) margin of Table Hill, with several landslides on the scarp face, and arrays of sinuous slump scars along the scarp crest. Ohau I moraine ridges trend from bottom to top of image on the northwesterly backtilted Table Hill platform. Fluvially eroded slopes deﬁne western and northern limits of Ohau I complex in center foreground and center distance, respectively. Ohau River, the Ohau and Pukaki hydro-electric canals, and Lake Ruataniwha (artiﬁcially dammed) are visible in distance. Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Fig. 5. Photograph of Ohau II (foreground, center-left), Ohau III (middle ground), and Ohau IV (left-distance) terminal moraine ridges. View is north. Prominent Ohau III icecontact outwash plains with well-preserved relict braided stream channels are visible at bottom and middle of photograph. Aoraki/Mount Cook is tallest glaciated peak in far distance. topography, which includes moraine ridges in the southern sector of the Ohau II belt, suggests deposition during recession of the Ohau glacier terminus. 3.4. Ohau III landforms Situated inboard of, and locally truncating, the Ohau II landforms is the Ohau III geomorphologic assemblage. The Ohau III glaciogenic deposits are correlated with the regional Mount John Formation (Gair, 1967; Amos et al., 2007; Cox and Barrell, 2007). The Ohau III moraine complex rims the southern perimeter of the Ohau glacier trough, from the eastern foot of the Ohau Range to the southern base of the Ben Ohau Range (Figs. 2 and 5). There are three main elements of the Ohau III complex. Sets of prominent, sharpcrested moraine ridges form the outer part of the complex. These ridges, standing as much as w10 m above the adjacent ground, are among the tallest and most continuous moraine ridges observed in the entire study area. A central zone of the Ohau III moraine belt comprises rather featureless undulating ground moraine topography with numerous shallow lakes or depressions and relatively few moraine ridges; on the innermost side of Ohau III is a prominent, eroded ice-contact slope. Well-developed outwash plains 119 Fig. 6. Panel A. Oblique aerial photograph with view southwest toward Ohau IV moraine ridges and associated outwash (foreground), as well as Ohau III moraine ridges and associated outwash (distance, beyond prominent eroded ice-contact slope at far margin of Ohau IV outwash plain). Ohau Canal (lower) and Ohau River (center) drain left from the lake. Panel B. Photograph of Ohau IV outwash plain (middle). Vantage is west. Prominent Ohau IV terminal moraine ridge is in distance, and sharply deﬁned eroded-ice-contact slope of innermost Ohau III is in immediate foreground. Ohau Range deﬁnes skyline. Note exceptionally well-preserved relict braided stream channels on Ohau IV outwash surface. emanating from the Ohau III moraine belt extend through the Clearburn gap and the Ohau River gap (Figs. 2 and 5). The Ostler Fault has produced a w20 m vertical offset of Ohau III outwash surface at both gaps (Amos et al., 2007, 2010). The fault is interpreted to be a northwest-dipping listric thrust, on account of a substantial backtilt of the outwash surfaces, localized within w2.5 km northwest of the fault scarp. This tilt pattern implies a rapid northwestward diminution of the tectonic uplift (Davis et al., 2005; Amos et al., 2007; Ghisetti et al., 2007). As a result, vertical movement associated with the Ostler Fault is highly localized along the hanging wall, and uplift at the Ohau IIeV moraine belts is negligible (e.g., <1 mm yr#1) in regard to surface-exposure dating calculations. The middle to inner parts of the Ohau III moraine belt south of Lake Ohau exhibit a combination of generally featureless undulating topography, scattered discontinuous moraine ridges w1e5 m high, numerous shallow lakes and an absence of abutting outwash deposits. This landscape formed during general recession of the Ohau glacier terminus. The innermost Ohau III eroded ice-contact slope extends laterally for w13 km, stands w30 m above the Ohau IV outwash surface (Fig. 6), and approximates the outer margin of the trough into Author's personal copy 120 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 which the Ohau glacier withdrew, following formation of the Ohau III moraine/outwash complex. This ice-contact slope could also represent an ice-margin advance or stillstand. 3.5. Ohau IV landforms The Ohau IV moraine belt, dominated by prominent, sharpcrested ridges, forms a rim alongside the southern and eastern shores of the lake (Figs. 2 and 6). The associated outwash plain has been incised w20e30 m by the modern Ohau River (Fig. 6). Many sub-rounded to sub-angular greywacke boulders, some exceeding 4 m in height, are embedded in or rest on the moraine ridges, as well as on the ice-contact heads of Ohau IV meltwater channels and outwash surfaces. The Ohau IV glaciogenic deposits, along with those of Ohau V and VI, are correlated with the regional Tekapo Formation (Gair, 1967; Amos et al., 2007; Cox and Barrell, 2007). 3.6. Ohau V landforms The innermost features of the Lake Ohau terminal moraine complex are subdued discontinuous ridges along the southern shoreline of Lake Ohau that, at their northeastern ends, project into the Lake Ohau basin (Fig. 2). Two arcuate moraine ridges projecting northeastward into the lake mark the last positions occupied by the Ohau glacier before it withdrew into what is now the lake basin. Geomorphologic relationships indicate that these moraines were constructed at the edge of newly formed Lake Ohau (Fig. 2). 3.7. Ohau VI landforms Glaciogenic landforms representing ice retreat from the Ohau V terminal positions are designated here as Ohau VI. These landforms include well-expressed lateral recessional moraines on the western side of the Ohau valley near the head of the lake, but discrete landforms are more difﬁcult to delineate on the eastern side of the valley, where broad ice-marginal benches span a large altitudinal range. The upper part of this ﬂight of benches may reﬂect recession from older parts of the Ohau glacial landform sequence (e.g., Ohau III). Included within the Ohau VI landform assemblage is an icemolded bedrock bench at the junction of the Hopkins and Dobson rivers w24 km north of the Ohau V terminal moraines (Figs. 2 and 7). A patchy veneer of lodgement till and scattered large greywacke boulders rest on the molded bedrock. This bedrock bench was exposed as the Ohau glacier receded from the southern end of the lake toward the head of the Lake Ohau catchment. 3.8. Irishman Stream moraines At the LLGM, glacier ice from the Irishman Stream tributary valley merged with the Ohau glacier. Following substantial ice recession, glacier ice readvanced, producing prominent bouldery moraines within the cirque at the head of the Irishman Stream valley (Fig. 8). These moraines have been documented in detail by Kaplan et al. (2010), and are the youngest glacial geomorphologic features considered in this study. 3.9. Summary of geomorphological interpretations We interpret the following sequence of events from the Lake Ohau glacial geomorphology. The Ohau glacier achieved its greatest recognized extent prior to the LLGM, as documented by the Ohau I landform assemblage. Incomplete preservation of the Ohau I moraine belt, exacerbated by uplift and tilting due to the Ostler Fault, reduces the certainty with which the ice extent and local paleogeography can be inferred. After uplift and tilting of the Ohau I landforms, the Ohau glacier expanded to construct the outboard Ohau II landforms, followed by ice retreat that produced the Fig. 7. Oblique aerial photograph of ice-molded landforms near the junction of the Hopkins (background) and Dobson (behind the photographer) rivers. Vantage is west. These bedrock landforms comprise part of the Ohau VI landform assemblage. Note broad ice-scour troughs and ridges aligned right to left on the crest of the bench, parallel to the Hopkins valley ice-ﬂow direction, and a dark line in the foreground at the edge of the bench crest, marking a low (w1 m high) Dobson valley recessional lateral moraine. Barrier Range deﬁnes the skyline. Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 121 (Fig. 9). Integrated with the geomorphologic mapping study, we targeted speciﬁc moraine crests for sampling based on the following criteria: (1) crests represent single, distinguishable icemarginal positions; (2) moraines do not appear to have been reworked by post-depositional processes, including human inﬂuence; and (3) landforms are clearly deﬁned and retain pristine features such as perched erratics and abandoned ice-marginal outwash fans with preserved braided stream channels (e.g., Fig. 6), all indicating little landform change since initial formation. Where possible, we sampled several boulders on each moraine ridge. Boulders were selected based on the following considerations: (1) Geomorphologic integrity. We assessed the immediate surroundings of each sampled boulder to ensure there were no indications of post-depositional disturbance by slope processes, ﬂuvial reworking, or human activities. Ideal boulders are those rooted in moraine ridge crests, and they occur away from steep slopes of landforms that may have experienced degradation as well as adjacent hill slopes that may have shed rock-fall debris onto the moraine area. We generally did not sample boulders with heights less than 50 cm above ground level. (2) Boulder surface integrity. We avoided sampling boulders that are heavily jointed, fractured, pitted, and/or exhibit evidence of spalling or granular disintegration. Preference was given to sampling from the top center of boulders with near-horizontal, planar, or gently rounded upper surfaces, keeping away from any steep edges at the boulder sides. Fig. 8. Irishman Stream moraine set. A.) Overview photograph of glacier landforms at the head of Irishman Stream valley. Vantage is east. B.) Photograph of outer prominent bouldery moraine ridge at head of Irishman Stream valley. Vantage is west (depicted by arrow in Panel A). inboard Ohau II recessional moraines. A subsequent ice advance, which in places overran parts of the Ohau II moraine belt, culminated with formation of the near-continuous outer Ohau III moraine ridges. The inner w3 km of the central portion of Ohau III moraine belt comprises subdued and discontinuous moraines formed during recession from the Ohau III maximum. The eroded ice-contact slope that forms the inner margin of the Ohau III moraine belt is thought to mark the edge of the glacier trough as it existed at Ohau III time. Alternatively, this ice-contact slope may represent an ice-margin advance or stillstand position. The Ohau IV moraines were formed at the culmination of a subsequent resurgence of the glacier, by which time meltwater outﬂow had become conﬁned to the Ohau River gap. Moraine ridges of the inboard Ohau V moraine belt mark brief pauses in recession from the Ohau IV ice limit. Subsequent terminus retreat and ice-surface lowering exposed the bedrock bench near the Hopkins/Dobson junction that comprises part of the Ohau VI landform assemblage. Recession continued well into mountain valleys, after which time a glacier at the head of Irishman Stream tributary valley (and probably other glaciers in the Ohau catchment headwaters) readvanced to produce Lateglacial moraines (Kaplan et al., 2010). 4. Methods 4.1. Sample collection We sampled only boulders on landforms with clear geomorphologic context relating directly to a former glacier margin Boulder surfaces were sampled using hammer and chisel or the ‘drill-and-blast’ method of Kelly (2003). We used clinometer and compass to document the surrounding topographic shielding for all sample locations. We photographed all boulders from several different perspectives, and we measured the ground-to-sample height on four sides of every boulder. 4.2. 10 Be extraction and AMS measurements Samples were processed for 10Be analysis at the LamontDoherty Earth Observatory (LDEO) Cosmogenic Isotope Laboratory using the methods of Schaefer et al. (2009). The LDEO procedure is available online at: http://www.ldeo.columbia.edu/tcn. Beryllium ratios (10Be/9Be) were measured with the CAMS accelerator at the Lawrence-Livermore National Laboratory. Samples processed before September 2007 were measured relative to the KNSTD standard (10Be/9Be ¼ 3.15e#12). Samples processed during and after September 2007 were measured relative to the 07KNSTD standard [10Be/9Be ¼ 2.85e#12; based on a revised 10Be half-life of 1.36 " 0.07 Ma (Nishiizumi et al., 2007)]. After measurement, 10 Be/9Be ratios were corrected for residual boron contamination and 10Be in procedural blanks. All blank corrections were less than 1.5%. All samples measured relative to the KNSTD standard were subsequently normalized to 07KNSTD by applying a correction factor of 0.9042 (Nishiizumi et al., 2007). 4.3. Exposure-age calculations We calculated surface-exposure ages using the sea-level highlatitude (SLHL) 10Be production rates of Putnam et al. (2010b) and scaling methods of Stone (2000; ‘St’), Desilets et al. (2006; ‘De’), Dunai (2001; ‘Du’), and Lifton et al. (2005, 2008; ‘Li’), as well as a version of the Stone (2000) scaling that incorporates a highresolution version of the Lifton et al. (2008) geomagnetic model (Putnam et al., 2010b), labeled ‘Lm’. Abbreviations follow the Author's personal copy 122 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Fig. 9. Examples of boulders sampled for 10Be surface-exposure dating. Panel A. Photograph, with vantage north, of boulder embedded in Ohau VI ground moraine overlying icemolded bedrock. Panel B. Photograph, vantage northwest, of large boulder embedded in top of outermost Ohau IV moraine ridge ice-contact slope. Panel C. Photograph, vantage west, of boulder resting on top of outermost Ohau III moraine ridge crest. Panel D. Photograph, vantage west-southwest, of large boulder embedded in ice-distal slope of outermost Ohau III moraine ridge. Panel E. Photograph, vantage north, of boulder resting on subdued Ohau II moraine ridge remnant that is surrounded by Ohau III outwash. Panel F. Photograph, vantage east, of boulder resting on Ohau II moraine ridge. nomenclature of Balco et al. (2008). Tectonic and glacioisostatic uplift rates at the Lake Ohau moraines are assumed to be negligible (Putnam et al., 2010b) for moraines away from the Ostler Fault (i.e., all moraines except for Ohau I). Even for the Ohau I moraine, a landscape uplift rate of 1 mm yr#1 (the proposed localized uplift rate on the northwest side of the Ostler Fault scarp adjacent to the study area; Amos et al., 2010) would decrease ages by less than 1% (i.e., within the analytical uncertainty of the data, reported below). Accordingly, we did not apply a correction for the effects of landscape uplift to any of the landforms dated in this study. Changes in local atmospheric pressures at sea level during the last ice age (i.e., since boulder deposition) could have resulted in slightly higher production rates due to equatorward-shifted westerlies and hence lower atmospheric pressures in southern middle latitudes (e.g., Staiger et al., 2007). In this regard, a 14C-dated LLGM-age production-rate test site reported by Putnam et al. (2010b) veriﬁed the suitability of an early-Holocene production-rate calibration data set for calculating the age of an w18,000-yr surface, indicating that temporal shifts in average air pressure, if any, have not had any discernible inﬂuence on 10Be production since the end of the LLGM. This ﬁnding is consistent with the modeling results of Staiger et al. (2007) for the atmosphere over New Zealand. For these reasons we did not apply a correction for changes in atmospheric pressure. An erosion rate was not incorporated into age calculations due to the resistant nature of the greywacke boulder surfaces (Birkeland, 1982; Schaefer et al., 2006, 2009). Furthermore, as no erosion correction was applied to the production-rate calibration data set of Putnam et al. (2010b), any common effects of erosion on this rock type have already been integrated into the SLHL 10Be production rate. We followed Schaefer et al. (2006, 2009) and Putnam et al. (2010b) and did not apply a snow-cover correction for the following reasons: (1) snowfalls are infrequent at elevations below w1500 m a.s.l. in the Southern Alps, (2) at lower elevations, winter snow rarely exceeds depths of w1 m and tends to melt away within days to weeks, and (3) salient boulder tops and moraine ridge crests are generally windswept and free of snow throughout the winter. Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 123 4.4. Glaciological modeling We constrained a glaciological model to mapped and dated moraines (Fig. 2) in order to estimate likely LLGM climatic parameters and average equilibrium-line altitude (hereafter, snowline) for the Ohau glacier. This aspect of work was carried out using the University of Maine Ice Sheet Model (UMISM), a 2D ﬁnite-element mass and momentum ice dynamics solver with embedded components for calculating isostasy, thermodynamics, sliding, and surface mass balance (Fastook and Prentice, 1994; Fastook et al., 2008) (see Appendix). The force balance in UMISM follows the Shallow Ice Approximation (SIA) (vertically integrated momentum combined with continuity) where the dominant stress is internal shear, and where longitudinal stresses are neglected. Although models of this sort are most commonly used for simulating ice sheets, they are also applicable in mountain settings for large trunk glaciers ﬂowing over gently sloping beds (Le Meur and Vincent, 2003; Le Meur et al., 2004, 2007; Leysinger Vieli and Gudmundsson, 2004; Schäfer et al., 2008). Other SIA modeling studies over domains comparable in scale to the Ohau valley include the Wasatch (Laabs et al., 2006) and Uinta (Refsnider et al., 2008) mountains, Utah, USA, and the Wind River Range, Wyoming, USA (Birkel et al., 2012). Primary input to UMISM was a 500-m resolution bedrock elevation map (resampled from 90 m SRTM data) arranged in a quadrilateral grid of nodal points over the Ohau LLGM ice catchment. Initial climate was deﬁned by monthly temperature and precipitation values for New Zealand extracted from the WorldClim data set (Hijmans et al., 2005). WorldClim is a 1 km gridded AD1950e2000 climatology derived from the interpolation of a global array of weather station data using thin plate smoothing splines. This baseline modern climate input was adjusted in the LLGM simulations by shifting the temperature seasonal cycle according to prescribed anomalies (e.g., DT ¼ #5 ! C implies 5 ! C cooling), and by scaling precipitation (e.g., P ¼ 75% implies 25% reduction). In order to account for evolving ice topography, the monthly temperature ﬁelds were modiﬁed using an atmospheric lapse rate of #5 ! C km#1, based on modern observations (Norton, 1985). However, we recognize that this is just a working scenario, as the value may have varied through time. In a sensitivity test of the LLGM Ohau glacier, we found that increasing the lapse rate to #6 ! C km#1 required only a þ0.2 ! C DT compensation relative to the #5 ! C km#1 baseline simulation in order to achieve the moraine-deﬁned glacier footprint. The climate signal is transferred to the ice-ﬂow solver through a calculation of net annual surface mass balance, or the difference between snow/ice accumulation and ablation. This difference is obtained using a typical degree-day method (e.g., Braithwaite and Olesen, 1989). Although degree-day mass-balance schemes lack the complexity of energy-balance models (which consider spatial and temporal variations in albedo, insolation, and cloudiness), they are widely used in glacier modeling studies for their ease of implementation and overall good ﬁt to observational data (Braithwaite, 2011). In our modeling, annual snow accumulation was determined by summing the precipitation amounts for all months when the temperature, Tmonth, is &0 ! C. Ablation totals were generated by ﬁrst summing melting degrees (md) (mdmonth ¼ Tmonth ' number of days in month) for all months when Tmonth is >0 ! C. Accumulated snow was then reduced at a prescribed snowmelt rate (mm per md). If melting degrees remained once all winter snow had been ablated, ice was in turn melted at an ice-melt rate. We used melt factors for snow and ice of 4.6 and 7.2 mm md#1, respectively, determined empirically for New Zealand (Anderson et al., 2006). Fig. 10 shows how choice of melt rates can inﬂuence snowline. Ice sheet models such as UMISM have changeable parameters that to some extent control the behavior of ice ﬂow. In order to Fig. 10. Mass balance versus elevation plots for the Ohau glacier model domain for three different snow/ice melt rate (SMR/IMR) regimes with base climate at DT ¼ #6.2 ! C, P ¼ 100%, and atmospheric lapse rate ¼ #5 ! C km#1. A) SMR ¼ 4 IMR ¼ 7, B) SMR ¼ 4.6 IMR ¼ 7.2, and C) SMR ¼ 5 IMR ¼ 8 (units mm md#1). Snowline, denoted by blue lines, ranges 1474e1553 m, which affords an uncertainty of "50 m relative to the 1525 m value derived from New Zealand-speciﬁc degree-day factors. Further tests indicate that a 50 m rise or fall in snowline can be compensated by changing DT w0.25 ! C. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.) Author's personal copy 124 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 produce meaningful paleo-glacier reconstructions, it is therefore necessary to tune the model so that output is consistent with existing observational data. The chief ice-ﬂow parameter tuned for this study was the sliding law constant, B (see Appendix). The value of B can be adjusted either to lessen or increase the amount of basal sliding due to melting and refreezing over rough beds. Low values afford relatively thin, fast-ﬂowing ice, whereas high values afford the opposite. Another parameter, E, the ice-ﬂow enhancement factor, can be tuned in order to change ice hardness to account for impurities. However, we left E at 1 (i.e., no ﬂow enhancement) because similar net functionality is achieved by varying the sliding law constant. Our experimental approach included more than a dozen sensitivity tests wherein DT was adjusted systematically for three scenarios of ice properties, deﬁned by B ¼ 0.02, 0.03, and 0.04 (Fig. 10). Success or failure was determined by whether or not the modeled Ohau glacier expanded to LLGM moraine limits. We also tested cases for two precipitation scenarios, P ¼ 100% and P ¼ 70%. The latter value accounts for hypothesized drier climate over South Island during glaciations (Drost et al., 2007; Whittaker et al., 2011). Taken together, the results from these experiments provide a means to estimate probable temperature conditions across the Ohau glacier at the LLGM. We note that our model is based on modern topography and lake-ﬂoor bathymetry, and it does not incorporate the effects of buried glacier trough geometry. We therefore take only the Ohau II/III simulation as generally indicative of LLGM ice extent and do not attempt to quantify a climatic value for the slightly shorter glacier length represented by the Ohau IV/V moraine belts. In addition, because we do not know whether or not the receding Ohau glacier was in equilibrium with climate when it exposed the bedrock bench at the Hopkins/Dobson conﬂuence, we did not use the glacier model to estimate climatic conditions associated with that recession, but instead characterized the deglaciation of the Ohau valley using glacier-length changes inferred from geomorphology. 5. Results 5.1. 10 Be surface-exposure dating Results from 10Be sample analyses and procedural blank Be/9Be ratios are given in Table 1a and b, respectively. Procedural blanks indicate a contribution of 2100e21,600 contaminating 10Be atoms (Table 1b), with an average contribution of 9760 " 5307 atoms ("1s). Combined, corrections for contaminating 10Be (determined using procedural blanks) and boron were generally w1%. Analytical uncertainties were typically 2e3%. Reported sample 10Be concentration uncertainties have been propagated with uncertainties in the measured blank 10Be concentration. All age calculations appear in Table 2, and have been referenced to calendar years BP (i.e., before AD1950, hereafter ‘yrs’ and ‘yrs ago’) by subtracting 56, 57, and 60 years for samples collected in AD2006, AD2007, and AD2010, respectively. Hereafter we discuss ages calculated using the ‘Lm’ scaling protocol [incorporating the highresolution geomagnetic model of Lifton et al. (2008), following Putnam et al. (2010b)] because this method yields results that tend to agree best with independent 14C constraints in the Southern Alps. A correction for geomagnetic effects integrated over the past 20,000 years is about 0.5%. Choosing alternative scaling methods or geomagnetic models to calculate ages yields agreement to within about 1.5%, and therefore would not affect any of the conclusions below. Uncertainties of elevations measured using differential global positioning system (GPS) are <1 m and thus have negligible effects on age uncertainties. 10 All exposure ages are plotted on maps in Fig. 11a and b. Based on the observed consistency of ages, we combined surface-exposure dates for each moraine belt (Fig. 12). We used c2 statistics to test the assumption that multiple 10Be dates from a single landform assemblage register coeval exposure to the cosmic-ray ﬂux, representing the time of landform construction (Table 3). The c2 test analyzes whether dispersion among samples in each given data set can be explained by analytical uncertainties alone, or whether geomorphological factors, such as multiple episodes of deposition, predepositional (i.e., ‘inherited’) accumulations of cosmogenic 10Be, or post-depositional modiﬁcation processes, have inﬂuenced a particular data set (Balco and Schaefer, 2006). 10Be surfaceexposure ages determined from Ohau IIIeV moraine belts exhibit low c2 values and hence high levels of internal consistency, implying that each belt of moraine landforms was constructed over a period that was within the uncertainty of the 10Be dating method as applied here. Where distributions of individual 10Be ages for a group of associated landforms are approximately normal, we used Chauvenet’s Criterion to test for statistical outliers at the 95% conﬁdence level (Bevington and Robinson, 1992; Dunai, 2010). Outliers identiﬁed by this method were excluded from further statistical treatment. With the exception of ages from the Ohau I and II moraine belts, all other sample sets (Ohau III, IV, V, VI) have approximately normal distributions (Fig. 11). After removing outliers identiﬁed by the Chauvenet method, we calculated the arithmetic mean and standard deviation to provide a conservative measure of the landform age and associated uncertainty. We present all derived landform ages and uncertainties in Table 4. Finally, systematic uncertainties attending 10Be production-rate and scaling protocols must be taken into account when evaluating the full range of acceptable landform ages. Because we use the local 10 Be production rate of Putnam et al. (2010b), scaling uncertainties are negligible. Choosing a different 10Be production-rate value would shift all 10Be ages, and hence the whole moraine chronology, systematically. To illustrate the effects of production-rate uncertainties on the Ohau moraine chronology, we present in Table 4 ‘minimum’ and ‘maximum’ possible landform ages calculated using the upper (þ2.1%) and lower (#2.1%) error bounds, respectively, of the local 10Be production-rate (Putnam et al., 2010b). Landform ages are given below, and also presented in Table 4. It is important to note that landform ages should only be compared for a speciﬁc production-rate scenario. In the following presentation and discussion of landform assemblages, we use the mean age (third column of Table 4), based on the median value of the production-rate uncertainty range. 5.1.1. Ohau I The three ages obtained from Ohau I moraines at Table Hill are between w130,000 and 150,000 yrs (Figs. 11 and 12; Table 3), indicating deposition during Marine Isotope Stage 6. However, the ages are discordant (e.g., c2experimental ¼ 23.42 which is greater than c2expected of 5.99 at 95% conﬁdence). One possibility is that the range is due to differential surface erosion of the boulders, thus the oldest age of 149,500 " 3100 yrs is closest to the true age of the moraine belt. However, the youngest sample came from a prominent moraine ridge that lies several ridges inboard of the ridge with the two older samples. Thus, another possibility is that these moraines are of composite age, and that the age spread is correct. We accommodate these possibilities conservatively, by suggesting a tentative age of 138,600 " 10,600 yrs, based on the arithmetic mean and standard deviation of the three dates. This age assignment is preliminary, pending further quantitative information of the age(s) of the Ohau I moraines. Author's personal copy Table 1a Lake Ohau moraine complex surface-exposure sample details and CAMS laboratory no. Sample ID Be data. Carrier added (g)a 10 Be/9Be " 1s (10#14)b [10Be] " 1s (104 atoms g#1)c Average 9Be current (mA)d AMS stde 3.0021 5.0339 5.0195 0.2007 0.2008 0.2014 19.02 " 0.45 22.34 " 0.44 20.60 " 0.67 84.28 " 2.00 99.14 " 1.95 91.80 " 3.02 15.2 (2) 13.6 (3) 15.0 (3) 07KNSTDB1,B2 07KNSTDB1,B2 07KNSTDB1,B2 0.999 0.999 0.999 0.999 0.999 0.987 0.999 0.999 0.999 0.999 4.0076 4.0054 4.0033 4.0186 3.4861 4.0183 4.0032 4.0156 4.0265 3.0082 0.2031 0.1974 0.2000 0.2023 0.1813 0.1995 0.2023 0.2012 0.2002 0.2012 7.42 7.45 5.86 7.23 6.07 5.44 5.14 6.06 6.35 4.73 " " " " " " " " " " 0.22 0.22 0.19 0.27 0.18 0.22 0.19 0.22 0.22 0.15 24.90 24.31 19.30 24.09 20.94 18.13 17.11 20.04 20.82 20.72 " " " " " " " " " " 0.75 0.71 0.62 0.92 0.63 0.73 0.65 0.73 0.74 0.69 11.6 12.8 16.4 11.7 19.1 7.9 10.0 11.8 13.3 15.2 (3) (3) (2) (4) (4) (4) (4) (4) (3) (3) KNSTDB3 KNSTDB3 07KNSTDB4 KNSTDB3 07KNSTDB5 07KNSTDB4 07KNSTDB4 07KNSTDB4 07KNSTDB4 07KNSTDB1,B2 2.00 0.96 1.49 1.37 2.59 1.88 0.86 1.70 1.19 1.5 1.58 2.37 1.60 1.65 1.36 1.18 2.01 1.92 0.87 1.48 0.88 1.04 0.91 1.05 1.02 1.79 0.998 0.999 0.992 0.993 0.999 0.999 0.999 0.999 0.999 0.996 0.999 0.999 0.999 0.999 0.999 0.999 0.999 0.998 0.999 0.999 0.989 0.999 0.999 0.998 0.999 0.999 7.0417 7.0285 7.0495 7.0649 7.0605 7.0138 7.3478 7.0161 4.1080 4.0176 4.0301 5.0339 5.0195 5.0214 6.0094 5.0152 5.0092 5.0107 5.0771 5.0755 5.0042 5.013 5.0076 6.0114 6.0044 6.0062 0.1989 0.2035 0.1983 0.1980 0.2035 0.2026 0.2021 0.1987 0.2020 0.2018 0.1930 0.2008 0.2014 0.2008 0.1993 0.1983 0.2012 0.2005 0.2015 0.2002 0.1986 0.2015 0.2002 0.1989 0.2009 0.1988 6.77 8.33 8.65 8.60 8.58 8.95 8.97 9.01 5.05 4.81 5.15 5.85 5.61 4.92 6.54 5.79 5.59 5.67 6.12 6.07 6.03 5.68 5.62 6.65 6.77 6.91 " " " " " " " " " " " " " " " " " " " " " " " " " " 0.18 0.22 0.21 0.25 0.33 0.25 0.23 0.28 0.15 0.16 0.20 0.22 0.15 0.16 0.21 0.16 0.19 0.19 0.19 0.19 0.21 0.15 0.16 0.22 0.23 0.22 12.59 15.96 16.05 15.96 16.37 17.09 16.31 16.87 16.41 15.96 16.29 15.33 14.84 12.97 14.26 15.03 14.73 14.88 15.99 15.74 15.73 15.04 14.75 14.47 14.90 15.04 " " " " " " " " " " " " " " " " " " " " " " " " " " 0.34 0.43 0.40 0.47 0.64 0.48 0.43 0.53 0.50 0.53 0.63 0.59 0.41 0.43 0.47 0.44 0.53 0.52 0.52 0.50 0.54 0.42 0.44 0.48 0.51 0.48 22.2 17.9 20.9 14.2 14.6 14.4 18.3 16.8 10.2 11.2 10.7 10.8 15.6 17.8 15.8 15.4 15.2 15.6 14.7 15.4 12.9 16.1 15.3 14.8 16.5 15.5 (2) (2) (2) (2) (4) (3) (2) (3) (5) (4) (5) (2) (3) (2) (2) (3) (2) (2) (2) (2) (2) (3) (3) (2) (3) (2) KNSTDB6 KNSTDB7 KNSTDB6 KNSTDB7 KNSTDB7 KNSTDB8 KNSTDB8 KNSTDB8 KNSTDB3 KNSTDB3 KNSTDB3 07KNSTDB9 07KNSTDB4 07KNSTDB4 07KNSTDB1,2 07KNSTDB1,2 07KNSTDB1,2 07KNSTDB1,2 07KNSTDB9 07KNSTDB9 07KNSTDB9 07KNSTDB10,11 07KNSTDB1,2 07KNSTDB1,2 07KNSTDB1,2 07KNSTDB1,2 2.21 2.07 1.45 1.50 2.07 1.13 2.81 1.55 1.65 1.27 0.999 0.998 0.999 0.998 0.992 0.992 0.999 0.999 0.995 0.999 7.1038 7.0410 7.0320 7.3979 7.0396 7.0048 7.0591 7.0317 7.0789 7.0112 0.2003 0.2011 0.1920 0.1981 0.1981 0.1985 0.2038 0.2013 0.2027 0.2035 6.81 6.77 7.12 7.40 6.62 6.89 6.93 6.42 7.09 6.94 " " " " " " " " " " 0.20 0.21 0.24 0.19 0.18 0.18 0.21 0.19 0.23 0.20 12.65 12.78 12.85 13.05 12.26 12.86 13.19 12.09 13.42 13.29 " " " " " " " " " " 0.38 0.40 0.43 0.34 0.34 0.34 0.41 0.37 0.43 0.39 15.6 15.9 14.8 21.8 21.7 22.3 15.9 15.4 14.3 17.8 (2) (2) (4) (2) (2) (2) (2) (2) (2) (2) 07KNSTDB9 KNSTDB7 KNSTDB7 KNSTDB6 KNSTDB6 KNSTDB6 KNSTDB8 07KNSTDB9 KNSTDB7 KNSTDB8 Longitude (DD) Elevation (m a.s.l.) Boulder size (L ' W ' H) (cm) Sample thickness (cm) Shielding correction #44.297439 #44.297739 #44.298589 169.98659 170.00296 170.00239 605.2 615.4 621.1 240 ' 140 ' 120 250 ' 150 ' 120 400 ' 330 ' 100 2.00 1.19 1.89 0.999 0.999 0.999 #44.35097 #44.3489 #44.34733 #44.34317 #44.29125 #44.31537 #44.31787 #44.3236 #44.32126 #44.341789 169.89591 169.89264 169.90467 169.90149 169.98240 169.95898 169.96605 169.97621 169.94935 169.83328 599 600 586 590 579 569.1 564.5 554.5 570.5 680 230 160 220 320 320 160 370 200 410 200 ' ' ' ' ' ' ' ' ' ' 220 100 200 180 190 120 330 140 400 190 80 60 90 60 144 80 120 90 160 60 1.81 0.99 1.65 2.78 1.34 2.01 1.34 1.61 1.45 2.35 #44.25613 #44.25611 #44.24982 #44.25058 #44.31698 #44.32263 #44.32507 #44.32681 #44.33402 #44.33353 #44.33335 #44.30317 #44.3041 #44.30449 #44.30615 #44.30711 #44.30738 #44.33238 #44.33351 #44.33352 #44.33376 #44.33366 #44.33412 #44.322689 #44.323181 #44.323181 169.98217 169.98219 169.97236 169.97101 169.94778 169.94554 169.94361 169.9406 169.88676 169.88824 169.88743 169.95896 169.9588 169.95875 169.95875 169.95823 169.95822 169.87581 169.88021 169.88027 169.88213 169.88232 169.88575 169.87317 169.87543 169.87543 592.1 591.6 582.2 585.9 581.8 574.2 572.5 573.9 605 602 603 582.2 583.4 581.7 577.5 576.9 575.5 620 618 618 613 613 606 614 612 612 190 240 210 100 280 260 220 130 400 350 450 260 400 210 270 350 210 300 300 170 430 190 280 280 230 310 ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 60 ' 160 170 ' 70 160 ' 100 60 ' 70 150 ' 180 120 ' 140 100 ' 60 100 ' 80 240 ' 170 320 ' 130 320 ' 90 200 ' 200 300 ' 310 200 ' 180 230 ' 180 220 ' 160 110 ' 100 170 ' 80 90 ' 90 110 ' 60 250 ' 130 100 ' 50 210 ' 130 190 ' 70 200 ' 80 225 ' 90 #44.28244 #44.27704 #44.27659 #44.27492 #44.27046 #44.269896 #44.29886 #44.29853 #44.2974 #44.29836 169.94186 169.98217 169.94047 169.93968 169.93488 169.9301 169.90032 169.90072 169.90386 169.90023 557.3 561.4 559.0 561.2 564.1 546.3 583.5 583.6 578.3 575.3 370 210 300 230 280 270 550 390 280 600 ' ' ' ' ' ' ' ' ' ' 290 ' 50 160 ' 120 240 ' 120 180 ' 100 90 ' 110 180 ' 110 220 ' 400 260 ' 130 180 ' 210 400 ' 240 ' ' ' ' ' ' ' ' ' ' Quartz weight (g) 125 (continued on next page) A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Ohau I moraines BE24238 OH-07-04 BE24240 OH-07-44 BE24241 OH-07-45 Ohau II moraines BE23323 OH-06-60a BE23324 OH-06-61 BE24135 OH-06-62 BE23325 OH-06-63 BE27913 OH-07-01 BE24136 OH-07-16 BE24137 OH-07-18 BE24138 OH-07-19 BE24139 OH-07-21 BE24239 OH-07-26 Ohau III moraines BE22764 OH-06-04 BE22912 OH-06-05 BE22768 OH-06-16 BE22915 OH-06-18 BE22919 OH-06-44 BE22927 OH-06-45 BE22928 OH-06-46 BE22929 OH-06-50 BE23322 OH-06-59 BE23326 OH-06-72 BE23327 OH-06-73 BE24131 OH-07-08 BE24133 OH-07-10 BE24134 OH-07-11 BE24232 OH-07-13 BE24233 OH-07-14 BE24234 OH-07-15 BE24235 OH-07-31 BE24128 OH-07-33 BE24129 OH-07-34 BE24130 OH-07-36 BE24226 OH-07-37 BE24236 OH-07-38 BE24229 OH-07-39 BE24230 OH-07-40 BE24231 OH-07-41 Ohau IV moraines BE24125 OH-06-08 BE22913 OH-06-10 BE22914 OH-06-11 BE22765 OH-06-13 BE22766 OH-06-14 BE22767 OH-06-15 BE22924 OH-06-22 BE24126 OH-06-23 BE22916 OH-06-24 BE22926 OH-06-26 10 Latitude (DD) Author's personal copy CAMS laboratory no. Sample ID a [10Be] " 1s (104 atoms g#1)c Average 9Be current (mA)d AMS stde 0.21 0.20 0.19 0.21 0.22 0.21 0.20 0.21 0.25 13.55 12.85 11.61 13.41 13.21 12.52 12.57 13.57 12.66 " " " " " " " " " 0.41 0.37 0.36 0.39 0.39 0.39 0.38 0.40 0.47 16.5 19.3 15.6 17.5 17.2 12.0 11.6 11.7 13.1 (2) (2) (2) (2) (2) (3) (4) (4) (2) KNSTDB7 KNSTDB6 07KNSTDB9 KNSTDB8 KNSTDB8 KNSTDB8 KNSTDB8 KNSTDB8 KNSTDB8 " " " " " " " 0.20 0.18 0.21 0.15 0.21 0.18 0.16 12.75 12.00 12.48 12.48 13.45 12.52 12.52 " " " " " " " 0.39 0.34 0.39 0.28 0.41 0.33 0.30 17.5 21.5 16.2 20.9 16.4 23.2 12.8 (2) (2) (2) (4) (2) (2) (4) KNSTDB8 KNSTDB6 KNSTDB7 KNSTDB6 KNSTDB7 KNSTDB6 KNSTDB7 " " " " " " " " 0.14 0.15 0.16 0.29 0.21 0.16 0.23 0.12 12.50 12.31 12.93 12.95 12.32 12.15 11.94 7.56 " " " " " " " " 0.22 0.24 0.27 0.46 0.32 0.26 0.27 0.15 21.0 15.8 19.1 17.2 18.8 17.9 16.7 20.0 (4) (4) (3) (4) (3) (4) (4) (4) 07KNSTDB12 07KNSTDB12 07KNSTDB12 07KNSTDB13 07KNSTDB13 07KNSTDB13 07KNSTDB13 07KNSTDB14,15 360 170 100 220 110 90 150 140 180 3.33 1.92 1.90 2.09 2.61 2.51 2.87 1.68 1.00 0.999 0.998 0.999 0.999 0.999 0.999 0.999 0.999 0.998 7.0471 7.0103 7.0804 7.0090 7.0289 7.1630 7.0079 7.0130 7.0027 0.2037 0.1978 0.1974 0.1937 0.1876 0.2014 0.2017 0.1988 0.2008 7.09 6.92 6.33 7.36 7.50 6.76 6.63 7.25 6.70 " " " " " " " " " 290 150 90 220 130 240 140 1.36 1.94 1.35 2.15 1.62 1.78 1.53 0.999 0.998 0.999 0.998 0.995 0.998 0.999 7.0118 7.0158 7.0039 7.0248 7.0228 7.0291 7.0364 0.2037 0.1978 0.1999 0.1962 0.2018 0.1974 0.1993 6.66 6.47 6.62 6.79 7.08 6.77 6.69 0.65 1.30 1.33 1.19 1.86 3.14 1.64 2.19 0.992 0.992 0.992 0.993 0.992 0.992 0.980 0.982 8.0589 7.9369 7.7802 7.7958 8.5528 7.8605 10.7265 10.3647 0.1823 0.1820 0.1825 0.1837 0.1837 0.1826 0.1818 0.1816 8.23 7.99 8.21 8.20 8.55 7.81 10.52 6.37 Elevation (m a.s.l.) Boulder size (L ' W ' H) (cm) #44.29664 #44.29227 #44.30176 #44.30282 #44.30214 #44.30577 #44.3067 #44.30368 #44.30026 169.90778 169.93141 169.87808 169.88131 169.89023 169.89184 169.88222 169.86948 169.86005 579.2 546.8 586.2 582.9 581.0 580 580 591 598 400 320 310 410 280 130 310 290 470 ' ' ' ' ' ' ' ' ' 250 280 280 110 110 120 130 150 230 ' ' ' ' ' ' ' ' ' #44.29694 #44.29279 #44.29291 #44.28507 #44.28508 #44.28456 #44.30005 169.90352 169.90362 169.90602 169.87905 169.87993 169.87991 169.88623 575.7 554.8 550.1 540.0 539.2 539.0 569.9 420 380 170 730 250 680 240 ' ' ' ' ' ' ' 230 220 160 260 210 430 240 ' ' ' ' ' ' ' #44.067938 #44.068170 #44.068120 #44.069717 #44.067300 #44.067294 #44.163124 #44.172895 169.866249 169.865613 169.865572 169.866875 169.865716 169.865681 169.813320 169.813869 721.9 718.8 718.2 694.0 722.3 722.3 644.3 622.4 300 310 250 230 205 205 185 115 ' ' ' ' ' ' ' ' 260 ' 50 260 ' 80 200 ' 90 120 ' 50 80 ' 40 115 ' 50 170 ' 60 170 ' 40 Carrier added (g)a Be/9Be " 1s (10#14)b Shielding correction Longitude (DD) Quartz weight (g) 10 Sample thickness (cm) Latitude (DD) Carrier 9Be concentration is 996 ppm for all samples except for those labeled OH-10, for which the carrier 9Be concentration is 1024 ppm. Boron-corrected 10Be/9Be. Ratios are not corrected for background 10Be detected in procedural blanks. Reported [10Be] values have been corrected for background 10Be detected in procedural blanks. d 9 Beþ3 measured after the accelerator. Reported currents are averaged over all AMS runs for a given sample. The number of AMS runs is given in parentheses. e AMS standards to which respective ratios and concentrations are referenced. Reported 10Be/9Be ratios for KNSTD and 07KNSTD are 3.15 ' 10#12 and 2.85 ' 10#12, respectively (Nishiizumi et al., 2007). ‘B’ refers to corresponding procedural blank listed in Table 1b. b c A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 BE22917 OH-06-27 BE22770 OH-06-31 BE24127 OH-06-84 BE22930 OH-06-85 BE22931 OH-06-86 BE23328 OH-06-90 BE23329 OH-06-92 BE23330 OH-06-95 BE23331 OH-06-96 Ohau V moraines BE22925 OH-06-25 BE22769 OH-06-28 BE22918 OH-06-29 BE22771 OH-06-78 BE22920 OH-06-79 BE22772 OH-06-80 BE22921 OH-06-88 Ohau VI moraines BE29526 OH-10-01 BE29527 OH-10-02 BE29528 OH-10-03 BE29521 OH-10-04 BE29520 OH-10-05 BE29519 OH-10-06 BE29518 OH-10-07 BE29593 OH-10-08 126 Table 1a (continued ) Author's personal copy 127 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Table 1b Procedural blank 10 Be data. CAMS laboratory no. Sample ID Carrier added (g)a 10 BE24237 Blank1 17 Aug 07 (B1) Blank2 17 Aug 07 (B2) Blank 06 Jan 07 B3) Blank2 06 July 07 (B4) Blank1 10 Nov 09 (B5) Blank 04 Aug 06 (B6) Blank 11 Sept 06 (B7) Blank 03 Oct 06 (B8) Blank1 06 July 07 (B9) Blank1 06 Aug 07 (B10) Blank2 06 Aug 07 (B11) Blank1 20 April 10 (B12) Blank1 29 April 10 (B13) Blank1 18 May 10 (B14) Blank2 20 May 10 (B15) 0.2002 0.04 " 0.02 0.2000 BE24242 BE23332 BE24140 BE27919 BE22763 BE22922 BE22932 BE24132 BE24214 BE24227 BE29537 BE29513 BE29594 BE29596 Be/9Be " 1s (10#14)b N10Be " 1s (103 atoms)c Average 9Be current (mA)d AMS stde 5.1 " 2.9 13.8 (3) 07KNSTD 0.12 " 0.03 16.4 " 3.6 13.8 (2) 07KNSTD 0.1987 0.04 " 0.02 5.5 " 2.1 10.4 (3) KNSTD 0.1929 0.06 " 0.03 7.2 " 3.9 17.0 (2) 07KNSTD 0.1789 0.02 " 0.02 2.1 " 2.6 16.7 (4) 07KNSTD 0.1981 0.07 " 0.01 10.2 " 1.8 19.2 (3) KNSTD 0.1988 0.08 " 0.02 6.8 " 2.3 15.5 (2) KNSTD 0.2029 0.06 " 0.02 8.7 " 2.2 14.9 (3) KNSTD 0.2012 0.07 " 0.02 9.8 " 2.6 15.0 (2) 07KNSTD 0.2021 0.07 " 0.03 8.9 " 4.4 17.9 (3) 07KNSTD 0.2015 0.05 " 0.02 6.2 " 2.8 16.1 (3) 07KNSTD 0.1824 0.15 " 0.02 18.7 " 2.9 16.7 (3) 07KNSTD 0.1822 0.17 " 0.02 21.6 " 2.9 17.6 (3) 07KNSTD 0.1816 0.08 " 0.02 10.3 " 2.4 14.1 (3) 07KNSTD 0.2004 0.06 " 0.03 8.9 " 4.1 18.0 (3) 07KNSTD a Carrier 9Be concentration is 996 ppm for all samples except for B13, B14, and B15, for which the carrier 9Be concentration is 1024 ppm. Boron-corrected 10Be/9Be. c Total 10Be contamination (in atoms) determined from each procedural blank. d 9 Beþ3 measured after the accelerator. Reported currents are averaged over all AMS runs for a given sample. The number of AMS runs is given in parentheses. e AMS standards to which respective ratios and concentrations are referenced. Reported 10Be/9Be ratios for KNSTD and 07KNSTD are 3.15 ' 10#12 and 2.85 ' 10#12, respectively (Nishiizumi et al., 2007). b 5.1.2. Ohau II Ten ages from the Ohau II moraine ridges range from 26,350 " 1000 to 33,860 " 1040 yrs (Fig. 11) and form a bimodal distribution with a dominant peak (n ¼ 6) and a younger secondary mode (n ¼ 4) (Fig. 12). The dominant peak is centered at 32,500 yrs ago. The full distribution (no outliers removed) has an arithmetic mean of 30,810 " 2450 yrs (1s) and a median of 31,550 yrs. The distribution yields a c2 ¼ 49.68 which is greater than c2expected of 16.92 at the 95% conﬁdence level (Table 3), indicating that the data set includes multiple age populations (Balco and Schaefer, 2006). As boulder heights do not show a robust correlation to surfaceexposure age (R2 < 0.01; see Section 6.1 below), and because there is no geomorphological evidence of landform degradation, we have no basis for inferring that post-depositional geomorphological processes have inﬂuenced the Ohau II age distribution. Two ages (OH-07-16 and OH-07-18) forming the young, secondary mode of the distribution come from remnants of moraines preserved as small islands in the Ohau III outwash plain, and it is conceivable that these remnants represent a younger moraineforming event. The remaining two ages (OH-06-62 and OH-0726) are not in morphostratigraphic order and are considered untenably young. To derive landform-age estimates we follow Ward and Wilson (1978) and separate the younger mode containing the two morphostratigraphically young ages from older mode. This procedure returned an age of 32,520 " 970 yrs (arithmetic mean of older population " 1s) for the outboard landforms and a tentative age of 27,400 " 1300 yrs ago (arithmetic mean of two morphostratigraphically consistent samples of the younger population " 1s) for inboard moraine remnants. 5.1.3. Ohau III Twenty-six ages from the Ohau III moraine belt exhibit a high level of internal consistency (Figs. 11 and 12). The ages of two samples (OH-06-04 and OH-07-11) are identiﬁed as outliers by Chauvenet’s Criterion and are excluded from further analysis. The resulting outlier-free distribution yields a c2 of 18.95 (Table 3), Author's personal copy 128 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Table 2 10 Be surface-exposure ages (in thousands of cal. yrs before AD1950, labeled ‘ka’, "1s) from the Lake Ohau moraines. Samples marked by a single asterisk (*) next to the sample ID are tentatively attributed to a younger episode of Ohau II moraine formation. Outlying ages are italicized and marked by a double asterisk (**) next to the sample ID. Bold ages (‘Lm’) are those discussed in text. Sample ID Ohau I moraines OH-07-04 OH-07-44 OH-07-45 Ohau II moraines OH-06-60a OH-06-61 OH-06-62** OH-06-63 OH-07-01 OH-07-16* OH-07-18* OH-07-19 OH-07-21 OH-07-26** Ohau III moraines OH-06-04** OH-06-05 OH-06-16 OH-06-18 OH-06-44 OH-06-45 OH-06-46 OH-06-50 OH-06-59 OH-06-72 OH-06-73 OH-07-08 OH-07-10 OH-07-11** OH-07-13 OH-07-14 OH-07-15 OH-07-31 OH-07-33 OH-07-34 OH-07-36 OH-07-37 OH-07-38 OH-07-39 OH-07-40 OH-07-41 Ohau IV moraines OH-06-08** OH-06-10 OH-06-11 OH-06-13 OH-06-14 OH-06-15 OH-06-22 OH-06-23 OH-06-24 OH-06-26 OH-06-27 OH-06-31 OH-06-84 OH-06-85 OH-06-86 OH-06-90 OH-06-92 OH-06-95 OH-06-96 Ohau V moraines OH-06-25 OH-06-28 OH-06-29 OH-06-78 OH-06-79** OH-06-80 OH-06-88 Ohau VI moraines OH-10-01 OH-10-02 St age (ka) De age (ka) Du age (ka) Li age (ka) Lm age (ka) 131.3 " 3.2 153.2 " 3.1 141.4 " 4.8 130.1 " 3.2 151.6 " 3.1 139.9 " 4.8 130.7 " 3.2 152.3 " 3.1 140.6 " 4.8 129.7 " 3.2 151.1 " 3.1 139.4 " 4.8 128.3 " 3.2 149.5 " 3.1 138.0 " 4.7 34.30 33.26 29.64 33.66 32.60 28.68 26.68 31.62 32.40 29.58 " " " " " " " " " " 1.05 0.98 0.97 1.30 0.94 1.17 1.02 1.17 1.16 0.99 34.35 33.28 29.61 33.70 32.60 28.68 26.71 31.62 32.41 29.45 " " " " " " " " " " 1.05 0.98 0.97 1.30 0.94 1.17 1.02 1.17 1.16 0.99 34.47 33.42 29.77 33.83 32.75 28.82 26.85 31.79 32.57 29.59 " " " " " " " " " " 1.05 0.99 0.97 1.31 0.95 1.18 1.02 1.17 1.17 0.99 34.40 33.33 29.65 33.76 32.66 28.72 26.76 31.68 32.47 29.45 " " " " " " " " " " 1.05 0.99 0.97 1.31 0.94 1.17 1.02 1.17 1.17 0.99 33.86 32.80 29.22 33.21 32.13 28.28 26.35 31.17 31.94 29.11 " " " " " " " " " " 1.04 0.97 0.95 1.28 0.98 1.15 1.00 1.15 1.15 0.98 17.42 21.96 22.48 22.25 22.93 23.98 22.74 23.63 22.31 21.89 22.25 23.71 22.80 19.94 21.98 23.14 22.84 22.26 23.74 23.46 23.71 22.44 22.13 21.60 22.26 22.58 " " " " " " " " " " " " " " " " " " " " " " " " " " 0.48 0.60 0.56 0.66 0.91 0.68 0.60 0.75 0.68 0.74 0.86 0.92 0.64 0.69 0.73 0.68 0.82 0.79 0.77 0.75 0.83 0.63 0.67 0.73 0.76 0.73 17.63 22.07 22.58 22.36 23.03 24.06 22.85 23.72 22.41 22.01 22.35 23.79 22.89 20.12 22.10 23.24 22.94 22.35 23.79 23.52 23.77 22.53 22.23 21.72 22.35 22.66 " " " " " " " " " " " " " " " " " " " " " " " " " " 0.49 0.61 0.56 0.66 0.91 0.69 0.61 0.75 0.69 0.74 0.87 0.92 0.64 0.69 0.73 0.69 0.83 0.79 0.78 0.75 0.83 0.63 0.67 0.73 0.77 0.74 17.66 22.14 22.66 22.43 23.11 24.16 22.94 23.82 22.49 22.08 22.43 23.89 22.98 20.17 22.18 23.33 23.03 22.42 23.89 23.61 23.86 22.61 22.31 21.78 22.43 22.74 " " " " " " " " " " " " " " " " " " " " " " " " " " 0.49 0.61 0.57 0.66 0.92 0.69 0.61 0.75 0.69 0.75 0.87 0.92 0.65 0.70 0.74 0.69 0.83 0.79 0.78 0.76 0.83 0.63 0.67 0.74 0.77 0.74 17.71 22.13 22.64 22.42 23.08 24.11 22.91 23.78 22.47 22.06 22.41 23.85 22.94 20.19 22.17 23.30 23.00 22.40 23.84 23.57 23.82 22.58 22.29 21.78 22.41 22.72 " " " " " " " " " " " " " " " " " " " " " " " " " " 0.49 0.61 0.57 0.66 0.91 0.69 0.61 0.75 0.69 0.74 0.87 0.92 0.64 0.70 0.74 0.69 0.83 0.79 0.78 0.76 0.83 0.63 0.67 0.74 0.77 0.74 17.39 21.78 22.28 22.06 22.72 23.73 22.54 23.40 22.12 21.72 22.06 23.48 22.59 19.84 21.81 22.93 22.63 22.06 23.49 23.22 23.47 22.25 21.94 21.44 22.07 22.38 " " " " " " " " " " " " " " " " " " " " " " " " " " 0.48 0.60 0.56 0.65 0.90 0.68 0.60 0.74 0.68 0.73 0.86 0.91 0.63 0.68 0.72 0.68 0.82 0.78 0.77 0.74 0.82 0.62 0.66 0.72 0.76 0.73 19.95 18.16 18.20 18.47 17.47 18.49 18.44 18.54 18.78 18.52 19.09 18.45 17.80 18.67 18.48 17.53 17.65 18.72 17.30 " " " " " " " " " " " " " " " " " " " 0.60 0.57 0.61 0.49 0.48 0.50 0.58 0.58 0.61 0.55 0.58 0.53 0.55 0.55 0.55 0.56 0.54 0.55 0.65 20.15 18.38 18.42 18.68 17.70 18.71 18.65 18.74 18.98 18.73 19.29 18.67 18.01 18.87 18.69 17.75 17.86 18.91 17.51 " " " " " " " " " " " " " " " " " " " 0.61 0.58 0.62 0.50 0.49 0.51 0.59 0.58 0.62 0.55 0.59 0.54 0.56 0.56 0.55 0.56 0.55 0.56 0.66 20.20 18.42 18.46 18.72 17.73 18.75 18.69 18.78 19.02 18.77 19.33 18.72 18.05 18.91 18.73 17.78 17.90 18.95 17.54 " " " " " " " " " " " " " " " " " " " 0.61 0.58 0.62 0.50 0.49 0.51 0.59 0.59 0.62 0.56 0.59 0.54 0.56 0.56 0.55 0.56 0.55 0.56 0.66 20.23 18.46 18.50 18.77 17.78 18.80 18.73 18.82 19.06 18.81 19.37 18.76 18.09 18.96 18.77 17.83 17.94 18.99 17.59 " " " " " " " " " " " " " " " " " " " 0.61 0.58 0.62 0.50 0.50 0.51 0.59 0.59 0.62 0.56 0.59 0.54 0.56 0.56 0.56 0.56 0.55 0.56 0.66 19.86 18.12 18.15 18.41 17.44 18.44 18.39 18.48 18.72 18.47 19.02 18.40 17.76 18.61 18.43 17.50 17.61 18.65 17.27 " " " " " " " " " " " " " " " " " " " 0.60 0.57 0.61 0.49 0.49 0.50 0.58 0.58 0.61 0.55 0.58 0.53 0.55 0.55 0.55 0.55 0.54 0.55 0.65 17.77 17.11 17.77 18.05 19.47 18.08 17.55 " " " " " " " 0.54 0.48 0.56 0.41 0.60 0.48 0.43 17.99 17.34 18.00 18.28 19.69 18.31 17.77 " " " " " " " 0.55 0.49 0.57 0.42 0.60 0.49 0.44 18.02 17.37 18.04 18.32 19.74 18.35 17.81 " " " " " " " 0.55 0.49 0.57 0.42 0.60 0.49 0.44 18.07 17.43 18.09 18.37 19.78 18.40 17.86 " " " " " " " 0.55 0.49 0.57 0.42 0.60 0.49 0.44 17.74 17.09 17.74 18.01 19.40 18.04 17.52 " " " " " " " 0.54 0.48 0.56 0.41 0.59 0.48 0.43 17.17 " 0.30 17.00 " 0.34 17.30 " 0.31 17.14 " 0.34 17.30 " 0.31 17.14 " 0.34 17.36 " 0.31 17.21 " 0.34 17.11 " 0.30 16.95 " 0.34 Author's personal copy 129 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Table 2 (continued ) Sample ID St age (ka) De age (ka) Du age (ka) Li age (ka) Lm age (ka) OH-10-03 OH-10-04 OH-10-05 OH-10-06 OH-10-07 OH-10-08** 17.90 18.25 17.05 16.96 17.77 11.44 18.03 18.38 17.19 17.10 17.94 11.64 18.03 18.39 17.18 17.09 17.96 11.66 18.09 18.45 17.25 17.16 18.02 11.71 17.83 18.17 17.00 16.91 17.72 11.49 " " " " " " 0.37 0.66 0.44 0.37 0.40 0.23 " " " " " " 0.37 0.67 0.45 0.37 0.41 0.24 which is less than a c2 of 35.17 expected for a normal population at the 95% conﬁdence level. Because the sample population is normally distributed, and because all measures of central tendency show close agreement, we use the arithmetic mean " 1s landform age of 22,510 " 660 yrs for the prominent outer Ohau III moraines (Figs. 11 and 12). Because only the outer parts of this moraine belt are bouldery, we do not know whether this age is also representative of the middle to inner, presumed recessional, parts of the moraine belt. 5.1.4. Ohau IV Nineteen ages from the outer ridges of the Ohau IV moraine belt (Fig. 11) are tightly distributed and form an approximately normal population (Fig. 12). The distribution yields a c2 of 20.48, which is " " " " " " 0.37 0.67 0.44 0.37 0.41 0.24 " " " " " " 0.38 0.67 0.45 0.38 0.41 0.24 " " " " " " 0.37 0.66 0.44 0.37 0.40 0.23 less than the expected value of 28.87 for a normal distribution of this size (Table 3). All measures of central tendency agree (Table 3). We identiﬁed one age (OH-06-08) as an outlier using Chauvenet’s Criterion. After the outlier is removed, we use the arithmetic mean " 1s to assign a landform age of 18,220 " 500 yrs. 5.1.5. Ohau V Seven ages from the Ohau V moraine ridges form an approximately normal distribution with one older age (OH-06-79; 19,400 " 590 yrs) forming a shoulder on the older tail of the summed probability curve (Fig. 12). The c2 determined for this data set is 10.46, which is less than the c2expected value of 12.59 for a distribution of this size, indicating that all sample variability can be explained by analytical uncertainty alone (Table 3). The age of Fig. 11. Glacial geomorphologic map of Lake Ohau area. Panel A. Southern sector of ﬁeld area. Panel B. Northern sector of ﬁeld area. 10Be surface-exposure ages are in white boxes. Lines indicate sample locations. Ages (black text) are in ka ago. Italics denote ages that are statistical outliers (see text). Red text gives the last four digits of sample ID (Tables 1 and 2). Refer to Fig. 2 for geomorphologic legend and location context within the study area. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.) Author's personal copy 130 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 mean " 1s age of 17,380 " 510 yrs for the Ohau VI recessional morainal landforms. 5.1.7. Chronology of Irishman Stream landforms Kaplan et al. (2010) reported 37 10Be surface-exposure ages from the Irishman Stream moraines. All these ages were calculated in the same fashion as those reported in this study. Kaplan et al. (2010) assigned an age of 13,000 " 500 yrs ago (arithmetic mean " 1s) to the outermost Irishman Stream moraine ridge (Figs. 2 and 8). Younger moraines inboard were formed at 12,200 " 400 yrs ago and 11,500 " 400 yrs ago. 5.2. Glaciological modeling results Fig. 11. (continued). 19,400 " 590 yrs that produces the shoulder of the distribution is identiﬁed as an outlier by Chauvenet’s Criterion, and is therefore excluded from further analysis. After this outlier is removed, we assign an arithmetic mean " 1s age of 17,690 " 350 yrs for the Ohau V moraine ridges. 5.1.6. Ohau VI Eight ages from Ohau VI recessional moraines and erratics on ice-molded bedrock form an approximately normal distribution, with one signiﬁcantly younger age (OH-10-08; 11,490 " 230 yrs) identiﬁed as an outlier by Chauvenet’s Criterion (Fig. 12). The full age distribution yields a c2experimental of 517.2, which is much larger than the c2expected value of 14.07, implying that the data set has been inﬂuenced by geomorphologic factors (Table 3). The boulder from which sample OH-10-08 was collected is near a talus slope at the base of a cliff face (Fig. 11), and it is possible that this boulder was emplaced by rock-fall 11,490 yrs ago, rather than by deposition from the receding glacier. Once this outlier age is removed the c2experimental for this data set is 8.48, which is less than the c2expected value of 12.59 for a data set of this size (Table 3). Based on this reﬁned data set we assign an arithmetic We conducted a series of 4000-year glaciological simulations to determine a probable range of DT for the LLGM Ohau glacier under physically plausible ice-ﬂow regimes. In the ﬁrst case we used modern precipitation values. Fig. 13 shows ice area, volume, and thickness results for optimum solutions R1, R2, and R3 with the sliding law constant set to 0.02, 0.03, and 0.04, respectively. In each of these simulations, the Ohau glacier grew to ﬁt lateral and terminal moraine targets (Fig. 14). We obtained values of DT ¼ #5.6 ! C in the case of relatively thin, fast ﬂowing ice (R1), #6.5 ! C for relatively thick, slow-moving ice (R3), and 6.2 ! C for the middle condition (R2). Modern snowline averaged across the Ohau catchment in the model lies at 2458 m, and at 1478, 1525, and 1598 m for simulations R1, R2, and R3, respectively. Respective snowline depression values are 980, 933, and 860 m, with a midpoint of 920 " 50 m. The latter estimate is consistent with the 875 " 50 m snowline depression estimated by Porter (1975) for the outer LLGM belt “Mount John” moraines of the nearby Lake Pukaki glacier trough. In a second set of simulations, we used the boundary conditions of R2, but reduced precipitation by 30%. From this we found that ice could meet LLGM lateral and terminal moraine targets only when DT equaled #6.9 ! C. Considering the full range of solutions reported above, we take the mid-point DT value of #6.25 " 0.5 ! C as a reasonable estimate of atmospheric temperature, relative to modern conditions, in the Ohau catchment during the LLGM. We note that our estimate of DT for the Ohau glacier is consistent with the #6 to #6.5 ! C estimate reported by Golledge et al. (2012) from a recent model of the whole Southern Alps iceﬁeld. 6. Geomorphological and glaciological implications of chronology 10 Be 6.1. Post-depositional geomorphologic stability of Ohau landforms Post-depositional moraine degradation can lead to boulder exhumation, which in turn can complicate interpretations of glacier history drawn from 10Be surface-exposure ages. Within the Ohau glacier landform assemblage, 10Be surface-exposure ages determined from boulders from the well-expressed moraine belts Ohau III, IV, and V show little variability beyond that which can be explained by analytical uncertainty alone (Table 3; Figs. 12 and 15). Although sampled boulders range from 50 cm to more than 400 cm tall, surface-exposure ages show no statistical correlation to boulder height (Fig. 15). Combined with observations of overall geomorphological stability of the Ohau landforms, we conclude that moraine degradation by diffusive processes (cf. Putkonen and Swanson, 2003; Applegate et al., 2012) has been negligible over at least the last w30,000 yrs, and possibly since w150,000 yrs ago. We therefore consider that landform ages derived here and discussed below can be conﬁdently attributed to the timing of ﬂuctuations of the Ohau glacier margin. Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 131 Fig. 12. Probability density functions (i.e., ‘camelplots’) for the six Ohau morainal landform assemblages. Center black line is arithmetic mean, while vertical black, red, and green lines are, respectively, 1s, 2s, 3s uncertainty thresholds. Thin black curves are Gaussian curves representing individual samples. Dotted black line represents summed probability distribution for data set, including outliers. Thick black curve is summed probability distribution with outliers excluded. Statistics are inset. Note that the younger mode of the Ohau II distribution may reﬂect, at least in part, a later episode of moraine formation. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.) 6.2. Chronology of LLGM ice-margin ﬂuctuations at Lake Ohau The 10Be chronology places the maximum LLGM extent of the Ohau glacier at 32,520 " 970 yrs ago, and affords inboard mean moraine dates of 27,400 " 1300 yrs ago, 22,510 " 660 yrs ago, 18,220 " 500 yrs ago, and 17,690 " 350 yrs ago. The w22,500-yr Ohau III moraine belt served as a target for glaciological modeling, which returned a snowline depression of 920 " 50 m, and temperature lowering of 6.25 " 0.5 ! C, relative to modern values. The prominent geomorphological break separating Ohau III and IV implies that an interstade preceded construction of the Ohau IV moraines at 18,220 " 500 yrs ago. Shortly after 18,000 yrs ago, following construction of the Ohau V moraines, the Ohau glacier evacuated the Lake Ohau basin. During this retreat, the Ohau glacier Author's personal copy 132 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Table 3 Summary statistics for moraine 10Be surface-exposure age distributions. Ages are in cal. yrs before AD1950. Data set Ohau I (all) Ohau II (all) Ohau II (no outliers) Ohau III (all) Ohau III (no outliers) Ohau IV (all) Ohau IV (no outliers) Ohau V (all) Ohau V (no outliers) Ohau VI (all) Ohau VI (no outliers) N Nremoved Mean age (yrs) "1s (yrs) Error-weighted mean (yrs) Error-weighted uncertainty (yrs) Standard error of the mean (yrs) Peak age (yrs) Median age (yrs) c2experimental c2expected c2 test ðc2experimental ! < c2expected (95%) 3 10 6 e e 4 138,590 30,810 32,520 10,600 2450 970 139,010 30,750 32,540 1990 330 440 7490 770 400 128,720 32,470 32,490 137,980 31,550 32,460 23.42 49.68 3.86 5.99 16.92 11.07 Fail Fail Pass 26 24 e 2 22,210 22,510 1280 660 21,960 22,460 140 140 250 140 22,310 22,310 22,260 22,330 134.8 18.95 37.65 35.17 Fail Pass 19 18 e 1 18,300 18,220 610 500 18,290 18,210 130 130 140 120 18,350 18,350 18,410 18,410 20.48 13.17 28.87 27.59 Pass Pass 7 6 e 1 17,930 17,690 720 350 17,860 17,700 190 190 270 140 17,770 17,760 17,740 17,740 10.46 2.87 12.59 11.07 Pass Pass 8 7 e 1 16,650 17,380 2140 510 15,670 17,280 120 140 760 190 17,070 17,070 17,050 17,110 517.2 8.48 14.07 12.59 Fail Pass produced the Ohau VI recessional moraines on the valley sides and exposed the glacially molded bedrock knob near the junction of the Hopkins and Dobson rivers. If it is assumed that the glacier terminus was located near the Hopkins/Dobson river conﬂuence when the bedrock bench became exposed, this would imply a terminus retreat of w24 km and hence a w40% reduction in length of the Ohau glacier tongue over the approximately 300 yrs that had elapsed between construction of the Ohau V and VI landforms. Over the same time interval, the Ohau glacier surface lowered at least w200 m to expose the Hopkins/Dobson River bedrock bench. Therefore, a mean net terminus retreat rate of w77 m yr#1 and ice-surface lowering rate of 0.7 m yr#1 occurred between w17,700 and w17,400 yrs ago. Consideration of the longest possible duration of ice retreat, based on Ohau V and Ohau VI moraine age error limits, affords minimum net terminus retreat and ice-surface lowering rates of w20 m yr#1 and 0.18 m yr#1, respectively. The Ohau glacier tributaries retracted well back into the high mountain valleys before a subsequent Lateglacial ice resurgence, documented by Kaplan et al. (2010) as culminating at 13,000 " 500 yrs ago. Doughty et al. (2012) targeted the Irishman Stream Lateglacial moraines for glaciological modeling and determined that a snowline w400 m lower than modern values and a DT of #2.7 " 0.6 ! C was necessary to grow the modeled glacier to the outer limit of the Lateglacial moraine belt. Thus, we consider from these results that the bulk of recession from LLGM moraine belts toward Lateglacial limits, indicating snowline rise and atmospheric warming of at least w520 m and w3.6 ! C, respectively, took place within the ﬁrst few millennia following the LLGM in the Lake Ohau drainage system of the Southern Alps of New Zealand. We note that evidence, if any, for the innermost limit of ice recession during the ﬁrst pulse of deglaciation was masked by glacier readvance in Irishman Stream valley during the ACR. Thus, because the full distance of glacier recession prior to Table 4 Mean, maximum, and minimum landform ages based on systematic production-rate uncertainty limits. Production rate limits (PRmax and PRmin) are based on ‘Lm’ scaling and the production-rate uncertainty of "2.1% (from Putnam et al., 2010b). Data set Basis for moraine age Mean moraine age (PRmean) (yrs) Min. moraine age (PRmax) (yrs) Max. moraine age (PRmin) (yrs) Ohau I (all) Chose: Mean " 1s; Basis: Scattered small data set; considered general minimum age of moraine belt. Chose: Mean " 1s after young mode (n ¼ 4) removed. Basis: bimodal distribution. Chose: Mean " 1s after old mode (n ¼ 6) and two morphostratigraphic outliers removed. Basis: bimodal distribution. Chose: Mean " 1s after two outliers removed. Basis: Approx. normal distribution. Chose: Mean " 1s after one outlier removed. Basis: Approx. normal distribution. Chose: Mean " 1s after one outlier removed. Basis: Approx. normal distribution. Chose: Mean " 1s after one outlier removed. Basis: Approx. normal distribution. 138,600 " 10,600 )135,680 " 10,600 &141,500 " 10,820 Ohau II (old mode) Ohau II (young mode; outliers removed) Ohau III (outliers removed) Ohau IV (outliers removed) Ohau V (outliers removed) Ohau VI (outliers removed) 32,520 " 970 )31,840 " 950 &33,200 " 990 27,400 " 1300 )26,820 " 1270 &27,980 " 1330 22,510 " 660 )22,030 " 650 &22,980 " 680 18,220 " 500 )17,830 " 480 &18,600 " 510 17,690 " 350 )17,320 " 350 &18,060 " 360 17,380 " 510 )17,020 " 500 &17,750 " 520 Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 133 Augustinus et al., 2011), span the past w30,000 yrs and complement the Ohau glacier record. These records show a distinct rise in the percentage of grass pollen, at the expense of shrub and tree pollen, beginning w29,000 yrs ago. Over the following w11,000 yrs the percentage of grass pollen varied, suggesting ﬂuctuations between colder and milder conditions. A sustained shift in the pollen spectra began between w19,000 and w18,000 yrs ago, with progressive decline in grass pollen in favor of shrub and tree pollen, which became dominant at these lowland sites by w17,000 yrs ago. Increasing dominance of shrub and tree pollen w18,000 to w17,000 yrs ago was commensurate with the onset of glacier recession from the Ohau and Pukaki LLGM moraine belts. 7.2. Comparison with records elsewhere in the Southern Hemisphere Fig. 13. A) Ice area, B) volume, and C) thickness results for three different ice-ﬂow regimes deﬁned by the sliding law constant in the Ohau glacier model domain: relatively thin, fast moving ice (R1, B ¼ 0.02, dark gray); moderate ice (R2, B ¼ 0.03, light gray); and thick, slow-moving ice (R3, B ¼ 0.04, black). the Lateglacial resurgence can not be discerned from geomorphology, we consider the magnitude of this snowline/temperature rise to be a minimum value. 7. Discussion 7.1. Comparison to other New Zealand glacier records and paleovegetation signatures 10 Be dates of the youngest terminal moraines near Lake Ohau are consistent with those from the youngest LLGM moraines alongside Lake Pukaki, which lies w20 km northeast of Lake Ohau. Based on detailed 10Be measurements near Boundary Stream on the right-lateral moraines of Lake Pukaki, the onset of ice withdrawal was 18,350 " 390 yrs ago (n ¼ 10, no outliers) (Putnam et al., 2010b). This age is indistinguishable from that of the Ohau IV moraine belt (18,220 " 500 yrs) that occurs alongside Lake Ohau. The 10Be dates therefore suggest that the Pukaki glacier began a major recession about 18,000 yrs ago in concert with the Ohau glacier. Well-documented and well-dated pollen records from lowland sites at Okarito, alongside the northwest ﬂank of the Southern Alps on South Island west coast (Vandergoes et al., 2005; Newnham et al., 2007b), and Auckland, on northern North Island (Sandiford et al., 2003; Newnham et al., 2007a; As a test of whether the Ohau glacier record reﬂects general glacial conditions in the middle latitudes of the Southern Hemisphere, we compare the Ohau chronology with chronologies for LLGM glacier advances in mid-latitude South America (Figs. 16 and 17). In the Chilean Lake District (CLD; 39! e43! S), 14 C ages assayed from plant macrofossils document times at which vegetated land surfaces were overrun by Andean piedmont glaciers and buried intact by glacial deposits (Denton et al., 1999b). The Last Glaciation terminal moraine complexes were formed by at least seven episodes of ice advance between w36,000 and w18,000 yrs ago (Fig. 15), and together these events comprise the LLGM of the CLD. Formation of Ohau LLGM moraine belts match closely in time with glacier advances into the CLD LLGM moraine belts (Denton et al., 1999b) (Fig. 17). The age of the Ohau II moraine belt (w32,500ew27,400 yrs) overlaps with the timing of glacier advances into the CLD LLGM moraine belt at 34,060 " 325 yrs ago (Puerto Octay site; n ¼ 5; no outliers), 31,120 " 70 yrs ago (Bahía Frutillar site; n ¼ 17; 8 outliers excluded), and 27,910 " 210 yrs ago (Canal Tenglo site; n ¼ 6; 2 outliers excluded). Note that all ages given here for the CLD are from 14C dates reported in Denton et al. (1999b), converted to calendar ages using the IntCal09 calibration curve (Reimer et al., 2009). The Last Glacial advance of the Corcovado ice lobe into the outer moraine belt on Isla Grande de Chiloe peaked at 18,035 " 215 yrs ago (Delcahue site; n ¼ 34; 6 outliers excluded) (Denton et al., 1999b), coinciding with construction of the Ohau IV and V moraine belts at 18,220 " 500 and 17,690 " 350 yrs ago, respectively. Similar ages of w18,000 yrs were obtained for the Last Glacial advance into the outer moraine belts of the Ancud, Reloncavi, and Llanquihue glacier lobes of the former Patagonian Ice Sheet in the CLD (Denton et al., 1999b). The onset of deglaciation in the CLD (Denton et al., 1999b) began at approximately the same time as glacier recession at Lakes Ohau and Pukaki in the Southern Alps. In the CLD region, ice had receded to within w10 km of modern glacier margins by 14,550 " 620 yrs ago (Heusser, 1990). Farther south, on the east side of the Southern Andes at Lago Argentino (w50! S), a readvance of a Southern Patagonian Iceﬁeld outlet glacier culminated at w12,990 " 80 yrs ago (Strelin et al., 2011). Retreat occurred immediately thereafter, with ice receding to Holocene limits by w12,000 yrs ago (Kaplan et al., 2011; Strelin et al., 2011). This Lateglacial signature of Patagonian glacier ﬂuctuations mirrors that of the Irishman Stream tributary glacier in the Lake Ohau catchment (Kaplan et al., 2010) and those observed elsewhere in the central Southern Alps (Putnam et al., 2010a). Paleoceanographic records (Fig. 16) also exhibit patterns similar to those deﬁned from the Ohau moraine chronology. Alkenone- Author's personal copy 134 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Fig. 14. Visualizations of simulation R2 at 4000 model years: A) ice thickness map, B) mass balance map, and C) shaded polygon surface rendered in Google Earth". Model domain is fully speciﬁed for the Lake Ohau catchment only, and does not yield realistic results for neighboring catchments that are not fully included in the model domain. Elevation contours on the surface of the Ohau glacier are at 100 m intervals. and Mg/Ca-inferred SSTs from cores off southern Australia (37! S; MD03-2611; Calvo et al., 2007) and Chile (41! S; ODP-1233; Lamy et al., 2004; Kaiser et al., 2005), as well as faunal-based SST reconstructions from the Indian Ocean (46! S; MD88-770; Labeyrie et al., 1996; Barrows et al., 2007), southeast Atlantic Ocean (41! S; TN057-21; Barker et al., 2009), and just west of South Island (42! S; SO136-GC3; Pelejero et al., 2006; Barrows et al., 2007) all record the onset of decline towards maximum glacial cold by w30,000 yrs ago. The STF was equatorward of its present-day position during the LGM (Bard and Rickaby, 2009; Sikes et al., 2009; De Deckker et al., 2012). A cold and latitudinally expanded Southern Ocean persisted, with some variability, until w18,000 yrs ago, after which time sea-surface warming commenced ubiquitously across the Southern Ocean (Lamy et al., 2004; Barrows et al., 2007; Calvo et al., 2007; Lamy et al., 2007; Barker et al., 2009; De Deckker et al., 2012). The foraminiferal SST reconstructions from cores TNO57-21 at 41! S in the South Atlantic (Barker et al., 2009) and MD03-2611 at 37! S south of Australia (De Deckker et al., 2012) exhibit sharply expressed cold periods marking northward excursions of the STF that coincide with the construction of the Ohau III and IV moraine belts at w22,500 and w18,200 yrs ago, respectively (Fig. 17). Sea surface warming and poleward migration of the STF commenced during the Ohau IVeVI recession phase (just after w18,000 yrs ago), and which was interrupted by a SST/STF reversal that culminated coevally with construction of the Irishman Stream moraines at w13,000 yrs ago. A ﬁnal southern mid-latitude SST warming phase detected in cores TNO57-21 and MD03-2611 coincided with recession of the Irishman Stream glacier from Lateglacial to Holocene extents (w13,000e11,700 yrs ago; Kaplan et al., 2010). The TNO57-21 and MD03-2611 faunal records also capture a prominent interstade, and southward movement of the STF, from w21,000 to w19,000 yrs ago that corresponds to the prominent geomorphological break between Ohau III and IV (i.e., after w22,500 and before w18,000 yrs ago), as well as the ‘Varas Interstade’ at w21,000 to w19,000 yrs ago in the Chilean Lake District (Mercer, 1972; Denton et al., 1999b). The observed close correspondence among millennial-scale events detected in SSTs and faunal assemblages south of Australia (Calvo et al., 2007; De Deckker et al., 2012), Atlantic SSTs (Barker et al., 2009), the Ohau glacier record presented here from the southwest Paciﬁc region, and the CLD glacier record from the southeast Paciﬁc, implies that all regions monitored a common Southern Hemisphere mid-latitude climate signal during the LLGM and Last Glacial termination. Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 135 Fig. 15. Plots comparing boulder height versus ages normalized to mean moraine age for the six Ohau morainal landform assemblages. Bottom panel shows all data plotted together. Solid black lines show trend, with respective regression correlation coefﬁcients plotted nearby (R2 values). Dotted black line (and respective R2 value) in Ohau VI panel gives resulting trend with one exceptionally young outlier age excluded from data set. Note that all data sets show no signiﬁcant correlation between boulder height and boulder age. Author's personal copy 136 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Fig. 16. Polar projection map of the Southern Hemisphere with locations of records mentioned in text. Arrows indicate ocean currents. Red dots are terrestrial sites discussed in text, blue dots give key marine sediment core locations, and white dots are locations of pertinent ice cores. Dark blue region illustrates sub-antarctic zone. Schematic physical oceanography adapted from Brown et al. (2001). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.) Furthermore, some aspects of the Ohau glacier record exhibit similarities to, as well as important differences from, Antarctic icecore isotope-inferred temperature records (w70e80! S; Petit et al., 1999; Blunier and Brook, 2001; EPICA Community Members, 2006; Kawamura et al., 2007; Lemieux-Dudon et al., 2010; see Fig. 16 for core locations). For example, the EPICA Dome C deuterium signature, placed on the timescale of Lemieux-Dudon et al. (2010), registered minimum ice-age temperatures over East Antarctica w34,700ew17,900 yrs ago (Fig. 17), consistent with the duration of the Ohau LLGM. Antarctic warming commenced at w17,900 yrs ago (Lemieux-Dudon et al., 2010), at the beginning of the last termination (Denton et al., 2010), in concert with the onset of deglaciation at Lakes Ohau and Pukaki. Resurgence of Southern Alps glaciers at 13,000 yrs ago fell within the Antarctic Cold Reversal (ACR) deﬁned from Antarctic ice cores (Kaplan et al., 2010; Putnam et al., 2010a). Advances into the Ohau LLGM moraine belt also occurred during a period of persistently low atmospheric CO2 concentrations w30,000ew18,000 yrs ago (Fig. 17; Monnin et al., 2001; Ahn and Brook, 2008). However, most Antarctic isotopeinferred temperature signatures do not register the strong millennial-scale variability during the LLGM evident in the Ohau glacier record, in the CLD glacier record (Denton et al., 1999a, 1999b), and in high-resolution SST reconstructions from southern middle latitudes (e.g., Barker et al., 2009). One exception is the d18O-inferred atmospheric temperature signature from the EPICA Dronning Maud Land core (EPICA Community Members, 2006), which exhibits warming events that coincide with interstadial intervals indicated by the Ohau glacier record, the CLD glacier record, and southern mid-latitude SST patterns (Fig. 17). Overall, southern mid-latitude SSTs records, as well as highlatitude ice-core signatures of atmospheric temperatures and CO2, are concordant with the duration of sustained cold atmospheric temperatures as recorded by glaciers in the Southern Alps and southern South America (Fig. 17). Millennial-scale climate ﬂuctuations were registered most strongly among mid-latitude records, with weak or absent counterparts in Antarctic ice-core isotopeinferred temperature records. Overall, most Southern Hemisphere paleoclimate data suggest that the ocean and atmosphere registered a common climate signal throughout the southern quarter of the globe during the LGM and beginning of the termination. Therefore, any proposed drivers of Southern Hemisphere climate must account for the long duration and the coeval termination of southern LGM conditions in both the ocean and the atmosphere, from 37! S to 77! S. 7.3. The LGM in the Southern Hemisphere In seeking an explanation for Mercer’s (1984) ‘ﬂy in the ointment’ conundrum of an apparently synchronous LGM in the Northern and Southern Hemispheres, we look towards the Southern Ocean as a likely causative agent. The Southern Ocean has a propensity to stratify under cold atmospheric conditions (Gordon, 1991; Sigman et al., 2004), such as may arise from long-duration southern winters (Huybers and Denton, 2008; Keeling and Visbeck, 2011). Southern Ocean buoyancy-driven stratiﬁcation results, at near-freezing water temperatures, from the density of seawater being more sensitive to changes in salinity than to changes in temperature (Sigman et al., 2004, 2010). Wintertime Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 137 cooling of surface water initiates sea-ice formation. Brine is rejected during winter sea-ice formation over the Antarctic continental shelf (Bouttes et al., 2010, 2011). Subsequent wind-driven export of sea ice into the open ocean promotes surface-water freshening and open-ocean stratiﬁcation (Sigman and Haug, 2003). Thus winter cooling and sea-ice expansion play essential roles in spreading stratiﬁcation into the Southern Ocean. Stratiﬁcation cools the surface of the Southern Ocean by inhibiting upwelling of relatively warm deep waters (Gordon, 1981), enhancing the spread of winter sea ice. Any associated northward movement of the southern westerly wind belt would further reduce upwelling (Toggweiler et al., 2006; Toggweiler, 2008, 2009). Collectively, marine proxies indicate that the Southern Ocean entered a fully stratiﬁed state as early as 70,000 years ago (Robinson et al., 2007; Anderson et al., 2009), consistent with increasing winter duration at that time. Opal records from the Antarctic Polar Front (Anderson et al., 2009), as well as records of % C. davisiana from the subantarctic zone (Hays et al., 1976), indicate that the Southern Ocean then remained in a generally stratiﬁed state, despite millennial-scale disturbances of stratiﬁcation between w60,000 and w36,000 yrs ago corresponding to Heinrich Stadials in the North Atlantic (Anderson et al., 2009). These disturbances were evidently unable to completely destratify the ocean and initiate deep convection (Sigman et al., 2007). The Southern Ocean entered a state of persistent full stratiﬁcation w36,000 yrs ago (Hays et al., 1976; Robinson et al., 2007; Anderson et al., 2009; Sigman et al., 2010; Burke and Robinson, 2012), which along with expanded winter sea ice around Antarctica (Gersonde et al., 2003, 2005; Allen et al., 2011), continued through the LGM. We suggest that stratiﬁcation in the Southern Ocean, modulated by the effects of Earth’s orbit on high-latitude winter duration (Huybers and Denton, 2008), resulted in cold temperatures that affected adjacent regions. Thus, Southern Ocean stratiﬁcation provides an explanation for the formation of Ohau glacier LLGM moraines commencing as early as w32,500 yrs ago, and continuing through to w18,200 yrs ago. Individual ice advance maxima of the Ohau glacier during this interval may reﬂect millennial-scale ﬂuctuations within generally cold climatic conditions. A difﬁculty with the Southern Ocean coupled sea-ice/ stratiﬁcation model is that, taken alone, the effects of winter duration do not explain why, at w32,500 yrs ago, when southern high-latitude winter duration was relatively short, the Southern Ocean was stratiﬁed and the Ohau glacier was just as extensive as it was w22,500 yrs ago, when winter duration was relatively long. Although we consider southern winter duration to be an underlying factor governing stratiﬁcation and SSTs in the Southern Ocean, and hence southern mid-latitude glaciers, we note also that Southern Alps glaciers would also have been strongly inﬂuenced by Fig. 17. Plot comparing records mentioned in text. a.) 45! N (blue) and 45! S (red) insolation intensity (Berger and Loutre, 1991). b.) Ohau glacier-inferred snowline and temperature reconstruction. This curve incorporates the Lateglacial reconstruction from Irishman Stream (Kaplan et al., 2010; Doughty et al., 2012). Solid black lines depict where age control is known. Dotted black lines are based on inference. c.) Patagonian glacier-inferred temperature changes (Denton et al., 1999a; Denton et al., 1999b). Curve has been updated to reﬂect the Lateglacial record of Strelin et al. (2011). d.) Modern analogue technique (MAT) SSTs and e.) G. ruber from core MD032611 south of Australia (De Deckker et al., 2012). The latter is a proxy for tropical water incursions south of Australia, which indicates when the STF is to the south. f.) South Atlantic SSTs (Barker et al., 2009). g.) EPICA Dronning Maud Land d18O (Antarctic temperature proxy) (EPICA Community Members, 2006). Thin gray curve connects raw data. Thick blue line is 10-pt moving average. h.) Atmospheric CO2 records of Monnin et al. (2001), Indermühle et al. (1999, 2000), and Ahn and Brook (2008), synchronized to GICC05 (Greenland methane) timescale (Barker et al., 2009). i.) Southern Ocean stratiﬁcation/upwelling record inferred from biogenic opal ﬂux measured in cores TN057-13PC and TN057-14PC (Anderson et al., 2009). Vertical yellow bands correspond with northern stadials (southern warming events) in Barker et al. (2009) and Bond et al. (1999). HS1, HS2, and HS3 correspond to Heinrich Stadials 1, 2, and 3, respectively. YD is Younger Dryas. ACR is Antarctic Cold Reversal. (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.) Author's personal copy 138 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 the position of the STF upwind of South Island. Millennial-scale migrations of the STF could have overwhelmed the more subtle effects of winter duration at the northern edge of the Southern Ocean between 36,000 and 18,000 yrs ago (De Deckker et al., 2012). In addition, once stratiﬁed, the vertical structure of the Southern Ocean, and its sea ice, could remain robust enough to withstand disturbance by orbital forcing. Thus it seems that a more powerful mechanism would have been required to destratify the Southern Ocean and re-initiate deep ventilation (Sigman et al., 2007). Before exploring mechanisms for disruption of Southern Ocean stratiﬁcation, we summarize the argument thus far. We speculate that stratiﬁcation in the Southern Ocean is driven fundamentally by orbital forcing, as shown by the marine records of Hays et al. (1976), in particular through the effect of high-latitude winter duration on sea-ice formation (Huybers and Denton, 2008). Orbitally lengthened winters would prolong the period during which sea ice is formed and exported to the open ocean by the circum-Antarctic wind ﬁeld, thereby promoting stratiﬁcation in the open Southern Ocean. Once formed, stratiﬁcation appears difﬁcult to disrupt. At about 23,000 yrs ago, northern high-latitude summers were of low insolation intensity and southern lobes of the Laurentide Ice Sheet were at or near maximum extents (Lowell et al., 1999; Curry and Grimley, 2006; Curry and Petras, 2011), while southern highlatitude winters were of long duration (Huybers and Denton, 2008; Huybers, 2009), and the Ohau glacier was at or near its maximum extent. Thus, this mechanism, involving the effects of southern winter duration on Southern Ocean sea ice and stratiﬁcation, resolves Mercer’s (1984) conundrum, and offers an explanation for generally coeval maxima of the Ohau glacier and northern ice sheets during the global LGM, despite opposing summer insolation intensity patterns. This explanation circumvents any need for mysterious teleconnections between the hemispheres in order to account for global synchrony of the LGM. Future development of additional Southern Ocean stratiﬁcation records from marine proxies could help to test the links between southern winter duration and stratiﬁcation on orbital timescales. We then come to the question of what brought about the end of the southern LGM? Retreat of the Ohau glacier occurred in unison with a regionally widespread warming trend, seen for example in SSTs (Barrows et al., 2007; Calvo et al., 2007; Barker et al., 2009), and indicated by disintegration of sea-ice cover around Antarctica (Shemesh et al., 2002; Allen et al., 2011). This southern warming coincided with Heinrich Stadial 1 (HS1) in the North Atlantic region (Bond et al., 1999; Barker et al., 2009; Denton et al., 2010). Thus, at face value, the southern warming beginning w18,000 yrs ago had the hallmarks of an asynchronous millennial-scale climate oscillation between the North Atlantic and the Southern Ocean (Denton et al., 2010), involving an oceanic (Crowley, 1992; Broecker, 1998) and/or atmospheric (Toggweiler et al., 2006; Anderson et al., 2009) bipolar seesaw mechanism between the hemispheres (Sigman et al., 2007). However, on this occasion, a full glacial termination resulted. Taken together, modeling of the Ohau glacier (this study) and the Irishman Stream cirque glacier (Kaplan et al., 2010; Doughty et al., 2012) shows an overall warming between the LLGM and ACR of at least w3.6 ! C. Coeval phenomena included a net poleward migration of the southern STF (Chiessi et al., 2008; Bard and Rickaby, 2009; Sikes et al., 2009; De Deckker et al., 2012), rising atmospheric CO2 (Monnin et al., 2001) and major upwelling south of the Polar Front (Anderson et al., 2009), which indicates breakdown of Antarctic stratiﬁcation. Denton et al. (2010) pointed out that poleward migration of the wind-driven southern STF is difﬁcult to explain solely by an oceanic mechanism, and proposed that a southward shift of the southern westerlies at the onset of HS1 stimulated a poleward shift of the southern STF (Denton et al., 2010; De Deckker et al., 2012). This southward shift of the westerlies (Anderson et al., 2009), perhaps combined with the effects of a bipolar seesaw operating through the interior of the ocean (Broecker, 1998; Sigman et al., 2007), completely disrupted the stratiﬁcation of the glacial Southern Ocean, and thereby triggered the last mountain glacier termination in Southern Hemisphere middle latitudes. In conclusion, we suggest that the ice-age signature as reﬂected in the Ohau LLGM record, arose fundamentally from orbitally driven changes in southern high-latitude winter duration that, in turn, inﬂuenced stratiﬁcation of the Southern Ocean. Previous studies have established relationships between Southern Ocean stratiﬁcation and SSTs (Hays et al., 1976; Anderson et al., 2009). We propose that a prolonged episode of Southern Ocean stratiﬁcation promoted generally cold southern mid-latitude atmospheric temperatures, and hence full-glacial iceﬁelds, during the global LGM. We then suggest that a bipolar seesaw oscillation, perhaps operating through both the atmosphere and the ocean, caused poleward migration of the southern STF and disruption of Southern Ocean stratiﬁcation beginning at 18,000 yrs ago, resulting in seasurface warming that produced the southern mid-latitude Last Glacial termination. 8. Conclusions 1.) A precise and accurate 10Be surface-exposure chronology, comprising 73 new dates on glacial landforms near Lake Ohau, as well as 37 previously reported dates from the Irishman Stream tributary of the Ohau catchment, indicates that the LLGM in the Southern Alps was achieved as early as 32,520 " 970 yrs ago, and that subsequent glacier advances produced slightly inboard moraine belts at 27,400 " 1300 yrs ago, 22,510 " 660 yrs ago and 18,220 " 500 yrs ago. Recession from the terminal moraine belt was underway by 17,690 " 350 yrs ago, and by 17,380 " 510 yrs ago, the Ohau glacier tongue had retreated by as much as w24 km up valley. This retreat represents a w40% decrease of the LLGM glacier length and a net retreat rate of 77 m yr#1. A glacier that fed the former Ohau glacier during the LLGM had receded well into the mountains by 13,000 " 500 yrs ago (Kaplan et al., 2010). 2.) LLGM snowline was 920 " 50 m below the present day value averaged across the Lake Ohau ice catchment. This snowline depression corresponds to a temperature depression of about 6.25 " 0.5 ! C, which is the median value determined from a variety of simulations in which ﬂow parameters and precipitation were allowed to vary. Snowline and temperature rose by at least w520 m and w3.6 ! C, respectively, between the LLGM and ACR. 3.) Glacier variations at Lake Ohau complement marine and terrestrial records of climate change in the southern middle latitudes, as well as temperature and atmospheric CO2 concentrations measured in Antarctic ice cores. We propose that enhanced sea ice and open-ocean stratiﬁcation caused a decrease in Southern Ocean SSTs and an equatorward shift in the subtropical front that initiated the LLGM in the Southern Alps. We think that this Southern Ocean signal could have been paced in part by winter duration, which is sensitive to orbital precession, thereby producing the LLGM at Lake Ohau coevally with the maximum of the Laurentide Ice Sheet. 4.) We argue that a disruption of Southern Ocean stratiﬁcation beginning w18,000 yrs ago triggered the onset of the Last Glacial termination in the Southern Alps. We attribute this breakdown of stratiﬁcation to a poleward shift of the southern westerlies, associated with HS1 cooling in the North Atlantic region (Denton et al., 2010). A southward shift of the southern STF and invigorated Southern Ocean upwelling led to an abrupt Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 onset of sea-ice retreat, SST warming, and CO2 degassing, thereby warming Southern Hemisphere atmospheric temperatures and inducing mountain glacier retreat at Lake Ohau in the Southern Alps of New Zealand. Acknowledgments This work was supported by funding from the Gary C. Comer Science and Education Foundation (CSEF), the Quesada Family Foundation, the National Oceanographic and Atmospheric Administration (NOAA), and by National Science Foundation grants EAR1102782, EAR-0345835 and EAR-0745781. D.M. Sigman, S. Barker, W.S. Broecker, R.F. Anderson, T.J. Hughes, P. Huybers, B.L. Hall, G.R.M. Bromley, M.J. Vandergoes, P. Upton, and R.B. Alley provided helpful discussions. In particular, we are grateful to J.D. Hays for engaging discussions about the role of polar ocean stratiﬁcation in ice-age cycles. N.R. Golledge, A.N. Mackintosh, B. Anderson, and T.J.H. Chinn contributed helpful insights into glaciological modeling in the Southern Alps. S. Barker kindly provided data from core TNO57-21. B. Lemieux-Dudon provided the updated EPICA Dome C chronology. K. Ladig and S. Travis helped with ﬁeldwork. J. Frisch assisted with laboratory work. We are grateful to the Kees family (Ohau Station), F. Hawkins, M. Lindsay, J. Papich, G. Burrows (Ribbonwood Station), S. and H. Williamson (Glenbrook Station), J. and K. Wigley (Glen Lyon Station), and the Department of Conservation ! nanga o Nga !i Tahu, for permitting us e Te Papa Atawhai, Te Ru access to their land. T. and K. Ritchie of Lake Ruataniwha Holiday Park provided excellent accommodation during long ﬁeld seasons. V. Jomelli and an anonymous reviewer provided helpful suggestions that improved the paper. A.E. Putnam was supported by CSEF, NOAA, a University of Maine teaching assistantship, and a Lamonte Doherty Earth Observatory (LDEO) postdoctoral fellowship while conducting this research. GNS Science’s Direct Crown Funded Programme ‘Global Change through Time’ supported D.J.A.B. This is LDEO contribution #7629. Appendix A A complete description of UMISM can be found in Fastook and Prentice (1994) and Fastook et al. (2008). Here we provide an overview of the model ice-ﬂow solver in order to give context for the tuning reported in the methods section. The tuning pertains to mainly the sliding and ice hardness terms below in Equations (2), (3) and (5). The core of UMISM is a differential equation for ice extent and thickness as a function of time derived from an integrated momentum conservation equation based on the ﬂow law of ice (Glen, 1955), coupled with a continuity equation for mass conservation: dh ¼ MB # V$ðUHÞ; dt (1) where dh/dt is the time-dependent ice-surface elevation, MB is the net-annual surface mass balance (accumulation minus ablation), U is the column-average ice velocity, and H is the ice thickness. U in Equation (1) is obtained by expanding: 3_ ¼ h s in A ; (2) which is a tensor equation expressing strain rates, 3_ , and stresses, s, through a non-linear power law. In this equation, n is the empirical ﬂow law constant [we use a value of 3 for temperate glaciers 139 (Paterson, 1994)], and A is a temperature dependent function that represents ice hardness in an Arrhenius relationship: A ¼ EA0 e#Q =RT ; (3) where A0 is a constant, Q is the activation energy for ice creep, R is the gas constant, T is the ice temperature, and E is a tuning (or ﬂow enhancement) parameter meant to account for ice impurities. By shallow-ice approximation, all stresses and strain rates are ignored in the force balance except basal drag. Thus, the stress in Equation (2) acting on the bed, or the “driving” stress, is expressed: (4) sxz ¼ rgHVh; where sxz is a stress in the x-direction acting on a surface with normal in the z direction, r is density of ice, g is gravitational acceleration, H is ice thickness, and Vh is ice-surface slope. Finally, U, is calculated from the sum of velocity due to internal deformation, UD, and velocity due to basal sliding, US, as expressed: U ¼ UD þ US ¼ " " # # rgVh m m 2 rgVh n nþ1 H þ H ; nþ2 A B (5) where the last term follows a general sliding law relationship developed by Weertman (1964) for beds at the melting point, and modiﬁed to incorporate the effect of basal water (Johnson and Fastook, 2002). Here n is the empirical ﬂow law constant in Equation (2), A is the ice hardness parameter described in Equation (3), B is the sliding law constant, and m is the sliding law exponent. References Ahn, J., Brook, E.J., 2008. Atmospheric CO2 and climate on millennial time scales during the last glacial period. Science 322 (5898), 83e85. Allen, C.S., Pike, J., Pudsey, C.J., 2011. Last GlacialeInterglacial sea-ice cover in the SW Atlantic and its potential role in global deglaciation. Quaternary Science Reviews 19e20, 2446e2458. Amos, C.B., Burbank, D.W., Nobes, D.C., Read, S.A.L., 2007. Geomorphic constraints on listric thrust faulting: Implications for active deformation in the Mackenzie Basin, South Island, New Zealand. Journal of Geophysical Research 112. http:// dx.doi.org/10.1029/2006JB004291. Amos, C.B., Burbank, D.W., Read, S.A.L., 2010. Along-strike growth of the Ostler fault, New Zealand: consequences for drainage deﬂection above active thrusts. Tectonics 29, TC4021. Anderson, B., Mackintosh, A., 2006. Temperature change is the major driver of lateglacial and Holocene glacier ﬂuctuations in New Zealand. Geology 34 (2),121e124. Anderson, B., Lawson, W., Owens, I., Goodsell, B., 2006. Past and future mass balance of ‘Ka Roimata o Hine Hukatere’ Franz Josef Glacier, New Zealand. Journal of Glaciology 52 (179), 597e607. Anderson, R.F., et al., 2009. Wind-driven upwelling in the Southern Ocean and the deglacial rise in atmospheric CO2. Science 323, 1443e1448. Anderson, B., et al., 2010. Climate sensitivity of a high-precipitation glacier in New Zealand. Journal of Glaciology 56 (195), 114e128. Applegate, P.J., et al., 2012. Improved moraine age interpretations through explicit matching of geomorphic process models to cosmogenic nuclide measurements from single landforms. Quaternary Research 77 (2), 293e304. Augustinus, P.C., D’Costa, D., Deng, Y., Hagg, J., Shane, P., 2011. A multi-proxy record of changing environments from ca. 30 000 to 9000 cal. a BP: Onepoto maar palaeolake, Auckland, New Zealand. Journal of Quaternary Science 26 (4), 389e401. Balco, G., Schaefer, J.M., 2006. Cosmogenic-nuclide and varve chronologies for the deglaciation of southern New England. Quaternary Geochronology 1 (1), 15e28. Balco, G., Stone, J.O., Lifton, N.A., Dunai, T.J., 2008. A complete and easily accessible means of calculating surface exposure ages or erosion rates from 10Be and 26Al measurements. Quaternary Geochronology 4, 93e107. Bard, E., Rickaby, R.E.M., 2009. Migration of the subtropical front as a modulator of glacial climate. Nature 460, 380e384. Barker, S., et al., 2009. Interhemispheric Atlantic seesaw response during the last deglaciation. Nature 457 (7233), 1097e1102. Barrell, D.J.A., Andersen, B.G., Denton, G.H., 2011. Glacial geomorphology of the central South Island, New Zealand. GNS Science Monograph 27. Lower Hutt, New Zealand. Map and 71 pp. Barrell, D.J.A., 2011. Quaternary glaciers of New Zealand. In: Ehlers, J., Gibbard, P.L., Hughes, P.D. (Eds.), Quaternary Glaciations e Extent and Chronology, Part IV e a Closer Look. Developments in Quaternary Science. Elsevier, Amsterdam, pp. 1047e1064. Author's personal copy 140 A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Barrows, T.T., Juggins, S., De Deckker, P., Calvo, E., Pelejero, C., 2007. Long-term sea surface temperature and climate change in the Australian-New Zealand region. Paleoceanography 22, PA22215. Berger, A., Loutre, M.F., 1991. Insolation values for the climate of the last 10000000 years. Quaternary Science Reviews 10 (4), 297e317. Bevington, P., Robinson, D., 1992. Data Reduction and Error Analysis for the Physical Sciences. WCB McGraw Hill. Birkel, S.D., et al., 2012. Climate inferences from a glaciological reconstruction of the Late Pleistocene Wind River Ice Cap, Wind River Range, Wyoming. Arctic, Antarctic, and Alpine Research 44 (3), 265e276. Birkeland, P.W., 1982. Subdivision of Holocene glacial deposits, Ben Ohau Range, New Zealand, using relative-dating methods. Geological Society of America Bulletin 93, 443e449. Blunier, T., Brook, E.J., 2001. Timing of millennial-scale climate change in Antarctica and Greenland during the last glacial period. Science 291 (5501), 109e112. Bond, G.C., et al., 1999. The north Atlantic’s 1e2 kyr climate rhythm: relation to Heinrich events, Dansgaard/oeschger Cycles and the Little Ice Age. Geophysical Monograph 112, 35e58. Bouttes, N., Paillard, D., Roche, D.M., 2010. Impact of brine-induced stratiﬁcation on the glacial cycle. Climate of the Past Discussions 6, 575e589. Bouttes, N., et al., 2011. Impact of oceanic processes on the carbon cycle during the last termination. Climate of the Past Discussions 7, 1887e1934. Braithwaite, R.J., Olesen, O.B., 1989. Calculation of glacier ablation from air temperature, West Greenland. In: Oerlemans, J. (Ed.), Glacier Fluctuations. D. Reidel Publishing Company, Dordrecht, p. 160. Braithwaite, R.J., 2011. Why do we expect glacier melting to increase under global warming? In: Blanco, J., Kheradman, H. (Eds.), Climate Change e Geophysical Foundations and Ecological Effects. InTech, p. 520. Broecker, W.S., van Donk, J., 1970. Insolation changes, ice volume, and the 18O record in deep-sea cores. Reviews of Geophysics 8 (1), 169e198. Broecker, W.S., 1966. Absolute dating and the astronomical theory of glaciation. Science 151, 299e304. Broecker, W.S., 1998. Paleocean circulation during the last deglaciation: a bipolar seesaw? Paleoceanography 13 (2), 119e121. Brown, E., et al., 2001. Ocean Circulation, second ed. The Open University. Burke, A., Robinson, L.F., 2012. The Southern Ocean’s role in carbon exchange during the last deglaciation. Science 335, 557e561. Burrows, C.J., 2005. Julius Haast in the Southern Alps. Canterbury University Press, Christchurch. Calvo, E., Pelejero, C., De Deckker, P., Logan, G.A., 2007. Antarctic deglacial pattern in a 30 kyr record of sea surface temperature offshore South Australia. Geophysical Research Letters 34 (13), L13707. Carter, L., et al., 1998. Ocean Circulation New Zealand. In: NIWA Chart Miscellaneous Series, vol. 76. NIWA, Wellington. Chiessi, C.M., et al., 2008. South Atlantic interocean exchange as the trigger for the Bolling warm event. Geology 36 (12), 919e922. Chinn, T., Winkler, S., Salinger, M.J., Haakensen, N., 2005. Recent glacier advances in Norway and New Zealand: a comparison of their glaciological and meterological causes. Geograﬁska Annaler 87A, 141e157. Chinn, T.J., 1996. New Zealand glacier responses to climate change of the past century. New Zealand Journal of Geology and Geophysics 39, 415e428. Clark, P.U., Mix, A.C., 2002. Ice Sheets and sea level of the Last Glacial Maximum. Quaternary Science Reviews 21, 1e7. Clark, P.U., et al., 2009. The Last Glacial Maximum. Science 325, 710e714. Cox, S.C., Barrell, D.J.A., 2007. Geology of the Aoraki Area. 1:250,000 Geological Map 15. GNS Science. GNS Science, Lower Hutt, New Zealand, 1 Sheet and 71 pp. Crowley, T.J., 1992. North Atlantic deep water cools the Southern Hemisphere. Paleoceanography 7 (4), 489e497. Curry, B.B., Grimley, D.A., 2006. Provenance, age, and environment of midWisconsinan slackwater lake sediment in the St. Louis Metro East area, USA. Quaternary Research 65, 108e122. Curry, B.B., Petras, J., 2011. Chronological framework for the deglaciation of the Lake Michigan lobe of the Laurentide Ice Sheet from ice-walled deposits. Journal of Quaternary Science 26 (4), 402e410. Davis, K., Burbank, D.W., Fisher, D., Wallace, S., Nobes, D., 2005. Thrust-fault growth and segment linkage in the active Ostler fault zone, New Zealand. Journal of Structural Geology 27 (8), 1528e1546. De Deckker, P., Moros, M., Perner, K., Jansen, E., 2012. Inﬂuence of the tropics and southern westerlies on glacial interhemispheric asymmetry. Nature Geoscience 5, 266e269. Dee, D.P., et al., 2011. The ERA-Interim reanalysis: conﬁguration and performance of the data assimilation system. Quarterly Journal of the Royal Meteorological Society 137, 553e597. Denton, G.H., et al., 1999a. Interhemispheric linkage of paleoclimate during the last glaciation. Geograﬁska Annaler 81A (2), 107e153. Denton, G.H., et al., 1999b. Geomorphology, stratigraphy, and radiocarbon chronology of Llanquihue drift in the area of the southern Lake District, Seno Reloncaví, and Isla Grande de Chiloé, Chile. Geograﬁska Annaler 81A, 167e229. Denton, G.H., et al., 2010. The last glacial termination. Science 328, 1652e1656. Desilets, D., Zreda, M., Prabu, T., 2006. Extended scaling factors for in situ cosmogenic nuclides: new measurements at low latitude. Earth and Planetary Science Letters 246, 256e276. Doughty, A.M., et al., 2012. Late-glacial temperatures in the Southern Alps of New Zealand based on glacier modelling at Irishman basin. Quaternary Science Reviews. http://dx.doi.org/10.1016/j.quascirev.2012.09.013. Drost, F., Renwick, J., Bhaskaran, B., Oliver, H., McGregor, J., 2007. A simulation of New Zealand’s climate during the Last Glacial Maximum. Quaternary Science Reviews 26, 2505e2525. Dunai, T., 2001. Inﬂuence of secular variation of the magnetic ﬁeld on production rates of in situ produced cosmogenic nuclides. Earth and Planetary Science Letters 193, 197e212. Dunai, T.J., 2010. Cosmogenic Nuclides: Principles, Concepts, and Applications in the Earth Surface Sciences. Cambridge University Press, New York. EPICA Community Members, 2006. One-to-one coupling of glacial climate variability in Greenland and Antarctica. Nature 444 (7116), 195e198. Fastook, J.L., Prentice, M., 1994. A ﬁnite-element model of Antarctica: sensitivity test for meteorological mass-balance relationship. Journal of Glaciology 40 (134), 167e175. Fastook, J.L., Head, J.W., Marchant, D.R., Forget, F., 2008. Tropical mountain glaciers on Mars: altitude-dependence of ice accumulation, accumulation conditions, formation times, glacier dynamics, and implications for planetary spin-axis/ orbital history. Icarus 198, 305e317. Gair, H.S., 1967. Sheet 20-Mount Cook, Geological Map of New Zealand. Department of Scientiﬁc and Industrial Research, Wellington. Gersonde, R., et al., 2003. Last glacial sea surface temperatures and sea-ice extent in the Southern Ocean (Atlantic-Indian sector): a multiproxy approach. Paleoceanography 18 (3). http://dx.doi.org/10.1029/2002PA000809. Gersonde, R., Crosta, X., Abelmann, A., Armand, L., 2005. Sea-surface temperature and sea ice distribution of the Southern Ocean at the EPILOG Last Glacial Maximum e a circum-Antarctic view based on siliceous microfossil records. Quaternary Science Reviews 24 (7e9), 869e896. Ghisetti, F.C., Gorman, A.R., Sibson, R.H., 2007. Surface breakthrough of a basement fault by repeated seismic slip episodes: the Ostler Fault, South Island, New Zealand. Tectonics 26. http://dx.doi.org/10.1029/2007TC002146. Glen, J.W., 1955. The creep of polycrystalline ice. Proceedings of the Royal Society of London 228, 519e538. Golledge, N.R., et al., 2012. Last Glacial Maximum climate in New Zealand inferred from a modelled Southern Alps iceﬁeld. Quaternary Science Reviews 46, 30e45. Gordon, A.L., 1981. Seasonality of Southern Ocean sea ice. Journal of Geophysical Research 86, 4193e4197. Gordon, A.L., 1991. Two stable modes of Southern Ocean winter stratiﬁcation. In: Chu, P.C., Gascard, J.C. (Eds.), Deep Convection and Deep Water Formation in the Oceans. Elsevier Science Publishers B. V.. Hays, J.D., Imbrie, J., Shackleton, N.J., 1976. Variations in the earth’s orbit: pacemaker of the ice ages. Science 194 (4270), 1121e1132. Henderson, R.D., Thompson, S.M., 1999. Extreme rainfalls in the Southern Alps of New Zealand. Journal of Hydrology (NZ) 38, 309e330. Heusser, C.J., 1990. Chilotan piedmont glacier in the southern Andes during the last glacial maximum. Revista Geologíca de Chile 17, 3e18. Hijmans, R., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A., 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25, 1965e1978. Huybers, P., Denton, G., 2008. Antarctic temperature at orbital timescales controlled by local summer duration. Nature Geoscience 1 (11), 787e792. Huybers, P., 2009. Antarctica’s orbital beat. Science 325 (5944), 1085e1086. Indermuhle, A., et al., 1999. Holocene carbon-cycle dynamics based on CO2 trapped in ice at Taylor Dome, Antarctica. Nature 398 (6723), 121e126. Indermühle, A., Monnin, E., Stauffer, B., Stocker, T.F., Wahlen, M., 2000. Atmospheric CO2 concentration from 60 to 20 kyr BP from the Taylor Dome ice core, Antarctica. Geophysical Research Letters 27 (5), 735e738. Johnson, J., Fastook, J.L., 2002. Northern Hemisphere glaciations and its sensitivity to basal melt water. Quaternary International 95e96, 65e74. Kaiser, J., Lamy, F., Hebbeln, D., 2005. A 70-kyr sea surface temperature record off southern Chile (Ocean Drilling Program Site 1233). Paleoceanography 20 (4), PA4009. http://dx.doi.org/10.1029/2005PA001146. Kaplan, M.R., et al., 2010. Glacier retreat in New Zealand during the Younger Dryas stadial. Nature 467, 194e197. Kaplan, M.R., et al., 2011. In-situ cosmogenic 10Be production rate at Lago Argentino, Patagonia: implications for late-glacial climate chronology. Earth and Planetary Science Letters 309, 21e32. Kawamura, K., et al., 2007. Northern Hemisphere forcing of climatic cycles in Antarctica over the past 360,000 years. Nature 448 (7156), 912e916. Keeling, R.F., Visbeck, M., 2011. On the linkage between Antarctic surface water stratiﬁcation and global deep-water temperature. Journal of Climate 24, 3545e3557. Kelly, M.A., 2003. The late Würmian Age in the western Swiss Alps e last glacial maximum (LGM) ice-surface reconstruction and 10Be dating of late-glacial features. Ph.D. thesis. University of Bern, 105 pp. Laabs, B.J.C., Plummer, M.A., Mickelson, D.M., 2006. Climate during the last glacial maximum in the Wasatch and southern Uinta Mountains inferred from glacier modeling. Geomorphology 75, 300e317. Labeyrie, L., et al., 1996. Hydrographic changes of the Southern Ocean (southeast Indian sector) over the last 230 kyr. Paleoceanography 11 (1), 57e76. Lamont, G.N., Chinn, T.J.H., Fitzharris, B.B., 1999. Slopes of glacier ELAs in the Southern Alps of New Zealand in relation to atmospheric circulation patterns. Global and Planetary Change 22, 209e219. Lamy, F., et al., 2004. Antarctic timing of surface water changes off Chile and Patagonian ice sheet response. Science 304 (5679), 1959e1962. Lamy, F., et al., 2007. Modulation of the bipolar seesaw in the southeast paciﬁc during Termination 1. Earth and Planetary Science Letters 259 (3e4), 400e413. Author's personal copy A.E. Putnam et al. / Quaternary Science Reviews 62 (2013) 114e141 Le Meur, E., Vincent, C., 2003. A two-dimensional shallow ice ﬂow model of glacier de Saint Sorlin, France. Journal of Glaciology 49, 527e538. Le Meur, E., Gagliardini, O., Zwinger, T., Ruokolainen, J., 2004. Glacier ﬂow modelling: a comparison of the shallow ice approximation and the full-Stokes solution. Comptes Rendus Physique 5 (7), 709e722. Le Meur, E., Gerbaux, M., Schäfer, M., Vincent, C., 2007. Disappearance of an alpine glacier over the 21st century simulated from modeling its future surface mass balance. Earth and Planetary Science Letters 261, 367e374. Lemieux-Dudon, B., et al., 2010. Consistent dating for Antarctic and Greenland ice cores. Quaternary Science Reviews 29 (1e2), 8e20. Leysinger Vieli, G., Gudmundsson, G.H., 2004. On estimating length ﬂuctuations of glaciers caused by changes in climatic forcing. Journal of Geophysical Research 109 (F1). http://dx.doi.org/10.1029/2003JF000027. Lifton, N., et al., 2005. Addressing solar modulation and long-term uncertainties in scaling secondary cosmic rays for in situ cosmogenic nuclide applications. Earth and Planetary Science Letters 239, 140e161. Lifton, N., Smart, B., Shea, M., 2008. Scaling time-integrated in situ cosmogenic nuclide production rates using a continuous geomagnetic model. Earth and Planetary Science Letters 268, 190e201. Lowell, T.V., Hayward, R.K., Denton, G.H., 1999. Role of climate oscillations in determining ice-margin position: hypothesis, examples, and implications. Geological Society of America Special Papers 337, 193e203. MARGO Project Members, 2009. Constraints on the magnitude and patterns of ocean cooling at the Last Glacial Maximum. Nature Geoscience 2, 127e132. McGlone, M.S., Moar, N.T., 1998. Dryland Holocene vegetation history, Central Otago and the Mackenzie Basin, South Island, New Zealand. New Zealand Journal of Botany 36, 91e111. Mercer, J.H., 1972. Chilean glacial chronology 20,000 to 11,000 carbon-14 years ago: some global comparisons. Science 176, 1118e1120. Mercer, J.H., 1984. Simultaneous climatic change in both hemispheres and similar bipolar warming: evidence and implications. In: Hansen, J.E., Takahashi, T. (Eds.), Climate Processes and Climate Sensitivity. Geophysical Monograph American Geophysical Union, Washington. D. C., pp. 307e313. Milankovitch, M., 1941. Kanon der Erdbestrahlung und seine Andwendung auf das Eiszeiten-problem. Royal Serbian Academy, Belgrade. Mix, A.C., Bard, E., Schneider, R., 2001. Environmental processes of the ice age: land, oceans, glaciers (EPILOG). Quaternary Science Reviews 20, 627e657. Monnin, E., et al., 2001. Atmospheric CO2 concentrations over the last glacial termination. Science 291, 112e114. Newnham, R.M., Loewe, D.J., Giles, T., Alloway, B.V., 2007a. Vegetation and climate of Auckland, New Zealand, since ca. 32,000 cal. yr ago: support for an extended LGM. Journal of Quaternary Science 22 (5), 517e534. Newnham, R.M., Vandergoes, M.J., Hendy, C.H., Lowe, D.J., Preusser, F., 2007b. A terrestrial palynological record for the last two glacial cycles from southwestern New Zealand. Quaternary Science Reviews 26, 517e535. Nishiizumi, K., et al., 2007. Absolute calibration of 10Be AMS standards. Nuclear Instruments and Methods in Physics Research B 258, 403e413. Norton, D., 1985. A multivariate technique for estimating New Zealand temperature normals. Weather and Climate 5, 64e74. Oerlemans, J., 1997. Climate sensitivity of Franz Josef Glacier, New Zealand, as revealed by numerical modelling. Arctic and Alpine Research 29 (2), 233e239. Paterson, W.S.B., 1994. The Physics of Glaciers. Pergammon, Oxford. Pelejero, C., Calvo, E., Barrows, T.T., Logan, G.A., De Deckker, P., 2006. South Tasman Sea alkenone palaeothermometry over the last four glacial/interglacial cycles. Marine Geology 230, 73e86. Petit, J.R., et al., 1999. Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399 (6735), 429e436. Porter, S.C., 1975. Equilibrium-line altitudes of Late Quaternary glaciers in the Southern Alps, New Zealand. Quaternary Research 5, 27e47. Purdie, H., et al., 2011. Interannual variability in net accumulation on Tasman Glacier and its relationship with climate. Global and Planetary Change. http:// dx.doi.org/10.1016/j.gloplacha.2011.04.004. Putkonen, J., Swanson, T., 2003. Accuracy of cosmogenic ages for moraines. Quaternary Research 59, 255e261. Putnam, A.E., et al., 2010a. Glacier advance in southern middle latitudes during the Antarctic Cold Reversal. Nature Geoscience 3, 700e704. 141 Putnam, A.E., et al., 2010b. In situ cosmogenic 10Be production-rate calibration from the Southern Alps, New Zealand. Quaternary Geochronology 5, 392e409. Refsnider, K.A., et al., 2008. Last glacial maximum climate inferences from cosmogenic dating and glacier modeling of the western Uinta ice ﬁeld, Uinta Mountains, Utah. Quaternary Research 69, 130e144. Reimer, P.J., et al., 2009. INTCAL09 and MARINE09 radiocarbon age calibration curves, 0e50,000 years cal BP. Radiocarbon 51 (4), 1111e1150. Robinson, R.S., Mix, A., Martinez, P., 2007. Southern Ocean control on the extent of denitriﬁcation in the southeast Paciﬁc over the last 70 ka. Quaternary Science Reviews 26 (1e2), 201e212. Roe, G., 2006. In defense of Milankovitch. Geophysical Research Letters 33 (L24703). http://dx.doi.org/10.1029/2006GL027817. Sandiford, A., Newnham, R.M., Alloway, B.V., Ogden, J., 2003. A 28 000e7600 yr BP pollen record of vegetation and climate change from the Pukaki Crater, northern New Zealand. Palaeogeography Palaeoclimatology Palaeoecology 201, 235e247. Schaefer, J.M., et al., 2006. Near-synchronous interhemispheric termination of the last glacial maximum in mid-latitudes. Science 312 (5779), 1510e1513. Schaefer, J.M., et al., 2009. High-frequency Holocene glacier ﬂuctuations in New Zealand differ from the northern signature. Science 324, 622e625. Schäfer, M., Gagliardini, O., Pattyn, F., Le Meur, E., 2008. Applicability of the shallow ice approximation inferred from model inter-comparison using various glacier geometries. Cryosphere Discussions 2 (4), 557e599. Shemesh, A., et al., 2002. Sequence of events during the last deglaciation in Southern Ocean sediments and Antarctic ice cores. Paleoceanography 17 (4). http://dx.doi.org/10.1029/2000PA000599. Sigman, D.M., Haug, G.H. (Eds.), 2003. The Biological Pump in the Past. Treatise on Geochemistry, vol. 6. Elsevier Pergamon. Sigman, D.M., Jaccard, S.L., Haug, G.H., 2004. Polar ocean stratiﬁcation in a cold climate. Nature 428 (6978), 59e63. Sigman, D.M., de Boer, A.M., Haug, G.H., 2007. Antarctic stratiﬁcation, atmospheric water vapor, and Heinrich Events: a hypothesis for Late Pleistocene deglaciations. In: Schmittner, A., Chiang, J.C.H., Hemming, S.R. (Eds.), Ocean Circulation: Mechanisms and Impacts. Geophysical Monograph Series, vol. 173. American Geophysical Union, Washington, pp. 335e349. Sigman, D.M., Hain, M.P., Haug, G.H., 2010. The polar ocean and glacial cycles in atmospheric CO2 concentration. Nature 466, 47e55. Sikes, E.L., et al., 2009. Southern Ocean seasonal temperature and Subtropical Front movement on the South Tasman Rise in the late Quaternary. Paleoceanography 24, PA2201. Staiger, J., et al., 2007. Atmospheric scaling of cosmogenic nuclide production: climate effect. Journal of Geophysical Research 112. http://dx.doi.org/10.1029/ 2005JB003811. Stone, J.O., 2000. Air pressure and cosmogenic isotope production. Journal of Geophysical Research 105, 23753e23759. Strelin, J.A., Denton, G.H., Vandergoes, M.J., Ninnemann, U.S., Putnam, A.E., 2011. Radiocarbon chronology of the late-glacial Puerto Bandera moraines, Southern Patagonian Iceﬁeld, Argentina. Quaternary Science Reviews 30 (19e20), 2551e2569. Sugden, D.E., John, B.S., 1976. Glaciers and Landscape: a Geomorphological Approach. Edward Arnold, London. Suggate, R.P., 1990. Late Pliocene and Quaternary glaciations of New Zealand. Quaternary Science Reviews 9, 175e197. Toggweiler, J.R., Russell, J.L., Carson, S.R., 2006. Midlatitude westerlies, atmospheric CO2, and climate change during the ice ages. Paleoceanography 21 (2), PA2005. Toggweiler, J.R., 2008. Origin of the 100,000-year timescale in Antarctic temperatures and atmospheric CO2. Paleoceanography 23 (2), PA2211. http://dx.doi.org/ 10.1029/2006PA001405. Toggweiler, J.R., 2009. Shifting westerlies. Science 323 (5920), 1434e1435. Vandergoes, M.J., et al., 2005. Regional insolation forcing of late Quaternary climate change in the Southern Hemisphere. Nature 436, 242e245. Ward, G.K., Wilson, S.R., 1978. Procedures for comparing and combining radiocarbon age determinations: a critique. Archaeometry 20, 19e31. Weertman, J., 1964. The theory of glacier sliding. Journal of Glaciology 5 (39), 287e303. Whittaker, T.E., Hendy, C.H., Hellstrom, J.C., 2011. Abrupt millennial-scale changes in intensity of Southern Hemisphere westerly winds during marine isotope stages 2e4. Geology 39 (5), 455e458.

RELATED PAPERS

RELATED TOPICS

Log In

The Last Glacial Maximum at 44°S documented by a 10Be moraine chronology at Lake Ohau, Southern Alps of New Zealand

The Last Glacial Maximum at 44°S documented by a 10Be moraine chronology at Lake Ohau, Southern Alps of New Zealand

Related Papers

RELATED PAPERS

RELATED TOPICS