Southern Exposure: New Paleoclimate Insights From Southern Ocean and Antarctic Margin Sediments

Much of what is known about the evolution of Antarctica's cryosphere in the geologic past is derived from ice-distal deep-sea sedimentary records. Recent advances in drilling technology and climate proxy methods have made it possible to retrieve and interpret high-quality ice-proximal sedimentary sequences from Antarctica's margins and the Southern Ocean. These records contain a wealth of information about the individual histories of the East and West Antarctic Ice Sheets and associated temperature change in the circum-Antarctic seas. Emerging studies of Antarctic drill cores provide evidence of dynamic climate variability on both short and long timescales over the past 20 million years. This geologic information is critical for testing and improving computer model simulations used to predict future environmental change in the polar regions. Identifying the mechanistic links between past Antarctic ice-volume fluctuations and oceanographic change is necessary for understanding Earth's long-term climate evolution. While recent successes highlight the value of ice-proximal records, additional scientific drilling and climate proxy development are required to improve current knowledge of Antarctica's complex paleoenvironmental history.

S p e c i a l i S S u e o N a N ta r c t i c o c e a N o g r a p h y i N a c h a N g i N g W o r l d   (Gille, 2002;Turner et al., 2005;Steig et al., 2009) associated with changes in circulation (Purkey and Johnson, 2010), sea ice extent (Stammerjohn et al., 2008), ice sheet stability (Rignot and Jacobs, 2002;Shepherd et al., 2004), and regional ecology (Vaughan et al., 2003). Climate models indicate that the polar regions are highly sensitive to rising atmospheric CO 2 concentrations and associated warming (Bitz et al., 2012), but the high-latitude response to on going and future warming in a high-CO 2 world remains uncertain due to a lack of understanding of complex interactions between the Antarctic cryosphere and the global climate system (IPCC, 2007 Kennett, 1977;Miller et al., 1987;Zachos et al., 2001;Cramer et al., 2009 Kennett, 1977;DeConto and Pollard, 2003). A leading hypothesis, supported by climate models and paleoclimate proxy data, is that Cenozoic Antarctic ice sheet development was driven by changes in atmospheric CO 2 (Figure 2; Pearson and Palmer, 2000;DeConto and Pollard, 2003;Pagani et al., 2005Pagani et al., , 2011DeConto et al., 2008;Pearson et al., 2009 (Barrett et al., 2007;Naish et al., 2008). In parallel, non-CaCO 3 -based paleoceanographic and paleoclimatic proxies (e.g., Schouten et al., 2002;Belt et al., 2007)    caSe StudieS iN aNtarctic paleoeNViroNmeNtal recoNStructioNS case Study 1: the last deglaciation/holocene Much of our knowledge of atmospheric chemical composition, circulation, and temperature change over the last 800,000 years comes from annually resolved ice core records obtained from Antarctica's ice sheets (e.g., Luthi et al., 2008). These records reveal a close coupling of atmospheric CO 2 and climate that is not fully understood, but likely regulated by physical and/ or biogeochemical processes occurring in the ocean (e.g., Sigman et al., 2010).
Although a central role for the Southern Ocean in glacial-interglacial CO 2 variations has been hypothesized, evidence of direct Antarctic temperature-CO 2  (deconto et al., 2008). The dashed line highlights preindustrial co 2 levels. The red envelope includes atmospheric co 2 concentrations for the representative concentration pathways (rcps) to 2100 (ipcc, 2007; meinshausen et al., 2011) and the red, gray, dashed gray, yellow, and blue lines indicate extended concentration pathways (ecps) to 2300 (meinshausen et al., 2011). The ecps are within the range reconstructed for the early cenozoic, illustrating the importance of geologic drilling in providing long-term paleoclimatic context. (c) global atmospheric temperature estimates based on the deep-sea benthic foraminifer oxygen isotope (δ 18 o) compilation of Zachos et al. (2001). The benthic foraminifer δ 18 o signal reflects variations in both deep ocean temperature and global ice volume. White bars indicate key antarctic margin reference sections recovered by the cape roberts project (crp), aNdrill (aNd), integrated ocean drilling program (iodp), and Shaldril (Shl); core recovery is indicated in parentheses. The green and blue-banded bars reveal the evolution of antarctica's ice sheets, as inferred from geologic data. pre-industrial Kominz et al (2008) Zachos et al (2001) Millions of years before present Sea Level (m) Atmospheric CO 2 (ppm) Pagani et al (2005) Average air temperature ( Although the observed increases in biogenic opal flux to the sediments indicate that some nutrients were utilized regionally by phytoplankton, ice core CO 2 records (Monnin et al., 2001) suggest that phytoplankton production did not keep pace with upwelling, or that production was iron/micronutrient limited, resulting in a substantial release of old carbon to the atmosphere (Skinner et al., 2010).

RCPs ECPs
Although it is now clear that Southern  (Rignot and Jacobs, 2002;Shepherd et al., 2004) and modeling (Pollard and DeConto, 2009) that ocean temperature estimates from the interface of Antarctica's cryosphere and the Southern Ocean are required to assess the relative importance of ocean heat on grounding line retreat and to validate existing ice sheet models.

Much of what is known about
Antarctica's climate in the recent geologic past (15-0 thousand years ago) comes from ultra-high resolution (annual to centennial scale) marine sediments recovered from basins and drifts on the Antarctic continental margin (e.g., Domack et al., 2001;Leventer et al., 2006;Bentley et al., 2009;Escutia et al., 2011). Geochemical, sedimentological, and paleontological studies of these well-dated sequences reveal a wealth of high-quality information on past oceanographic fluctuations, sea ice conditions, meltwater influx, nutrient dynamics, and biologic productivity (e.g., Domack et al., 2001;Leventer et al., 2006;Bentley et al., 2009). Furthermore, these studies provide a geological perspective on modern the Site 1098 TEX 86 -based temperature variability is consistent with local/ regional terrestrial and marine temperature records, implying atmospheric forcing of western Antarctic Peninsula ocean temperatures during the Holocene (Shevenell et al., 2011).
Although the Site 1098 TEX 86 -based temperature trends are generally consistent with independent local and regional paleoenvironmental records, further refinement of the absolute temperature estimates is required (Shevenell et al., 2011). In particular, the warm absolute TEX 86 -derived temperatures, contrasting relative temperature trends interpreted from diatom paleocological studies (Sjunneskog and Taylor, 2002), and similarities between inferred temperature patterns, local (65°S) spring insolation, and spring-blooming diatom abundances (Sjunneskog and Taylor, 2002) at  (Shevenell et al., 2011). While the absolute temperature values will evolve as regional teX 86 calibrations improve, application of multiple existing calibrations to the Site 1098 data set indicate that the observed temperature trends are robust (Shevenell et al., 2011). teX 86 -derived temperatures are compared with local insolation (blue and purple), antarctic ice core (green), and sub-antarctic sea surface temperatures (SSts; red and blue) from the pacific Sector of the Southern ocean, revealing similarities on orbital and millennial timescales (see Shevenell et al., 2011, for details). relationship in the Southern Ocean and regional marine archaeal ecology (e.g., seasonality and depth of lipid production; Church et al., 2003;Shevenell et al., 2011;Kim et al., 2012 Well-dated early to mid-Pliocene

Model Snapshots
McKay et al., 2012). Sedimentological and diatom assemblage data also reveal warm regional ocean temperatures at the onset of many early Pliocene interglacials, providing independent support for TEX 86 -derived temperatures (Naish et al., 2009;McKay et al., 2012).
In studies of Antarctic margin drill cores, an important challenge is to obtain ice-proximal evidence that addresses East Antarctica's contribution to early Pliocene eustasy (e.g., Raymo et al., 2011) (Savin et al., 1975;Kennett, 1977;Flower and Kennett, 1994;Shevenell et al., 2004). During the transition, deep-sea δ 18 O records indicate major ice growth on Antarctica following the warmest interval of the Neogene, when globally distributed records indicate that the global carbon cycle was operating differently and atmospheric CO 2 may have been relatively low (Figure 2; Flower, 1999;Zachos et al., 2001;Pagani et al., 2005;Kurschner et al., 2008). Because a definitive link between ice growth and atmospheric CO 2 concentrations has yet to be established for the middle Miocene (Flower, 1999;Pagani et al., 1999) Verducci et al., 2009;Majewski and Bohaty, 2010), the Ross Sea region (Naish et al., 2007;Warny et al., 2009;Harwood et al., 2008Feakins et al., 2012), the Antarctic Peninsula (J.B. Anderson et al., 2011), and the McMurdo Dry Valleys (Lewis et al., 2008) in the past decade provides an excellent example of the power of integrating terrestrial, shallow marine, and deep-sea records to develop a consistent picture of ice-proximal to distal high-latitude climate change. Southern Ocean SST and bottom water temperature records derived from the Mg/Ca of planktonic (Shevenell et al., 2004) and benthic (Shevenell et al., 2008) foraminifer CaCO 3 suggest the presence of warm waters around Antarctica during the Miocene Climatic Optimum when Southern Ocean waters were relatively warm and atmospheric CO 2 was only slightly higher than present ( Figure 5; Shevenell et al., 2008;Lear et al., 2010). Although ice growth began at 15.5 Ma, about two million years prior to the node in Earth's orbital parameters identified at ~ 13.84 Ma (Holbourn et al., 2005), Southern Ocean surface cooling did not commence until 14.2 Ma, when SSTs cooled 6-7°C in a stepwise fashion, reaching a minimum at 13.8 Ma (Shevenell et al., 2004). Geomorphological evidence from the McMurdo Dry Valleys suggests a shift from wet-based to coldbased glaciation between 14.07 ± 0.05 and 13.85 ± 0.03 Ma, with an estimated mean surface temperature change of 8°C, similar to that observed in Southern Ocean SST records (Shevenell et al., 2004;Lewis et al., 2008). While these observations are consistent over a large geographic area and suggest a role for ocean heat, the dynamic middle Miocene behavior of the Antarctic cryosphere during an interval of relatively low atmospheric CO 2 presents a significant and ongoing challenge for ice sheet and climate modelers. Further research is also required to document trends and reduce pCO 2 uncertainty in the middle Miocene (Pagani et al., 1999;Pearson and Palmer, 2000;Kurschner et al., 2008 Shevenell et al., 2004Shevenell et al., , 2008. heavy mineral provenance studies at aNd-2a support a warmer climate between ~ 17 and 15.5 ma and indicate that the east antarctic ice Sheet (eaiS) had retreated away from the coast and into the transantarctic mountains (hauptvogel and passchier, 2012). pollen data from aNd-2a provide evidence for tundra vegetation around the ross Sea region between 15.7 and 15.5 ma (pictured), suggesting this interval was a period of maximum regional warmth (indicated by the red bar; Warny et al., 2009). heavy mineral provenance work from aNd-2a indicates that at 15.5 ma, the eaiS began to expand toward the coast until ~ 14.3 ma (hauptvogel and passchier, 2012). ice volume estimates from odp Site 1171 indicate orbitally paced ice growth beginning at ~ 15.5 ma during the warm miocene climatic optimum (black arrows indicate glaciations) and continuing until the globally recognized deep-sea δ 18 o increase at 13.8 ma (blue dashed line; Shevenell et al., 2008).