Antarctic Climate and Forcing at Glacial Interglacial Time scales

Rather than give a historical account of increasing knowledge in this area (which would focus strongly on Vostok), we will here use the longest (in age) record from Dome C to describe what is known, and then discuss what the similarities and differences of other records indicate.

Figure 11.3 shows the record (using water isotopes) of Antarctic temperature at Dome C over the last 800 ka (Jouzel et al., 2007). The age scale has been derived using an ice flow and snow accumulation model constrained, and in some parts tuned, to independent age markers of known age (Parrenin et al., 2007a). The maximum uncertainty of the absolute ages is estimated to be better than 6 kyr throughout the core. Through most of its span it shows good coherence with signals recorded in the benthic marine oxygen isotope stack (Fig. 11.3, representing some combination of global ice

Figure 11.3: Top panel: marine benthic oxygen isotope stack on LR04 age scale (Lisiecki and Raymo, 2005). Lower panels: 800 kyr record, on EDC3 age scale of Deuterium, and temperature (difference from the last millennium, derived using a correction for seawater isotopic content and modelled ice-sheet altitude) (Jouzel et al., 2007). Numbers above the marine curve represent selected Marine Isotope Stages.

Figure 11.3: Top panel: marine benthic oxygen isotope stack on LR04 age scale (Lisiecki and Raymo, 2005). Lower panels: 800 kyr record, on EDC3 age scale of Deuterium, and temperature (difference from the last millennium, derived using a correction for seawater isotopic content and modelled ice-sheet altitude) (Jouzel et al., 2007). Numbers above the marine curve represent selected Marine Isotope Stages.

volume and deep-water temperature) whose time scale was derived in a completely different way (Lisiecki and Raymo, 2005).

The first obvious feature of the record is that a generally cold climate (up to 9°C colder than the late Holocene in this part of Antarctica) was interrupted approximately every 100 kyr by warmings (such as the Holocene interglacial) lasting approximately 10-30 kyr. The last four interglacials (back to Marine Isotope Stage 11) had periods that were 2-4.5°C warmer than the last millennium. Before Marine Isotope Stage 11, a different pattern is seen: although 100 kyr still emerges as the dominant period of the record (Jouzel et al., 2007), the earlier interglacials are significantly cooler than the later ones, and the system seems to spend a higher proportion of time in the warmer phase. There is not yet any convincing explanation for this change in amplitude of the signal (discussed below).

The next question that arises is whether the Dome C record can be considered representative for a larger part of the Antarctic. Comparison with

Vostok (around 500 km away) shows excellent agreement, but more importantly a very similar signal (albeit with small differences in the age scale) is seen at Dome Fuji, on the opposite side of the East Antarctic plateau (Watanabe et al., 2003; Kawamura et al., 2007). This similarity suggests that the Dome C record can stand as a signal of climate across much of the polar plateau.

The 100 kyr periodicity seen in the ice core record is of course already well-known from other climate records, and assumed to derive in some way from orbital forcing amplified by internal feedbacks. One of the most important amplifiers is the greenhouse gas concentration of the atmosphere. The records of greenhouse gases from Dome C (completed by those from Vostok in as yet unmeasured sections of Dome C) extend so far to 650 ka (and are currently being continued to 800 ka). They show very strong congruence with many features of the temperature record, and are consistent with CO2 in particular playing a significant role in such amplification (Figs. 11.3 and 11.4).

Concentrations of CO2 are typically around 180-200 ppmv in the coldest parts of each cycle, and reached 280-300 ppmv during the warm periods back to Marine Isotope Stage 11 (Siegenthaler et al., 2005). They were however considerably lower (~ 250 ppmv) during the warm periods before Marine Isotope Stage 11, scaling with Antarctic temperature (Fig. 11.4). The very strong similarity of CO2 concentrations and Antarctic temperature strongly suggests that the Southern Ocean plays a leading role in controlling the atmospheric concentration of CO2 on these time scales, as already implied by more detailed records of the last deglaciation (Monnin et al., 2001). Concentration of CO2 in the atmosphere today (ca. 380 ppmv) is nearly 30% higher than the highest value seen in the previous 650 kyr.

Methane (CH4) concentrations (Spahni et al., 2005) show strong similarities to the Antarctic temperature record (Fig. 11.4), with high concentrations in warm periods, low concentrations in glacial maxima and lower concentrations in 'weaker' interglacials compared to 'stronger' ones. However in this case, even at the low resolution shown, it is apparent that there are higher-frequency millennial-scale variations superimposed on the orbital-scale trends. In the most recent climatic cycle (Blunier and Brook, 2001) these correspond to the D-O events that will be discussed later. Methane concentrations are most likely mainly controlled by changing wetland extents in the northern high latitudes and tropics, along with changes in sinks (Valdes et al., 2005), and thus are relevant to Antarctic climate only as a relatively minor amplifier.

Ice cores have provided other data that are important for understanding Antarctic climate and ice sheets. The snow accumulation rate is clearly

Figure 11.4: Records of Deuterium, non-sea-salt Calcium flux (Wolff et al., 2006), CO2 (Siegenthaler et al., 2005) and CH4 (Spahni et al., 2005) measured at Dome C. Note that there maybe small timing mismatches because the gases are shown on the EDC2 age scale, while the other records are on EDC3.

Figure 11.4: Records of Deuterium, non-sea-salt Calcium flux (Wolff et al., 2006), CO2 (Siegenthaler et al., 2005) and CH4 (Spahni et al., 2005) measured at Dome C. Note that there maybe small timing mismatches because the gases are shown on the EDC2 age scale, while the other records are on EDC3.

important for the mass balance of the ice sheet. In central Antarctica, it has generally been treated by a modelling approach in which the accumulation rate is thermodynamically related to the local temperature. Snow accumulation scales non-linearly with temperature, and was around half its present value in central East Antarctica during the coldest periods of the last 800 kyr than it is at present. Ice sheet elevation in central Antarctica is assumed to have varied little over the last 800 kyr: simple modelling exercises (Parrenin et al., 2007b) used as input to age models suggest variations of less than 200 m at both Dome C and Dome Fuji. Measurements of air volume (which has atmospheric pressure as one among its controlling variables) provide only a limited constraint, confirming that altitude changes of more than a few hundred metres have not occurred. Of course, it is possible for the areal extent of an ice sheet to change considerably without any great change in the altitude of its central parts.

The dust flux to Antarctica (represented in Fig. 11.4 by non-sea-salt calcium) was much higher during glacial maxima than during warm periods. This is believed to be a result of changes in climate in southern South America (Wolff et al., 2006), although the exact nature of the relevant changes is not yet clear. In any case the result is a change in aerosol content of the atmosphere over Antarctica (with possible radiative impacts) and a change in the amount of dust deposited in the Southern Ocean which may also play a role in stimulating phytoplankton productivity, and thus carbon dioxide drawdown from the atmosphere. Finally in this section, it has been suggested that the sea salt content of Antarctic ice cores might yield a record of past sea ice extent (Wolff et al., 2006). Sea salt, as expected under this idea, is at higher concentrations during cold periods, consistent with the findings of marine sediments (Gersonde et al., 2005) that sea ice was significantly extended at such times.

Was this article helpful?

0 0

Post a comment