There is broad agreement that we will be better placed to predict future climate change if we can gain a better understanding of how and why climate varied in the recent geological past. In general the geological records of changing climate that are preserved in the sediments of the world's oceans provide the basic framework for describing the history of climate. On the other hand, the long and continuous cores of ancient ice that have been recovered from Greenland and from Antarctica uniquely contain records of changes in the composition of the atmosphere: changes in the concentrations of greenhouse gases and of dust, both of which have a direct influence on climate. The over-riding objective of this project is to greatly improve the precision with which these two different types of record can be examined on the same time scale. Without this ability, we cannot build numerical models that model changes in past climate, taking account of the composition of the atmosphere. With a fully consistent timescale the massive investments in coring the oceans and the polar ice sheets will yield major benefits in increasing our understanding of climatic processes.
Scientific objectives and approach:
We expect that a correct common time scale will help to clarify some, or all, of the following observations:
a) major, rapid temperature changes that occurred over Greenland and over most of the North Atlantic
b) rapid rises in atmospheric methane
c) changes in atmospheric carbon dioxide observed in air trapped in Antarctic ice
d) changes in the dust flux to the polar ice caps
e) changes in the accumulation of cosmogenic isotopes in the ice cores
Numerical modelling is the principle tool for improving predictions of future climate. Numerical modelling is also a valuable tool for investigating past climate, with benefits both in enhancing the robustness of the models, and in understanding climate.
The main objective of this project is to document the precise temporal relationships between abrupt climatic changes that are known to have occurred in many parts of the globe during the past 350,000 years, and also to gain a global understanding of millennial-scale variability. It is absolutely essential to create a global stratigraphy for millennial variability if we are to understand the climatic processes involved.
The earliest, and still the most important documentation of millennial-scale climatic variability comprises the record of air temperature over Greenland as recorded in all of the long ice cores that have been recovered (Camp Century, Dye3, GISP2, GRIP). Figure. 1 shows one published version of this temperature record. The time interval of interest, from 64,000 yr BP to 24,000 yr BP, shows sharp (within decades) and large (over 10 degrees) changes in temperature. These changes have been repeated at intervals varying between several hundred and about two thousand years.
Figure. 1 Isotope record reflecting air temperature over Greenland between 24 and 64 Kyr
Several hypotheses have been put forward to explain this variability:
A) The "binge and purge" hypotheses. One group of hypotheses links the variability explicitly to ice sheets, suggesting that instabilities ("binge-and-purge" cycles) in the North American Ice sheet were the ultimate cause of high-frequency variability. The oceanic response would then arise through the impact of freshwater fluxes into and out of the ocean on the stability of the North Atlantic deep thermohaline circulation.
B) Internal ocean dynamics. A second group of hypotheses links millennial variability to internal ocean circulatory oscillations.
C) Atmospheric transport. A third group of explanations draws on a forcing arising from changes in the atmospheric transport (hydrological cycle changes) of water vapour between ocean basins, which in turn affects the salinity contrast between the oceans.
Figure 2. Atmospheric methane measured in ancient air bubbles in Greenland ice
Evidence from the Greenland ice cores themselves also demonstrates that many other aspects of the climate system varied in parallel with Greenland air temperature, although the majority of the parameters analysed probably document a response to the climate only of the Atlantic sector. One parameter that may or may not have an origin outside the North Atlantic is the concentration of methane in air bubbles trapped in the ice. Figure. 2 shows a composite published methane record from Greenland cores.
So far as can be detected methane rise and fall was essentially synchronous with temperature changes; this has been carefully investigated without any conflicting evidence emerging.
The Antarctic also underwent significant temperature variability on millennial time scales, as is indicated in Figure 3. From Antarctica there is an ice core record available covering about 400,000 years. Over the shorter interval that overlaps with the Greenland record, the character of the Antarctic signal is significantly different; temperature appears to have risen and fallen gradually, in contrast to the sudden transitions that are observed in the Greenland record. The time scale of the Antarctic oscillations is still uncertain.
Figure 3. Deuterium isotope record from Vostok ice core reflecting Antarctic air temperatures
Clearly, we cannot understand this variability if we cannot determine which region was affected, or whether there were phase leads-and-lags among the regions affected. We propose to tackle this challenge through co-ordinated work on ice cores and deep-sea sediment cores. The inspiration comes from work presented at AGU in December 1999. A core from the southern part of the North Atlantic (off southern Portugal)shows that while the surface temperature record very closely resembles the temperature record over Greenland (and can undoubtedly be accurately correlated with Greenland), the oxygen isotope record recovered from benthic foraminifera has a quite different character (fig 4). The benthic record closely resembles the record from Antarctica, when the Antarctic record is synchronised with Greenland using the atmospheric methane records.
Figure 4. Temperatures over Greenland compared to planktic delta18O in core MD 95-2042, and benthic delta 18O in the same core compared to the Vostok temperatures signal between 24 and 64 ka (N J Shackleton, M A Hall, and E Vincent, 2000, Paleoceanography vol 15, p565-569