Info

Wisconsin - t

Bubbles

FIGURE 14.60 Relationship of ice core depth to years before the present (bp) and to the type of ice for the Central Greenland Ice Sheet. The Holocene and Wisconsin periods are also marked (adapted from Gow et al., 1997).

Figure 14.59 shows schematically the processes by which gases and particles are trapped from the atmosphere into snow and ice at high latitudes (Delmas, 1992). As snow is deposited, the surface is initially quite porous. As more snow accumulates, it compacts the underlying layers, forming a porous structure known as firn. Atmospheric gases continue to penetrate the porous firn. At some depth (typically ~ 100 m), there is a close-off zone in which recrystallization starts to seal off the pores, trapping the atmospheric constituents in bubbles in the ice. At larger depths, the bubbles are completely sealed off and the trapped gas is preserved. As snow accumulation continues, the ice below is further compacted and the ice sheet spreads down and out. At depths larger than ~1200 m, the hydrostatic pressure is sufficiently large that the air is forced into the ice to form clathrates so that distinct bubbles are no longer evident (Miller, 1969). Clathrates are solid ice lattices that incorporate another molecule such as C02 or CH4 into their crystal lattice (e.g., see Kven-volden, 1993). Once an ice core is drilled, various depths can be dated and the air trapped in the ice as either bubbles or clathrates is recovered for analysis, for example, by crushing the ice sample (Wilson and Long, 1997).

The bubbles found at larger depths therefore correspond to older atmospheres. Figure f4.60 shows the relationship between the ice core depth and age (in number of years before the present time, bp) as well as the characteristics of the ice for samples from the Greenland Ice Sheet Project 2, GISP2 (Gow et al., 1997). Also shown are the periods corresponding to the Holocene (the past 10,000 years) and the Wisconsin ice age, for which some data are shown below, fee at depths of -260, 1000, 2430, and 2759 m corresponds to ages of 1000, 5100, 50,000, and 103,000 years bp, respectively (Grootes and Stuiver, 1997).

Because the firn is ventilated by atmospheric air while the bubbles are forming over a period of time and ice depths, the air eventually trapped in the bubbles is a time-integrated sample that is younger than the snow deposit itself. For example, in one recent study (Smith et al., 1997), the air bubbles were, on average, 220-700 years younger than the ice in which they were embedded, but the difference can be as much as several thousand years (e.g., see Rommelaere et al., 1997). These exchange processes with the atmosphere, gas diffusion, and the porosity and tortuosity of the ice pores have to be taken into account in relating the depth of the core to the age of the trapped air.

As seen earlier (Section B.2), air trapped in these ice cores can be recovered and analyzed to provide a snapshot of the composition of the atmosphere tens or even hundreds of thousands of years ago (but note cautions with respect to potential artifacts, e.g., in situ formation of C02 from carbonate in the bubbles). In addition, ice core composition can be used to infer the local temperature of the atmosphere when the snow/ice was deposited, so-called paleothermometry (Delmas, 1992). The isotopic composition, particularly the IR0/'"0 and D/H ratios, of the ice is related to the temperature at the level of the precipitating cloud that generated the snow/ice. Isotopic fractionation occurs during the natural water cycle and this leads to a relationship between the isotopic composition and the precipitation temperature (Dansgaard, 1964). Once this relationship is established, the isotopic composition of water in the ice core can be used to estimate the corresponding atmospheric temperature. Such relationships, using 1 O as an example, are usually expressed in the form

where SlxO in per mil (%) is defined as 1000(fl -Ra)/R{), R is the isotope ratio of the sample, and R() is the ratio of a standard sample. In the case of oxygen, the standard is usually standard mean ocean water (SMOW). The values of the slope and intercept (a and b, respectively) vary from location to location and likely with time as well (Cuffey and Clow, 1997; Jouzel et al., 1997). However, it appears that such relationships are still useful for inferring historic temperatures (Jouzel et al., 1997; Salamatin et al., 1998).

Ice core data have provided evidence that quite rapid and large oscillations in climate have occurred over the period of record. Figure 14.61, for example, shows the temperature changes in central Greenland for the past ~ 110,000 years. Recent millenia are characterized by relatively small rates of temperature change. Indeed, in summarizing the results of the Greenland Summit Ice Core Projects (GISP2 and GRIP) published in a special issue of the Journal of Geophysical Research (Vol. 102 (C12), pp. 26315-26886, November 30, 1997), Hammer, Mayewski, Peel, and Stuiver state

The ice-core records tell a clear story: humans have come of age agriculturally and industrially in the most stable climate regime of the last 110,000 years. However, even this relatively stable period is marked by change... [which is] more characteristic of the Earth's climate than is stasis.

Figure 14.61 illustrates the much larger changes, more than 20°C from one extreme to the other, that have occurred historically when viewed over these long time periods. Associated with decreases in temperature are decreased methane concentrations, increased dust loadings, and decreased snow accumulations (e.g., see Fig. 14.62). (It should be noted that the amplitude of the temperature changes can vary from site to site; for example, Dahl-Jensen et al. (1998) showed that the

0 0

Post a comment