Stable Isotope Analysis

The study of stable isotopes (primarily deuterium and 180) is a major focus of paleodimatic research. Most work has been on stable isotope variations in ice and firn, and in the tests of marine fauna recovered from ocean cores. However, increasing attention is being placed on other natural isotope recorders, such as speleothems

Greenlands Ice Core ChartDye Images Greenland
FIGURE 5.1 Location of principal ice-coring sites in the Canadian Arctic and Greenland, and in Antarctica.
Camp Century Greenland
7S" W W '20" 135'
Camp Century Greenland

FIGURE 5.2 Location of principal ice-coring sites in low latitudes.

TABLE 5.2 Locations of Ice Cores with Records Going Back into the Last Glaciation3

Drill Site

Location6

Max. Depth (m)

Camp Century

N.W. Greenland

1387

GISP2 (Summit)

C. Greenland

3053

GRIP

C. Greenland

3029

Dye-3

S. Greenland

2037

Renland

E. Greenland

324

Agassiz

N. Ellesmere Island

338

Devon

Devon Island

299

Barnes

Baffin Island

Penny

Baffin Island

334

Byrd

West Antarctica

2164

J9 (Ross ice shelf)

West Antarctica

Dome C

East Antarctica

905

Vostok

East Antarctica

3350

Law Dome

East Antarctica

1203

Taylor Dome

East Antarctica

375

Dome Fuji

East Antarctica

2500

Dunde

Western China

140

Guliya

Western China

309

Huascarän

Peru

166

Sajama

Bolivia

133

Dasuopu

Western China

168

a Other short cores (<100 m) and surface samples from ice sheet margins in Greenland and Antarctica have also recovered ice from the last glacial period. b See Figures 5.1 and 5.2 for locations.

a Other short cores (<100 m) and surface samples from ice sheet margins in Greenland and Antarctica have also recovered ice from the last glacial period. b See Figures 5.1 and 5.2 for locations.

(stalactites and stalagmites), tree rings, ostracods, and peat (Swart et al., 1993). In this section a brief introduction to the theory behind stable isotope work is provided and applications to ice-core analysis are discussed. The importance of stable isotopes in other branches of paleoclimatic research are dealt with in Chapters 6, 7, and 10.

Water is the most abundant compound on Earth. The primary compound in all forms of life, it is perhaps the most important agent in weathering, erosion, and geological recycling of materials, and, of course, plays a crucial role in the global energy balance. The study of "fossil water," either directly in the form of firn and ice, or indirectly through materials deposited from solution in "fossil water" (e.g., speleo-thems) thus has important implications in many aspects of paleo-environmental reconstruction.

In common with most other naturally occurring elements, the constituents of water, oxygen, and hydrogen may exist in the form of different isotopes. Isotopes result from variations in mass of the atom in each element. Every atomic nucleus is made up of protons and neutrons. The number of protons in the nucleus of an element (the atomic number) is always the same, but the number of neutrons may vary, resulting in different isotopes of the same element. Thus, oxygen atoms (which always have 8 protons) may have 8, 9, or 10 neutrons, resulting in three isotopes with atomic mass numbers of 16, 17, and 18, respectively (160, 170, and lsO). In nature these three stable isotopes occur in relative proportions of 99.76% (160), 0.04% (170), and 0.2% (lsO). Hydrogen has two stable isotopes, JH and 2H (deuterium) with relative proportions of 99.984% and 0.016%, respectively. Consequently, water molecules may exist as any one of nine possible isotopic combinations with mass numbers ranging from 18 (JH2 160) to 22 (2H2 180). However, as water with more than one "heavy" isotope is very rare, generally only four major isotopic combinations are common, and only two are important in paleoclimatic research (1H2H160, generally written as HDO, and 1H2180).

The basis for paleoclimatic interpretations of variations in the stable isotope content of water molecules is that the vapor pressure of H2160 is higher than that of HD160 and H2lsO (10% higher than HDO, 1% higher than H2lsO). Evaporation from a water body thus results in a vapor that is poorer in deuterium and lsO than the initial water; conversely, the remaining water is (relatively speaking) enriched in deuterium and lsO. At equilibrium, for example, atmospheric water vapor contains 10 %o (parts per thousand or per mil) less lsO and 100 %o less deuterium than mean ocean water. When condensation occurs, the lower vapor pressure of HDO and H2lsO results in these two compounds passing from the vapor to the liquid state more readily than water made up of lighter isotopes. Hence, compared to the vapor, the condensation will be enriched in the heavy isotopes (Dansgaard, 1961). Further condensation of the vapor will continue this preferential removal of the heavier isotopes, leaving the vapor more and more depleted in HDO and H2180 (Fig. 5.3). As a result, continued cooling will give rise to condensate with increasingly lower HDO and H2lsO concentrations than when the condensation process first began. The greater the fall in temperature, the more condensation will occur and the lower will be the heavy isotope concentration, relative to the original water f (%)

Rayleigh Analysis

assumed (i.e., an equilibrium Rayleigh condensation process). —, 8 vapor;-, 8 liquid; f = percentage of the original water vapor condensed (Epstein and Sharp, Journal of Geology, 67,© 1959 by the University of Chicago).

assumed (i.e., an equilibrium Rayleigh condensation process). —, 8 vapor;-, 8 liquid; f = percentage of the original water vapor condensed (Epstein and Sharp, Journal of Geology, 67,© 1959 by the University of Chicago).

source (Fig. 5.4). Isotopic concentration in the condensate can thus be considered as a primary function of the temperature at which condensation occurs (subject to certain reservations to be noted in Section 5.2.2).

5.2.1 Stable Isotopes in Water: Measurement and Standardization

In the majority of paleoclimatic studies using stable isotopes, oxygen is generally the element of primary interest, though deuterium is important in ice core research. In oxygen isotope work, the water sample is isotopically exchanged with carbon dioxide of known isotopic composition:

The relative proportions of 160 and lsO in carbon dioxide from the sample are then compared with the isotopic composition of a water standard (Standard Mean

SOUTH POLE

1h218o 1h2h16o Hdo

FIGURE 5.4 Schematic diagram to illustrate isotopic depletion of water vapor en route to the Antarctic ice sheet. As an air mass cools, precipitation produced is preferentially enriched in lsO, leaving the remaining vapor relatively depleted. Consequently, with further condensation, the precipitation contains less and less lsO (i.e., lower 8lsO values).This isobaric effect is accentuated by uplift (adiabatic) effects over the ice sheet itself, so that the lowest delta lsO values are found in the ice-sheet interior (Dansgaard et al., 1971; Robin, 1977).

SOUTH POLE

FIGURE 5.4 Schematic diagram to illustrate isotopic depletion of water vapor en route to the Antarctic ice sheet. As an air mass cools, precipitation produced is preferentially enriched in lsO, leaving the remaining vapor relatively depleted. Consequently, with further condensation, the precipitation contains less and less lsO (i.e., lower 8lsO values).This isobaric effect is accentuated by uplift (adiabatic) effects over the ice sheet itself, so that the lowest delta lsO values are found in the ice-sheet interior (Dansgaard et al., 1971; Robin, 1977).

Ocean Water or SMOW11) and the results expressed as a departure (8lsO) from this standard, thus

11 In order that isotopic analyses in different laboratories be comparable, a universally accepted standard is used, known as SMOW (Standard Mean Ocean Water; Craig, 1961b). This is not an actual oceanic water sample, but is based on a US National Institute of Standards and Technology distilled water sample (NIST-1). However, the zero point on the SMOW scale has been adjusted so that it is more or less equivalent (-0.1%o) to the isotopic composition of real ocean water (measured in samples from depths of 200500 m in the Atlantic, Pacific, and Indian Oceans; Epstein and Mayeda, 1953). Isotopic studies based on carbonate fossils use as a standard a Cretaceous belemnite from the Peedee Formation of North Carolina (PDB-1). Carbon dioxide released from PDB-1 = + 0.2%o relative to C02 equilibrated with SMOW (Craig, 1961b). Recent updates to international reference standards are described by Coplen (1996).

All measurements are made using a mass spectrometer and reproducibility of results within ± 0.1% is generally possible.

A 8lsO value of -10 therefore indicates a sample with an 180/160 ratio 1% or 10%o less than SMOW. Under our present climate, the lowest 8lsO value recorded in natural waters is —58%o (-454%o in 8D) in snow from the highest and most remote parts of Antarctica (Qin et al., 1994).

5.2.2 Oxygen-18 Concentration in Atmospheric Precipitation12

In Section 5.2.1, and Fig. 5.3, the isotopic composition of water in equilibrium with water vapor was considered. In reality, we cannot consider the process to be always at equilibrium between vapor and condensate, nor can the process be considered to occur in isolation. Exchanges between atmospheric water vapor, water droplets in the air, and water at the surface (which may be isotopically "light") do occur continuously, so this complicates any simple temperature-isotope effect that we might expect to find (Koerner and Russell, 1979). There are also kinetic effects on fractionation that occur during evaporation and condensation, and the latter can be especially important at very low temperatures (see Section 5.2.5). Overall, the lsO content of precipitation depends on

(a) the lsO content of the water vapor at the start of condensation (this could be very low if evaporation occurred over an inland lake or ice body where lsO concentrations are less than mean ocean water);

(b) the amount of moisture in the air compared to its initial moisture content;

(c) the degree to which water droplets undergo evaporation en route to the ground and whether any of this re-evaporated vapor re-enters the precipitating air mass (Ambach et al, 1968);

(d) the temperature at which the evaporation and condensation processes take place; and

(e) the extent to which clouds become supersaturated, with respect to ice, at very low temperatures.

In spite of these complications, empirical studies have demonstrated that geographical and temporal variations in isotopes do occur, reflecting temperature effects due to changing latitude, altitude, distance from moisture source, season, and long-term climatic fluctuations (Dansgaard et al., 1973; Koerner and Russell, 1979; Petit et al., 1991). As any interpretation of ice-core isotopic records is rooted in an evaluation of these factors, it is important to consider them in more detail.

5.2.3 Geographical Factors Affecting Stable Isotope Concentrations

For the last 30 years or so, precipitation samples from many locations throughout the world have been analyzed for their 8180 content (Rozanski et al., 1992, 1993). Figure 5.5 shows that 8lsO values of January and July precipitation generally reflect the distribution of temperature, decreasing at higher latitudes and at higher elevations

12 In this section, 8I80 is generally referred to, but the same principles apply to variations in 8D.

FIGURE 5.5 Mean 8lsO in January (upper figure) and July precipitation (lower figure) collected at precipitation stations throughout the world, based on analyses by the International Atomic Energy Authority over the last few decades (Lawrence and White, 1991).

(e.g., in the high interior parts of Greenland and in the Andes). The influence of the Gulf Stream is also apparent on the January map (Lawrence and White, 1991). Changes in temperature from winter to summer are also reflected in 8lsO, and this leads to an annual cycle in 8lsO of snowfall that can be used to count annual accumulation layers in ice cores (see Section 5.3.2).

Figure 5.5 indicates a strong latitudinal influence on 8lsO. Lower 8lsO values are found at higher latitudes as a result of the loss of heavy isotopes in water condensed en route to those regions. This is sometimes referred to as an isobaric effect, implying a systematic change brought about by overall cooling at a particular level in the atmosphere, rather than cooling brought about by a change in elevation (adiabatic cooling). With increasing elevation, adiabatic cooling of the precipitating air mass leads to precipitation that is more and more depleted in 180 due to preferential removal of the heavier isotope in the condensation process. For example, the Quelccaya Ice Cap in Peru receives moisture that has undergone adiabatic cooling as the air rises over the Andes from the Amazon Basin. This uplift reduces 8lsO in precipitation by ~ll%o (Grootes et al, 1989). On large ice sheets, the adiabatic effect is superimposed on a "distance from moisture source" factor that results in lower 8lsO concentrations as the distance from oceanic moisture sources increases (Koerner, 1979). Hence, at high elevations in central Antarctica, thousands of kilometers from the southern oceans, atmospheric precipitation has the lowest heavy isotope concentrations of any natural water occurring today (Morgan, 1982; Qin et al, 1994).

These different influences on 8lsO in precipitation lead to the geographical patterns seen in Fig. 5.5. There is a very strong correspondence between temperature and 8180 in extra-tropical regions (Fig. 5.6). The overall 8lsO-surface temperature

Present Climate (observed)

lllllill'ilvl i'

0 0

Post a comment