Stable isotope variations in ice cores

The atoms of oxygen (O) and hydrogen (H) present in glacier ice occur in different isotopes. Isotopes are different forms of an element that result from variations in atomic mass, or the combined number of protons and neutrons in each atomic nucleus. The number of protons in atoms of each element is constant. Mass variations are therefore a result of variations in the number of neutrons in the atom. Oxygen atoms have eight protons, but may have eight, nine or ten neutrons, giving three different isotopes with atomic

Figure 3.1 Variations in surface water oxygen isotope ratios during glacial maxima (low sea-level) and interglacials (maximum sea-level). (Adapted from Lowe and Walker, 1997)

Glacial

Glacial maximum; /^/X

'/, expanded ice sheets ^//a fi,fiO - -3D%o '///A

MSL

\ ICa. Iï0m

\ 15L

Continental land mas;

\ SlaO ocean + I.S%o

Interglacial

Ice sheets

reduced in siie OVy

MSL

Continental land mass \

5ieO ocean

\ ~0.0%o

masses of 16 (160), 17 (170) and 18 (180). Hydrogen atoms have one proton, but may have no or one neutron, giving two isotopes (XH and 2H), the latter also known as deuterium (D; 0.016 per cent). Water molecules may therefore consist of any of nine possible combinations of these five isotopes. Three of these combinations are, however, common: 1H2160, WO, and 1HDlsO. The relative amount of oxygen isotopes in nature is 99.76 per cent lsO, 0.04 per cent 170 and 0.2 per cent 160.

The isotopic composition of precipitation falling on a glacier or ice sheet depends on the history of evaporation and condensation in the hydrological cycle. During the evaporation process, water molecules consisting of light isotopes turn to vapour more easily than those composed of heavy isotopes. This pro cess is called fractionation. The resulting vapour is relatively enriched in :H and 160. As condensation proceeds, more of the remaining heavy isotopes will be removed, the vapour becoming more and more depleted in 18Ô and D. Cooling of water vapour as it rises in the atmosphere and/or is transported inland over ice sheets will result in precipitation with increasingly lighter isotopic composition. As a result, the isotopic composition of the precipitation reflects the temperature when the precipitation occurred (Fig. 3.1). Measurements have demonstrated that there is a high correlation between temperature and oxygen isotope composition (Fig. 3.2), with a calibration of 0.33 per mil "C"1 (Cuffey et al, 1995). Despite the fact that temperature is not the only factor determining the oxygen isotope

Temperature ("C)

Figure 3.2 Isotopic composition of snow versus local annual mean surface temperature. (Modified from Jouzel et al, 1997)

Temperature ("C)

Figure 3.2 Isotopic composition of snow versus local annual mean surface temperature. (Modified from Jouzel et al, 1997)

composition, the isotopic composition of glacier ice has been used for temperature reconstructions. Jouzel et al. (1997) reviewed the empirical temporal slopes (curve gradients) in isotope models from polar regions. They found that the temporal slopes were lower than modern slopes, the difference most probably due to changes in the evaporative origins of moisture, changes in the seasonality of the precipitation, changes in the strength of the inversion layer, or a combination of these factors. Despite problems with calibrating an isotope palaeothermometer, the use of isotopes as a temperature proxy seems justified.

The oxygen isotope composition in ice cores, measured by mass spectrometry, is given as deviations (<5180) from the Standard Mean Ocean Water (SMOW). Atmospheric water becomes depleted (average about 10 per cent) in the heavier lsO isotope during evaporation from the ocean surface. There is, however, a seasonal variation of about 15%o in the isotopic composition. Seasonal variations can therefore be detected by precise oxygen isotope meas urements. Due to diffusion effects, the amplitude of the <5lsO signal decreases with depth, but significant variations can still be detected back to 160,000 yr bp.

Hydrogen isotopes act in the same manner as oxygen isotopes. The hydrogen/ deuterium ratio in the atmospheric water (snow) is determined by saturation vapour pressure and molecular diffusity in air. A deuterium profile from the Vostok ice core shows similar variations as in the Greenland oxygen isotope records (Jouzel et al, 1990).

Downcore variations in stable isotope content are used to reconstruct the pattern and amplitude of climate variations. Figure 3.3 shows the oxygen isotope record from the GRIP core in Greenland and a deuterium profile from the Vostok core in Antarctica. Since isotopic fractionation is temperature dependent, the fluctuations mainly reflect global temperature variations. The curves are similar to those obtained from deep ocean cores. Both the Eemian and the Holocene are easily detected in the records.

Summit Greenland (GRIP)

Vostock Antarctica

3 140

Summit Greenland (GRIP)

Vostock Antarctica

3 140

-450

Figure 3.3 Stable oxygen isotope variations during the last 160,000 years recorded in the GRIP ice core, and deuterium ratios in the Vostok core from Antarctica. (Adapted from Lowe and Walker, 1997)

-450

Figure 3.3 Stable oxygen isotope variations during the last 160,000 years recorded in the GRIP ice core, and deuterium ratios in the Vostok core from Antarctica. (Adapted from Lowe and Walker, 1997)

In the GRIP and GISP2 data records there are numerous, high-frequency <5180 oscillations postdating the Eemian interglacial. Between 80,000 and 20,000 yr bp, some 20 interstadial events are recorded. These are interpreted to reflect abrupt temperature changes of the order of 5-8°C. These so-called

Dansgaard-Oeschger events lasted for about 500-2000 years and therefore cannot be explained by orbital forcing mechanisms. Instead, they are interpreted to reflect feedback mechanisms involving ice sheet/glacier fluctuations, variations in the ocean system and atmospheric circulation fluctuations.

'I I"" I" I I"" I mi |ii I I ■ » 'I "I I I'" 'I '"I'

1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000

Cal. bp

'I I"" I" I I"" I mi |ii I I ■ » 'I "I I I'" 'I '"I'

1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000

Cal. bp

'I ■ ■ I■ ■ I III I ||| HIM M|M Mill III up

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000

Cal. bp

Figure 3.4 Late-glacial and Holocene oxygen isotope variations in the GRIP (upper panel) 0ohnsen et ah, 1997) and GISP2 (lower panel) ice cores (Stuiver el ah, 1995).

'I ■ ■ I■ ■ I III I ||| HIM M|M Mill III up

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000

Cal. bp

Figure 3.4 Late-glacial and Holocene oxygen isotope variations in the GRIP (upper panel) 0ohnsen et ah, 1997) and GISP2 (lower panel) ice cores (Stuiver el ah, 1995).

Obtaining temperature series from stable isotope records from ice cores is not straightforward. This is especially the case for older parts of the record, because of glacier deformation and flow of ice from other regions. Different modern isotope values between the source area and the core site could give iso-topic variations unrelated to real temperature variations. The GRIP and GISP2 cores were drilled close to the summit of the Greenland ice sheet, and if the ice divide has not moved significantly during the last interglacial/glacial cycle, ice flow has been minimal.

The relationship between oxygen isotopic ratios in precipitation and climatic conditions is also difficult to quantify (Lorius et ah, 1989). Several approaches have been proposed, including comparison of isotopic values in seasonal snow/ice with meteorological data (Dansgaard et ah, 1975), use of statistical methods to correlate borehole temperature records and isotope data, and constructing calibration models for oxygen isotope ratios to palaeotemperature (Cuffey et al, 1992).

Interpretation of the isotopic signal in the GISP2 ice core indicates that the site is influenced by both the Icelandic Low to the SE and the Davis Strait/Baffin Bay storms to the SW and W (Barlow et al, 1997). The GISP2 isotope signal is influenced by the North Atlantic Oscillation: the seesaw in winter temperatures between west Greenland and northern Europe.

Oxygen isotope profiles from the GISP2 summit area show rapid smoothing of the 180/160 signal near the surface. Below a depth of about 2 m the smoothed ilsO signal is fairly well preserved, interpreted to reflect average local weather conditions; the longer climate variations also have regional and global significance (Grootes and Stuiver, 1997). Between approximately 75,000 and 11,650 yr bp (the Younger Dryas/Preboreal transition) the oxygen isotope record is characterized by frequent, rapid switches between intermediate interstadial and low stadial values. Spectral analysis of the variations superimposed on the orbitally induced changes yields significant periodicities of 1500 and 4000 years. Similar fluctuations as recorded in the oxygen isotope signal in the GISP2 ice core have also been found in other climate records, strongly suggesting that the GISP2 oxygen isotope signal is the local expression of more regional and worldwide climate events. Meltwater from ice sheets adjacent to the North Atlantic influenced ocean circulation during the B0lling-Aller0d-Younger Dryas complex of interstadials and stadials. The Holocene is characterized by relatively stable mean isotopic values (Fig. 3.4), however, with dominant 6.3, 11 and 210 year oscillations. The latter two are also recognized in the solar-modulated records of the cosmogenic isotopes 10Be and 14C, indicating that variations in solar irradiance is the main cause of these periodicities. Cooling by volcanic eruptions is recorded in the oxygen isotope signal; the effects, however, are small and volcanic eruptions are considered not to trigger large climate variations (Grootes and Stuiver, 1997).

The INTIMATE (INTegragion of Ice-core, MArine and TErrestrial records) group proposed that the GRIP ice core in Greenland be designated a stratotype for the Last Termination period (ca. 22,000-11,500 yr bp), and that the oxygen isotope profile be used as the basis for an event stratigraphy (Fig. 3.5) divided into stadials and interstadials according to their isotopic variations (Björck et al, 1998).

Was this article helpful?

0 0

Post a comment