Stable Isotopes in Paleoclimatology Introduction

Lecture 25 Spring 2009

STABLE ISOTOPES IN PALEOCLIMATOLOGY INTRODUCTION

At least since the classic work of Louis Agassiz in 1840, geologists have contemplated the question of how the Earth’s climate might have varied in the past. But until 1947, they had no means of quan- tifying paleotemperature changes. In that year, Harold Urey initiated the field of stable isotope geochemistry. In his classic paper “The thermodynamic properties of isotopic substances”, Urey calculated the temperature dependence of oxygen isotope fractionation between calcium carbonate and water and proposed that the isotopic composition of carbonates could be used as a paleothermometer (Urey, 1947). Urey’s postdoctoral associate Samuel Epstein and several students tested Urey’s idea by growing molluscs in water of various temperatures (Epstein et al., 1953). They found the following empirical relationship: 18 18 0.5 ∆ = δ Ocal – δ Owater = 15.36 – 2.673 (16.52 + T) 25.1 This equation was in good, though not exact, agreement with the theoretical prediction of Urey (Figure 25.1). Thus was born the modern field of paleoclimatology. THE RECORD OF CLIMATE CHANGE IN DEEP SEA SEDIMENTS It is perhaps ironic that while glaciers are a continental phenomenon, our best record of them is from the oceans. In part, this is because each period of continental glaciation destroys the record of the previous one. In contrast, deep-sea sediments are generally not disturbed by glaciation. Thus while much was learned by studying the effects of Pleistocene glaciation in North America and Europe, much was left unresolved: questions such as the precise chronology, cause, temperatures, and ice volumes (ice area could of course be determined, but this is only part of the problem). The question of chronology has been largely resolved through isotopic studies of deep-sea biogenic sediments, and great progress has been made toward resolution of the remaining questions. The principles involved in paleoclimatology are simple. As Urey formulated it, the isotopic composition of calcite secreted by organisms should provide a record of paleo-ocean temperatures because the fractionation of oxygen isotopes between carbonate and water is temperature dependent. In actual practice, the problem is somewhat more complex because the isotopic composition of the test of an organism will depend not only on temperature, but also on the isotopic composition of water in which the organism grew, vital effects (i.e., dif- Figure 25.1. Fractionation of oxygen isotopes be- ferent species may fractionate oxygen isotopes tween calcium carbonate and water as a function of somewhat differently), and post-burial isotopic temperature for biologically precipitated calcite exchange with sediment pore water. As it (molluscs; Epstein et al., Craig, 1965), biologically turns out, the latter two are usually not very precipitated aragonite, and abiologically precipi- important for carbonates (at least for late Terti- tated calcite. Also shown are the calculated frac- ary/Quaternary sediments), but the former is. tionation factors of Urey (1947) for 0° C and 25°C.

295 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

The Quaternary δ 18O Record The first isotopic work on deep-sea sediment cores with the goal of reconstructing the temperature history of Pleistocene glaciations was by Emiliani (1955), who was then a student of Urey at the Uni- versity of Chicago. Emiliani analyzed δ18O in foraminifera from piston cores from the world ocean. Remarkably, many of Emiliani’s findings are still valid today. He concluded that the last glacial cycled had ended about 16,000 years ago, and found that temperature increased steadily between that time and about 6000 years ago. He also recognized 14 other glacial–interglacial cycles over the last 600,000 years, and found that these were global events, with notable cooling even in low latitudes. He concluded that bottom water in the Atlantic was 2° C cooler, but that bottom water in the Pacific was only 0.8° C cooler during glacial periods. He also concluded that the fundamental driving force for Quater- nary climate cycles was variations in the Earth’s orbital parameters. Emiliani had the field of oxygen isotope paleoclimatology virtually to himself until about 1970. In retrospect, it is remarkable how much Emiliani got right. By that time, others saw the value of this approach and got into the act. Their work resulted in significant modifications to some of Emiliani’s con- clusions. One of the main improvements was simply refining the time scale using paleomagnetic stratigraphy and, later, some of the geochronological tools we discussed earlier in the course (10Be, Th isotopes, etc.). In his initial work, Emiliani had only 14C dating available to him, and he dated older sections simply by extrapolating sedimentation rates based on 14C dating. Another important modification to Emiliani’s work was a revision of the temperature scale. Emiliani had realized that the isotopic composition of the ocean would vary between glacial and interglacial times as isotopically light water was stored in glaciers, thus enriching the oceans in 18O. Assuming a δ18O value of about –15‰ for glacial ice, Emiliani estimated that this factor accounted for about 20% of the observed variations. The remainder he 18 attributed to the effect of temperature on ! O isotope fractionation. However, Shackle- 2.0 1.0 0 ton and Opdyke (1973) argued that storage Interglacial of isotopically light water in glacial ice was 18 actually the main effect causing oxygen iso- ! O = –30 topic variations in biogenic carbonates, and that the temperature effect was only sec- ondary. Their argument was based on the !18O = 0.0 observation that nearly the same isotopic variations occurred in both planktonic (surface-dwelling) and benthic (bottom dwelling) foraminifera. Because of the way in which the deep water of the ocean is formed and circulates, Shackleton and Op- Glacial dyke argued that deep-water temperature should not vary between glacial and inter- !18O glacial cycles. Analyzing tests of both ben- = –30 thic and planktonic organisms allowed a better calculation of temperature changes. The isotopic composition of tests of benthic !18O = 1.5 organisms, i.e., those growing in deep water, could be used to determine the change in seawater isotopic composition. This would allow a more precise calculation of surface water temperature change from the Figure 25.2. Cartoon illustrating how δ18O of the ocean isotopic composition of planktonic tests changes between glacial and interglacial periods.

296 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

(shells of organisms growing in surface water). Nevertheless, the question of just how much of the variation is deep-sea carbonate sediments is due to ice build-up and how much is due to the effect of temperature on fractionation continued to be de- bated. After Shackleton and Opdyke’s work, climate modeling suggested deep water temperatures may indeed vary, though probably not as much as Emiliani had calculated. It is now clear that the average δ18O of glacial ice is less than –15‰, as Emiliani had assumed. Typical values for Greenland ice are –30 to -40‰ (relative to SMOW) and as much as -50‰ for Antarctic ice. If the exact isotopic composition of ice and the ice volume were known, it would be a straightforward exercise to calculate the effect of continental ice build up on ocean isotopic composition. For example, the present volume of continental ice is 27.5 × 106 km3, while the volume of the oceans is 1350 × 106 km3. Assuming glacial ice has a mean δ18O of –30‰ relative to SMOW, we can calculate the δ18O of the hydrosphere as –0.6‰ (neglecting freshwater reservoirs, which are small). At the height of the Wisconsin Glaciation, the volume of glacial ice is thought to have increased by 42 × 106 km3, corresponding to a lowering of sea level by 125 m. If the δ18O of ice was the same then as now (–-30‰), we can readily calculate that the δ18O of the ocean would have increased by 1.59‰. This is illustrated in Figure 25.2. To see how much this affects estimated temperature changes, we can use Craig’s* (1965) revision of the Ep- stein calcite-water geothermometer: 2 T°C = 16.9 – 4.2 ∆cal-water + 0.13∆ cal − water 25.2 According to this equation, the fractionation should be 33‰ at 20° C. At 14°C, the fractionation is 31.5‰. If a glacial foram shell were 2‰ lighter, Emiliani would have made a correction of 0.5‰ for the change in oxygen isotopic composition of seawater and attributed the remainder of the difference, 1.5‰, to temperature. He would have concluded that the ocean was 6˚ C cooler. However, if the change in the isotopic composition of seawater is actually 1.5‰, leaving only a 0.5‰ difference due to temperature, the calculated temperature difference is only about 2° C. Thus the question of the volume of glacial ice, and its isotopic composition must be resolved before δ18O in deep-sea carbonates can be used to calculate paleotemperatures. It is now generally assumed that the δ18O of the ocean changed by 1.5‰ between glacial and interglacial periods, but second order local variations also occur (due to evaporation and precipitation), leaving some uncertainty in exact temperatures. Comparison of Figure 25.3. A. “Stacking of five cores” se- sealevel curves derived from dating of terraces and coral 18 lected by Imbrie et al. (1984). Because the reefs indicate that each 0.011‰ variation in δ O repre- absolute value of δ18O varies in from core to sents a 1 m change in sealevel. core, the variation is shown in standard de- By now, thousands of deep-sea cores have been ana- viation units. B. Smoothed average of the lyzed for oxygen isotope ratios. Though most reveal the 18 five cores in A. After Imbrie et al. (1984). same general picture, the δ O curve varies from core to core. In addition to the changing isotopic composition of

* Harmon Craig was also a student of Harold Urey.

297 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

the ocean, the δ18O in a given core will depend on several other factors. (1) The temperature in which the organisms grew. (2) The faunal assemblage, as the exact fractionation will vary from organism to organism. For this reason, δ18O analyses are often performed on a single species. However, these “vital effects” are usually small, at least for planktonic foraminifera. (3) Local variations in water isotopic composition. This is important in the Gulf of Mexico, for example. Melt-water released at the end (termination) of glacial stages flooded the surface of the Gulf of Mexico with enough isotopically light melt-water to significantly change its isotopic composition relative to the ocean as a whole. (4) Sedi- mentation rate varies from core to core, so that δ18O as a function of depth in the core will differ between cores. Changes in sedimentation rate at a give locality will distort the appearance of the δ18O curve. (5) Bioturbation, i.e., burrowing activity of seafloor animals, which may smear the record. The Cause of Quaternary Glaciations For these reasons, correlating from core to core and can sometimes be difficult and constructing a “standard” δ18O record is a non-trivial task. Nevertheless, it is the first step in understanding the global climate change signal. Figure 25.3 shows the global δ18O record constructed by averaging 5 cores (Im- brie, et al., 1984). A careful examination of the global curve shows a periodicity of approximately 100,000 years. The same periodicity was apparent in Emiliani’s initial work and led him to conclude that the glacial-interglacial cycles were due to variations in the Earth’s orbital parameters. These are often referred to as the Milankovitch cycles, after M. Milankovitch, a Serbian astronomer who argued they caused the ice ages in the early part of the twentieth century†. Three parameters describe these variations: e: eccentricity, ε: obliquity (tilt), and precession: e sin ω, where ω is the longitude of perihelion (perihelion is the Earth’s closest approach to the Sun). The eccentricity (i.e., the degree to which the orbit differs from circular) of the Earth’s orbit about the Sun, and the degree of tilt, or obliquity, of Obliquity the Earth’s rotational axis vary slightly. Pre- Eccentricity 21.5° cession refers to the change in the direction in 24.5° which the Earth’s rotational axis tilts when it is closest to the Sun (perihelion). These Sun Earth variations, which are illustrated in Figure 25.4, affect the pattern of solar radiation, or insolation, that the Earth receives. Changes in Precesion these parameters have negligible effect on the total insolation, but they do affect the pattern ! of insolation. For example, tilt of the rotational axis determines seasonality, and the sin ! = 0 sin ! = 1 latitudinal gradient of insolation. It is this Figure 25.4. Cartoon illustrating the “Milankovitch pa- gradient that drives atmospheric and oceanic rameters”. The eccentricity is the degree the Earth’s or- circulation. If the tilt is small, seasonality bit departs from circular. Obliquity is the tilt of the will be reduced (cooler summers, warmer Earth’s rotation axis with respect to the plane of the winters), and the average annual insolation ecliptic. Obliquity varies between 21.5° and 24.5°. Pre- gradient will be high. Precession relative to cision is the variation in the direction of tilt at the Earth’s the eccentricity of the Earth’s orbit also af- closest approach to the Sun (perihelion). The parameter fects seasonality. For example, the Earth ω is the angle between the Earth’s position on June 21 presently is closest to the Sun in January. As (summer solstice), and perihelion.

† While Milankovitch was a strong and early proponent of the idea that variations in the Earth’s orbit caused ice ages, he was not the first to suggest it. J. Croll of Britain first suggested it in 1864, and published several subsequent papers on the subject.

298 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

a result, northern hemisphere winters (and southern hemisphere summers) are somewhat milder than they would be otherwise. For a given latitude and season, precession will result in a ±5% difference in insolation. While the Earth’s orbit is only slightly elliptical, and variations in eccentricity are small, these variations are magnified because insolation varies with the inverse square of the Earth-Sun dis- tance. Variation in tilt approximates a simple sinusoidal function with a period of 41,000 yrs. Variations in eccentricity can be approximately described with characteristic period of 100,000 years. In actuality variation in eccentricity is more complex, and is more accurately described with periods of 123,000 yrs, 85,000 yrs, and 58,000 yrs. Similarly, variation in precession has characteristic periods of 23,000 and 18,000 yrs. While Emiliani suspected δ18O variations were related to variations these “Milankovitch” parameters, the first quantitative approach to the problem was that of Hayes et al. (1976). They applied Fourier analysis to the δ18O curve, a mathematical tool that transforms a complex variation such as that in Fig- ure 25.3 to the sum of a series of simple sine functions. Hayes et al. then used spectral analysis to show that much of the spectral power of the δ18O curve occurred at frequencies similar to those of the Milankovitch parameters. By far the most elegant and convincing treatment, however, is that of Imbrie (1985). Imbrie’s treatment involved several refinements and extension of the earlier work of Hayes et al. (1976). First, he used improved values for Milankovitch frequencies. Second, he noted these Mi- lankovitch parameters might vary with time, as might the climate system’s response to them. The Earth’s orbit and tilt are affected by the gravitational field of the Moon and other planets. In addition, other astronomical events, such as bol- ide impacts, can affect them. Thus Imbrie treated the first and second 400,000 years of Figure 25.3 separately. The power spectrum for these two parts of the δ18O curve is shown in Figure 25.5. Imbrie observed that climate does not respond instanta- neously to forcing. For example, maximum temperatures are not reached in Ithaca until mid or late July, 3 to 4 weeks after the maximum insolation, which occurs on June 21. Thus there is a phase difference between the forcing function (insolation) and climatic response (temperature). Imbrie also pointed out that the climate might respond differently to different forcing functions. As an example, he used tem- Figure 25.5. Power spectrum of the com- perature variations in the Indian Ocean, which respond both posite δ18O curve shown in Figure 25.3 as to annual changes in insolation and to semiannual changes a function of frequency. Peaks in the in ocean upwelling. The response to these two forcing func- spectrum correspond with the frequen- tions differs in different localities. The extent to which cli- cies of the variations of the Milankovitch mate responds to a particular forcing function is the gain. parameters. The residual spectrum The phase lag may also differ from locality to locality. shows the variance remaining after sub- Mathematically, the climatic response can be expressed as: tracting a phase and gain model based on y = g1(x1 – φ1) + g2(x2 – φ2) 25.3 the Milankovitch parameters. The upper where y is the climatic response (temperature) x and x are figure shows the power spectrum for 0- 1 2 the two forcing functions (insolation and upwelling), g and 400 kyr BP, the lower figure for the pe- 1 g are the gains associated with them and φ and φ are the riod 400-782 kyr. After Imbrie (1985). 2 1 2

299 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

TABLE 25.1 GAIN AND PHASE MODEL OF IMBRIE phase lags). (1985) Imbrie (1985) constructed a model for response of global climate (as measured Frequency σy σx k g φ by the δ18O curve) in which each of the 6 band (m) (u) (m/u) ka Milankovitch forcing functions was associated with a different gain and phase. e123 19.5 0.167 0.58 68 –5 The values of gain and phase for each e95 19.3 0.250 0.83 64 -3 parameter were found statistically by e59 12.2 0.033 0.79 292 -12 minimizing the residuals of the power ε 15.0 0.394 0.92 35 -9 spectrum (Figure 25.5). Table 25.1 gives the essential parameters of the model. σ p23 13.0 0.297 0.95 42 -6 x is the strength of each forcing function, p18 5.3 0.154 0.81 28 -3 and σy is the strength of the response (given in meters of sealevel reduction), k

is the coefficient of coherency, g is the gain (σy/σx), and φ is the phase difference between input function and the climatic response. The resulting model is shown in comparison with the data for the past 400,000 years and the next 25,000 years in Figure 25.6. The model has a correlation coefficient, r, of 0.88 with the data. Thus about r2, or 77%, of the variation in δ18O, and therefore presumably in ice volume, can be explained Imbrie’s Milankovitch model. The correlation for the period 400,000–782,000 yrs is somewhat poorer, around 0.80, but nevertheless impressive. Since variations in the Earth’s orbital parameters do not affect the average annual insolation the Earth receives, but only its pattern in space and time, one might ask how this could cause glaciation. The key factor seems to be the insolation received during summer by high northern latitudes. This is, of course, the area where large continental ice sheets develop. The southern hemisphere, except for Antarctica, is largely ocean, and therefore not subject to glaciation. Glaciers apparently develop when summers are not Figure 25.6. Gain and phase model of Imbrie warm enough to melt the winter’s accumulation of relating variations in eccentricity, tilt, and snow. precession to the oxygen isotope curve. Top Nevertheless, the total variation in insolation is shows the variation in these parameters over small, and not enough by itself to cause the climatic the past 400,000 and next 25,000 years. Bot- variations observed. Apparently, there are feedback tom shows the sum of these functions with mechanisms at work that serve to amplify the funda- appropriated gains and phases applied and mental Milankovitch forcing function. One of these compares them with the observed data. Af- feedback mechanisms was identified by Agassiz, and ter Imbrie (1985). that is ice albedo, or reflectance. Snow and ice reflect much of the incoming sunlight back into space. Thus as glaciers advance, they will cause further cooling. Any additional accumulation of ice in Antarctica, however, does not result in increased albedo, because the continent is fully ice covered even in non- glacial time, hence the dominant role of northern hemisphere insolation in driving climate cycles. What other feedback mechanisms might be at work is still a matter of much speculation. Isotope geo-

300 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

chemistry provides some interesting insights into 2 of these possible feedback mechanisms, carbon di- oxide and ocean circulation, which we discuss below. CARBON ISOTOPES, OCEAN CIRCULATION, AND CLIMATE In the previous lecture, we noted the need for feedback mechanisms to amply the Milankovitch signal, which is the primary driving force of Quaternary climate oscillations. We observed that the Mi- lankovitch variations change only the distribution of solar energy received by the Earth, not the total amount. If this were the only factor in climate change, we would expect that the glaciation in the southern and northern hemispheres would be exactly out of phase. This, however, is not the case. Thus there must be feedback mechanisms at work capable of producing globally synchronous climate variation. Broecker (Broecker, 1984 and subsequent papers) argued that one of these was the deep circulation of the ocean. The role of surface ocean in climate is well understood: for example, the south-flowing California Current keeps the West Coast of the U.S. relatively dry and maintains more moderate temperatures in coastal regions than they would otherwise be. The role of the deep, or thermohaline, circulation of the oceans is less obvious, but perhaps no less important. Whereas the surface ocean circulation is wind- driven, the deep circulation is driven by density, which is in turn controlled by temperature and salinity. In the present ocean, most deep ocean water masses “form” in high latitudes. Once these deep-water masses form, they do not return to the surface for nearly a thousand years. The principal site of deep- water formation is the Southern Ocean where the Antarctic Intermediate Water (AAIW) is formed in the Antarctic Convergence and Antarctic Bottom Water (AABW), the densest of ocean water masses, is formed in the Weddell Sea. A lesser amount of deep water is also formed in the North Atlantic during winter around Iceland; this water mass is called North Atlantic Deep Water (NADW). The only “deep water” formed at intermediate latitudes is Mediterranean Intermediate Water (MIW), which sinks as a result of evaporation in the Mediterranean increasing salinity and hence density. MIW, however, is a smaller water masses than the others, and furthermore has only intermediate density. Formation of deep water thus usually involves loss of thermal energy by the ocean to the atmosphere. Therefore, the present thermohaline circulation of the oceans keeps high latitude climates milder than they would otherwise be. In particular, energy extracted from the Atlantic Ocean water in the formation of NADW keeps the European climate relatively mild. We saw in Lecture 20 that δ13C is lower in deep water than in surface water (Figure 20.11). This results from biological cycling: photosynthesis in the surface waters discriminates against 13C, leaving the dissolved inorganic carbon of surface waters with high δ13C, while oxidation of falling organic particles rich in 12C lowers δ13C in deep water: in effect, 12C is “pumped” from surface to deep water more effi- ciently than 13C. δ13C values in the deep water are not uniform, varying with the “age” of the deep water: the longer the time since the water was at the surface, the more enriched it becomes in 12C and the 13 13 lower the δ C. Since this is also true of total inorganic carbon and nutrients such as PO4 and NO3, δ C 13 correlates negatively with nutrient and ΣCO2 concentrations. NADW has high δ C because it contains water that was recently at the surface (and hence depleted in 12C by photosynthesis). Deep water is not formed in either the Pacific or the Indian Oceans; all deep waters in those oceans flow in from the Southern Ocean. Hence deep water in the Pacific, being rather “old” has low δ13C. AABW is a mixture of young NADW, which therefore has comparably high δ13C, and recirculated Pacific deep water and hence has lower δ13C than NADW. Thus these water masses can be distinguished on the basis of δ13C. Examining δ13C in benthic foraminifera in cores from a variety of locations, Oppo and Fairbanks (1987) concluded that production of NADW was lower during the last glacial maximum and increased to present levels in the interval between 15000 and 5000 years ago. Figure 25.7 shows an example of data from core RC13-229, located in the South Atlantic. δ13C values decrease as δ18O increases. As we saw in the previous lecture, δ18O in marine carbonates is a measure of glacial ice volume and climate.

301 4/22/09 Geol. 656 Isotope Geochemistry

. Lecture 25 Spring 2009

4.5 As the climate warmed at the end of the last glacial interval, δ13C values in bottom water in the

4 more South Atlantic increased, reflecting an increase in the proportion of NADW relative to AABW 3.5 in this region. From δ13C variations in Mediter- ranean and Central Atlantic cores, Oppo and δ18O 3 Fairbanks (1987) also concluded that the pro- Ice Volume duction of MIW was greater during the last gla- 2.5 cial maximum. Thus the mode of ocean circula-

less tion apparently changes between glacial and 2 interglacial times; this change may well amplify 0.6 the Milankovitch signal. 0.4 Further evidence for the role of North Atlantic more 0.2 δ13C 0 -0.2 NADW -0.4 less -0.6 0 5 10 15 20 25

Age, ka Figure 25.7. Variation in δ18O and δ13C in benthic foraminfera from core RC13-229 from the eastern South Atlantic. δ13C data suggest the proportion of NADW in this region increased as the climate warmed. Data from Oppo and Fairbanks (1987). Deep Water was provided by a 1992 study by Charles and Fairbanks. Working with a high-resolution core (i.e., high sedimentation rate) from the Southern Ocean, they found δ13C increased dramatically just over 12,000 years ago (Figure 25.8), indicating a greatly increased flux of NADW to the region. The increase in NADW occurred just as ice volume was beginning to decrease Figure 25.8. Calculated northern hemi- rapidly (as judged from increasing sealevel in Barba- sphere insolation changes (due to orbital dos). The rather sudden melting of the continental ice changes), glacial melt water discharge caps is hard to explain by the slowly increasing North- (calculated from sealevel rise rates de- ern Hemisphere insolation at that time. If production 14 termined from C dating of fossil coral of NADW suddenly started, it would have warmed the 13 reefs off Barbados), and δ C in benthic North Atlantic climate, accelerating melting of glaciers. foraminfera in core RC11-83 (41°36’ S, Thus NADW production may represent a positive 94° 48’ E) from the Southern Ocean. The feedback amplifying the primary “Milankovitch” sig- 13 δ C data are a measure of the propor- nal. However, why NADW production should sud- tion of NADW in Circumpolar Deep denly increase dramatically remains a mystery. Water. European pollen assemblages are also shown. From Charles and Fair- banks (1992).

302 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

THE TERTIARY MARINE δ18O RECORD Imbrie’s (1985) analysis suggests that the climate system’s response to Milankovitch forcing has changed significantly even over the last 800,000 years. The present glacial- interglacial cycles began only 2 million years ago, yet the Mi- lankovitch forcing must have been present before that. This suggests the climate system’s response has changed even more drastically over the past few million years. In addition to ocean circulation and at- 18 mospheric CO2, the positions Figure 25.9. Variation in δ O in planktonic (a) and benthic (b) fora- of landmasses relative to the minifera over the past 60 million years in 3 DSDP cores from the poles and elevation of land- South Pacific. Note that δ18O is relative to PDB, rather than SMOW. masses may play an important This is conventional for carbonates. After Shackleton and Kennett role in global climate. All (1975). these factors have varied during the Tertiary, so it is perhaps not surprising that there have been significant climatic changes through the Tertiary. These changes have been recorded by δ18O in deep-sea sediments. Figure 25.9 shows the Tertiary δ18O variations in benthic and planktonic foraminifera recorded in three DSDP cores from the South Pacific. The data show an increase in δ18O through this time. Because extensive northern hemisphere glaciation only began in the Pleistocene or late Pliocene, variations in δ18O through most of the Tertiary are thought to primarily reflect temperature changes rather than changes in ice volume, though the latter, mainly in Antarctica, were also important. The δ18O record therefore testifies to gradual cooling through the Tertiary. Superimposed on the general increase in δ18O are some important “events” in which δ18O changes more rapidly. Going backward through time, these include the shifts that mark the onset of Pleistocene glaciation, the rapid increase in δ18O from mid-Miocene through Pliocene, and the nearly 1‰ increase at or near the Eocene-Oligocene boundary, and a more steady decrease, amounting to nearly 2‰ during the Eocene. The increase in δ18O in the latter half of the Miocene is thought to be due to growth of the West Antarctic ice sheet, and the increase at the Eocene-Oligocene boundary is thought to be associated with the initiation of the East Antarctic ice sheet. Prior to this time, Antarctica would have been largely ice-free. Comparison of cores from different parts of the ocean shows that the changes are globally synchronous. We’ll examine these changes in more detail in a subsequent lecture. Studies of spatially distributed cores suggest that global temperatures were some 2° C warmer during the Eocene that at present. Perhaps more significantly, the latitudinal gradient in temperature may have been only half the present one. This suggests oceanic and atmospheric circulation was different from the present, and on the whole much more efficient at transporting heat from equator to poles. Why this was so remains unclear. The Eocene-Oligocene shift is thought to represent the beginning of present system where temperature variations dominate thermohaline circulation in the oceans, and initiation of extensive East Antarc- tic glaciation. As we found in the previous section, deep ocean water masses are formed at high latitudes and are dense mainly because they are cold. Typically deep water today has a temperature be-

303 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

tween 2° and –2° C. Before the Eocene, deep water appears to have been much warmer, and thermohaline circulation may have been dominated by salinity differences. (The formation of Mediterranean Intermediate Water, which forms as a result of evaporative increase in salinity, can be viewed as a rem- nant of this salinity-dominated circulation.) It was probably not until late Miocene that the present thermohaline circulation was completely established. Even subsequent to that time, important variations may have occurred, as we have seen. The mid-Miocene increase in δ18O probably represents the expansion of the Antarctic ice sheets to cover West Antarctica. This interpretation is supported by δD analyses of sediment pore water. Even though pore water exchanges with sediment, water dominates the deuterium budget so that δD values are approximately conservative (diffusion also affects δD, but this effect can be corrected for). An increase of about 10‰ δD occurs between mid and late Miocene, which is though to reflect the accumulation of deuterium-depleted water in Antarctic ice sheets. CONTINENTAL ISOTOPIC RECORDS Climate change has left an isotopic record on the continents as well as in the deep sea. As with the

deep-sea records, it is the isotopic composition of H2O that is the paleoclimatic indicator. The record may be left directly in ice, in carbonate precipitated from water, or in clays equilibrated with water. We will consider examples of all of these in this lecture. As we noted with the deep-sea carbonate record, the preserved isotopic signal can be a function of several variables. Continental records tend to be even more difficult to interpret than marine ones. All the isotopic records we will consider record in some fashion the isotopic composition of precipitation in a given region. The isotopic composition of precipitation depends on a host of factors: (1) The isotopic composition of the oceans (the ice volume effect). (2) The isotopic composition of water in the source area (the δ18O of surface water in the ocean varies by a per mil or more because of evaporation, precipitation and freezing and is correlated with salinity). (3) Temperature and isotopic fractionation in the source area (when water evaporates a temperature dependence isotopic fraction occurs; kinetic affects will also occur, and will depend on the vigor of mixing of water at the sea surface; higher wind speeds and more turbulent mixing will reduce the kinetic fractionation). (4) Atmospheric and oceanic circulation patterns (as we saw in earlier lectures, the isotopic composition of water vapor is a function of the fraction of vapor remaining, which is not necessarily a simple function of temperature; changes in atmospheric and oceanic circulation may also result in changes in the source of precipitation in a given region). (5) Temperature in the area where the precipitation falls, as this determines the fractionation between vapor and water. (6) Seasonal temperature and precipitation patterns. The isotopic record might reflect water falling during only part of the year, and the temperature recorded may therefore be that of only a single season rather than an annual average. For example, even in a wet area such as Ithaca, recharge of ground water occurs almost entirely in winter; during summer, evaporation usually exceeds precipitation. (7) Evaporation of water or sublimation of ice. The isotopic record might be that of water remaining after some has evaporated. Since evaporation involves isotopic fractionation, the preserved isotopic record will not necessarily be that of the precipitation that falls. All of these are climatic factors and are subject to change between glacial and interglacial periods. Changes in these factors do not mean that the stable isotope record in a given region is not recording climatic changes, but they do mean that the climatic changes recorded might not be global ones.

304 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

Vostok Ice Core Climatologists recognized early on that continental ice preserves a stratigraphic record of climate change. Some of the first ice cores recovered for the purpose of examining the climatic record and analyzed for stable isotopes were taken from Greenland in the 1960’s (e.g., Camp Century Ice Core). Sub- sequent cores have been taken from Greenland, Antarctica, and various alpine glaciers. The alpine glaciers generally give isotopic records of only a few thousand years, but are nevertheless useful, recording events such as the Little Ice Age. The Greenland and Antarctic cores provide a much longer record. The most remarkable and useful of these cores is the core recovered by the Russians from the Vostok station in Antarctica (Jouzel, et al., 1987, 1996), and is compared with the marine δ18O record in Figure 25.10. The marine record shown is the SPECMAP record (Figure 25.3), which is a composite record based on that of Imbrie et al. (1984), but with further modification of the chronology. The core * 18 18 provides a 425,000 year record of δD and δ Oice, as well as CO2, and δ OO2 in bubbles (the latter was published subsequently and is not shown in Figure 25.10; we will return to this in a subsequent lec-

Figure 25.10. Climate record of the upper 2700 m of the Vostok Ice core. The upper curve shows δD in the Vostok ice core. The second curve shows the calculated temperature difference relative to the present mean annual temperature, and the lower curve shows the marine carbonate SPECMAP δ18O curve. (Vostok data is from Jouzel, et al., 1987, 1993, 1996).

* The European Project for Ice Coring in Antarctica (EPICA) has subsequently drilled a second core and obtained a record going back to 650,000 years BP.

305 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

ture), provides a measure of δ18O in the atmosphere and, indirectly, the ocean), which corresponds to a full glacial cycle. A s Jouzel et al. (1987) converted δD to temperature variations after subtracting the effect of changing ice volume on δD of the oceans. The conversion is based on a 6‰/°C relationship between δD and temperature in Antarctic snow (they found a similar relationship using circulation models). The hydrogen isotopic fractionation of water is a more sensitive function of temperature than is oxygen fractionation. Since this relationship might have changed between glacial periods and the present, there is some uncertainty in these temperature estimates, but they point out that they are consistent with a relationship between crystal size and temperature. Their results, taken at face value, show dramatic 10° C temperature variations between glacial and interglacial times. Dating of the Vostok ice core was based only on an ice-flow model. Nevertheless, the overall pattern observed is in remarkable agreement with the marine δ18O record, particular from 110,000 years to the present. The record of the last deglaciation is particularly similar to that of the marine δ18O record, and even shows evidence of a slight return trend toward glacial conditions from 12 kyr to 11 kyr BP, which corresponds well to the well documented Younger Dryas event of the North Atlantic (although the amplitude of this event in the Vostok core is much smaller than in North Atlantic records). It is also very significant that spectral analysis of the Vostok isotope record shows strong peaks in variance at 41 kyr (the obliquity frequency) and at 25 kyr, which agrees with the 23 kyr precessional frequency when the age errors are taken into consideration. Thus the Vostok ice core data appear to confirm the importance of Milankovitch climatic forcing. It is interesting and significant that even in this Figure 25.11. δ18O records as a function of age from the core, taken at 78° S, it is primarily insola- GRIP and GISP2 ice cores from Summit, Greenland. The tion at 65° N that is the controlling influ- GRIP data is a 200 year average and thus appears some- ence. There are, however, some differences what smoother than the GISP2 data. Data from Grootes between the Vostok record and the marine et al. (1993) and Stuiver et al. (1995) (GISP2 δ18O), Meese record, and we will consider these further et al. (1994 ) and Sowers et al., (1993) (GISP2 time scale), in a subsequent section. and Dansgaard et al. (1993) and Taylor et al. (1993) (GRIP Figure 25.10 shows only the upper 2600m data). of the Vostok core, which was subsequently deepened to a total depth of 3310m. Jouzel et al. (1996) and Petit et al. (1999) reported the δD, δ18O, and calculated temperatures for these data. Data on Na and dust concentrations in the ice and

306 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

18 15 3 4 CO2, CH4, O2, N2, N2O concentrations and δ O, δ N, and He/ He (the latter provide a record of the cosmic particle flux) in gas bubbles in the ice have also been published on these cores. Together, these data provide a record of both climate and atmospheric composition, going back 422,000 years. We’ll consider some of these data in the next lecture. Ice Records from Summit, Greenland: GRIP and GISP2 To compliment the remarkable record of the Vostok core, drilling was begun in the late 1980’s on two deep ice cores at the summit of the Greenland ice cap. A core drilled by a European consortium project, called GRIP (Greenland Ice Core Project), was located exactly on the ice divide; a core by a U. S. consortium, called GISP2 (Greenland Ice Sheet Project), was drilled 28 km to the west of the GRIP site. Drill- ing on these 3000 m cores was completed in 1992 and 1993 respectively. These cores provide very detailed climate records of the Holocene and the most recent glacial maximum. The also provide a record of climate in the northern hemisphere, and in the North Atlantic in particular, the region which un- doubtedly holds the key to Quaternary glacial cycles. δ18O records for both cores are shown in Figure 25.11. Down to depths of approximately 2700 m, which corresponds to roughly to the past 110,000 years, there is an excellent correlation in δ18O between the two cores. Both also agree well with the marine δ18O records and the Vostok record. Below this depth, δ18O and other parameters are not correlated between the cores, and the δ18O variations are not consistent with those in the marine or Vostok records. This is due to ice flow and folding in the Green- land cores. The last glacial interval, spanning the period from roughly 110,000 years ago to 14,000 years ago was a time during which climate alternated from periods nearly as warm as the present to much colder ones. The highly detailed record provided by the Greenland cores demonstrated that the transition between the two climate “states” was very rapid, occurring on time periods of a century and less in many cases. These rapid climate variations have been subsequently correlated to δ18O variations in high-resolution (i.e., high sedimentation rate) sediment cores from the North Atlantic. The δ18O record from the GRIP core suggests that the 1500 last interglacial period,

which ended about 115,0000. years ago was also charac- terized by highly unstable 1000 climate. While marine cores do confirm the climatic instability of the last glacial pe-

riod, they do not do so for Ca, ppb the last interglacial. Nor is 500 this instability seen in the Vostok core. Whether the rapidly varying δ18O and δD observed in the GRIP core at 0 depths corresponding to the 0 20 40 60 80 100 120 last glacial period actual reflect climatic variations or Age, ka instead result from folding Figure 25.12. Ca2+ variation in ice from the GISP2 core. High and and interlayering of ice dur- variable Ca2+ ion concentrations are found during cold periods, lower ing flow remains a matter of concentrations during warm periods. Ca ion, derived mainly from debate, though the weight of calcite in arid region soils, serves a measure of the atmospheric dust the evidence now supports concentration. From Mayewski and Bender (1995).

307 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

the latter interpretation. A number of other chemical and physical parameters are being or have been measured in these cores. Perhaps the most important finding to date is that cold periods were also dusty periods (again, this had previously been suspected from marine records). Ice formed in glacial intervals has high concentrations of Ca2+ (Figure 25.12), derived from soils in arid regions, as well as dust, indicating higher atmospheric dust transport during glacial periods, reflecting conditions that were both dustier and windier. Windier conditions could well result if thermohaline circulation was reduced, as the pole to equator temperature gradient would increase. Atmospheric dust may be an important feedback in the climate cycle: dust can act as nuclei for water condensation, increasing cloud cover and cooling the climate (Walker, 1995). It may also serve as a feedback in another way. There is now firm evidence that the abundance of dissolved Fe in surface waters limits biological productivity in some regions. In parts of the ocean far from continents wind blown dust is a significant source of Fe. Increased winds during the last glacial period may have fertilized the ocean with Fe, effectively turning up the biological pump and drawing

down atmospheric CO2. This idea however, remains specula- tive. Devil’s Hole Vein Calcite Record Another remarkable isotopic record is that of vein calcite in Devil’s Hole in Nevada. Devil’s Hole is an open fault zone near a major groundwater discharge area in the southern basin and range in southwest Nevada (Devil’s Hole is located in the next basin east from Death Valley). The fissure is lined with calcite that has precipitated from supersaturated ground water over the past 500,000 years. A 36 cm long core was recovered by SCUBA divers and analyzed by Winograd, et al. (1992). The results are compared with the Vostok and SPECMAP records in Figure 25.13. Ages of the Devil’s Hole core are based 22 U-Th ages determined by mass spectrome- try. Though the Devil’s Hole record is strongly similar to the SPECMAP record, there are some significant differences. In particular, Winograd et al. (1992) noted that Termination II, the end of the second to the last glacial epoch, in the Devil’s Hole and Vostok records precedes that seen in the SPECMAP record by about 13 kyr (140 kyr vs. 127 kyr). This is an im- * portant point because Termination II in the SPECMAP re- Figure 25.13. Comparison of Devil’s cord corresponds with a peak in northern hemisphere sum- Hole (DH-11) δ18O, marine carbonate mer insolation. Since Termination II in the Devil’s Hole re- (SPECMAP) δ18O, and Vostok ice core cord, which is much better dated than the other two, appears δD climate records. Dashed lines show to precede the peak in summer insolation, Winograd et al. ar- the times of Terminations I-V. After gued that the Milankovitch theory must be wrong, i.e., that Winograd, et al. (1992). insolation variations due to orbital changes cannot be driving glacial cycles.

* The ends of glacial epochs are called Terminations. They are quantitatively defined as the mid-point in the δ18O rise and are numbered successively backward in time.

308 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

The subtle differences between the Vostok, Devil’s Hole, and SPECMAP records have been carefully

. considered by Grootes (1993). He noted that the age control on the SPECMAP record is weak. Ages on this record have been adjusted to correlate with sea level changes as determined by dating of coral reefs and terraces and “tuned” to Milankovitch cycles. Recent high precision mass spectrometric U-Th dates on coral terraces from Barbados and New Guinea are in fact about 7 thousand years older than earlier alpha counting dates, which were used to adjust 5 the SPECMAP time scale. Thus the age of Termi- C Midwestern N. America nation II in the marine carbonate record needs to be G J revised upward from 127 kyr to 135 kyr. Once this

r Western USA

e 0 t J revision is made, Grootes notes that the completion J J Other of the glacial-interglacial change coincides in the 3 c Wa J records (at about 130-132 kyr), but the beginnings

ori GJ e -5 J t JJ differ. JJ GG Grootes (1993) argues that all 3 records may be

Me C JC G

W C GG G correct but may be recording different aspects of O -10 JCC J C

M G climate change. These differences may provide

S C C

O GG G some insight into the exact manner in which glacial 8 G 1 G G 18 δ G G GG C epochs end. The onset of the δ O increase in the -15 C G GG GG Vostok record, which occurs at 145 kyr, signifi- G cantly precedes the northern hemisphere increase in insolation, but it does coincide well with an in- -20 -20 -15 -10 -5 0 5 crease in southern hemisphere summer insolation. The Vostok temperature increase may well reflect δ18O Soil Carbonate PDB this increased southern hemisphere insolation. Figure 25.14. Relationship between δ18O in local This is consistent with an earlier inference that average meteoric water and soil carbonate. From melting of what were probably quite substantial ice Cerling and Quade (1993). shelves around Antarctica preceded melting of the northern hemisphere ice cap and caused the initial in- !18O Pedogenic Carbonate crease in δ18O in the ocean. Southern hemisphere insola- -12-10-8 -6 -4 tion waned at about 138 kyr, and subsequent warming H 0 H H HHH H H HH HHHHHHHHHBHHHH and sea level drive would have been driven by northern 2 G HG BB B hemisphere warming and interhemispheric coupling by GG G G 4 G G ocean currents and CO2. G G G G GG GG GGEEG 6 G EEEEE Grootes (1993) was also able to explain much of the re- E GG E EE G GE J E Age, Ma J GJ J E E maining discrepancy between the Devil’s Hole and 8 J JJJ GGJ J G J J G SPECMAP records. The isotopic variations in Devil’s Hole 10 J J J J G are due local temperature changes, changes in ocean iso- 12 J J G topic composition, and all the other factors we discussed 14 J J J J J JJ J above. Grootes (1993) first corrected the Devil’s Hole re- 16 J J J cord for changing oceanic δ18O. The effect of this is to 18 make the residual δ18O variations greater than the uncor- 18 Figure 25.15. δ O in paleosol carbonate rected ones. He suggested that differences in temperature nodules from the Potwar Plateau in north- at the site of evaporation and increased wind velocity† ern Pakistan. Different symbols corre- during the glacial maxima just before Termination II spond to different, overlapping sections would reduce the fractionation recorded in the Devil’s that were sampled. After Quade et al. Hole area. The increase in Devil’s Hole δ18O may reflect (1989).

† Both the increased equator-to-pole temperature gradient and increased concentration of dust in ice cores indicate higher wind speeds at glacial maxima.

309 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

this reduced fractionation. A change in ocean-atmosphere circulation patterns may have effectively blocked cold Arctic air from reaching the Devil’s Hole area and moderated temperatures there. The controversy surrounding the Devil’s Hole record emphasizes the complexity of factors influencing continental isotopic records and the difficulty in their interpretation. Soils and Paleosols

The concentration of CO2 dissolved in soil solutions is very much higher than in the atmosphere, reaching 1% by volume. As a result, soil water can become supersaturated with respect to carbonates. In soils where evaporation exceeds precipitation, soil carbonates form. The carbonates form in equilib- rium with soil water, but the isotopic composition of soil water tends to be heavier than that of mean annual precipitation. There are 2 reasons for this. First, soil water enriched in 18O relative to meteoric water due to preferential evaporation of isotopically light water molecules. Second, rain (or snow) falling in wetter, cooler seasons in more likely to run off than during warm sea- Hawaii sons. Thus there is a strong correlation 0 meteoric water line 18 Southern US between δ O in soil carbonate and mete- Coastal oric water, though soil carbonates tend to -40 be about 5‰ more enriched than ex- Oregon pected from the calcite-water fractiona- -80 Colorado tion (Figure 25.14). Because of this cor- !D‰ relation, the isotopic composition of soil -120 carbonate may be used as a paleoclimatic -160 Idaho indicator. Montana meteoric water line Figure 25.15 shows one example of δ18O in paleosol carbonates used in this -20 -10 0 +10 +20 way. The same Pakistani paleosol sam- 18 ples analyzed by Quade et al. (1989) for ! O ‰ 18 δ13C (Figure 24.7) were also analyzed for Figure 25.16. Relationship between δD and δ O in modern 18 δ18O. The δ13C values recorded a shift to- meteoric water and kaolinites. Kaolinites are enriched in O 2 ward more positive values at 7 Ma, by about 27‰ and H by about 30‰. After Lawrence and which apparently reflect the appearance Taylor (1971). 18 of C4 grasslands. The δ O shows a shift to more positive values at around 8 Ma, or a million years -15.0 before the 13C shift. Quade et al. interpreted this δ 15.5 -12.5 18.3 as due to an intensification of the Monsoon sys- 18.8 16.3 14.9 18.8 18.0 tem at that time, and interpretation consistent 19.0 -10.0 19.1 16.7 16.9 17.4 with marine paleontological evidence. 20.4 Clays, such as kaolinites, are another important 19.7 -7.5 21.3 constituent of soil. Savin and Epstein (1970) 16.5 showed that during soil formation, kaolinite and 13.2 22.4 19.5 18.7 21.0 21.7 montmorillonite form in approximate equi- 22.4 22.3 21.1 librium with meteoric water so that their δ18O -5.0 values are systematically shifted by +27 ‰ relative the local meteoric water, while δD are shifted by about 30‰. Thus kaolinites and mont-

morillonites define a line parallel to the meteoric Figure 25.17. δ18O in Cretaceous kaolinites from water line (Figure 25.16), the so-called kaolinite North American compared with contours of δ18O line. From this observation, Lawrence and Tay- (value shown in outline font) of present-day mete- lor (1972) and Taylor (1974) reasoned that one oric water. After Lawrence and Meaux (1993).

310 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

should be able to deduce the isotopic composition of rain at the time ancient kaolinites formed from their δD values. Since the isotopic composition of precipitation is climate dependent, as we have seen, ancient kaolinites provide another continental paleoclimatic record. Lawrence and Meaux (1993) conclude, however, that most ancient kaolinites have exchanged hydrogen subsequent to their formation, and therefore a not a good paleoclimatic indicator (this conclusion is, however, controversial). On the other hand, they conclude that oxygen in kaolinite does preserve the original δ18O, and that can, with some caution, be used as a paleoclimatic indicator. Figure 25.17 compares the δ18O of ancient Cretaceous North American kaolinites with the isotopic composition of modern precipitation. If the Cretaceous climate were the same as the present one, the kaolinites should be systematically 27‰ heavier than modern precipitation. For the southeastern US, this is approximately true, but the difference is generally less than 27‰ for other kaolinites, and the difference de- creases northward. This indicates these kaolinites formed in a warmer environment than the present one. Overall, the picture provided by Cretaceous kaolinites confirm what has otherwise be deduced about Cretaceous climate: the Cretaceous climate was generally warmer, and the equator to pole temperature gradient was lower. REFERENCES AND SUGGESTIONS FOR FURTHER READING Broeker, W. S. 1984. Terminations. in Milankovitch and Climate, ed. A. Berger, J. Imbrie, J. Hayes, G. Kukla and B. Satzman. 687-689. Dordrecht: D. Reidel Publishing Co. Cerling, T. E. and J. Quade, Stable carbon and oxygen in soil carbonates, in Climate Change in Continental Isotopic Records, Geophysical Monograph 78, edited by P. K. Swart, K. C. Lohmann, J. McKenzie and S. Savin, p. AGU, Washington, 1993. Cerling, T. E., The stable isotopic composition of modern soil carbonate and its relationship to climate, Earth Planet. Sci. Lett., 71, 229-240, 1984. Charles, C. D. and R. G. Fairbanks. 1992. Evidence from Southern Ocean sediments for the effect of North Atlantic deep-water flux on climate. Nature. 355: 416-419. Craig, H., Measurement of oxygen isotope paleotemperatures, in Stable Isotopes in Oceanographic Studies and Paleotemperatures, edited by E. Tongiorgi, p. 161-182, CNR Lab. Geol. Nucl., Pisa, 1965. Dansgaard, W., Jonhsen, S. J., Clauson, H. B., Dahl-Jensen, D., Gundestrup, N. S., Hammer, C. U., Hvidberg, C. S., Steffensen, J. P., Sveinbjornsdottir, A. E., Jouzel, J., and G. Bond. 1993. Evidence for general instability in past climate from a 250-kyr ice-core record. Nature, 364:218-220. Emiliani, C., Pleistocene temperatures, J. Geol., 63, 538-578, 1955. Epstein, S., H. A. Buchbaum and H. A. Lowenstam, Revised carbonate-water isotopic temperature scale, Bull. Geol. Soc. Am., 64, 1315-1326, 1953. Grootes, Interpreting continental oxygen isotope records, in Climate Change in Continental Isotopic Re- cords, Geophysical Monograph 78, edited by P. K. Swart, K. C. Lohmann, J. McKenzie and S. Savin, p. 37-46, AGU, Washington, 1993. Grootes, P.M., Stuiver, M., White, J.W.C., Johnsen, S., and Jouzel, J. 1993. Comparison of oxygen isotope records from the GISP2 and GRIP Greenland ice cores. Nature 366:552-554. Hayes, J. D., J. Imbrie and N. J. Shackleton, Variations in the Earth's orbit: pacemaker of the ice ages, Science, 194, 1121-1132, 1976. Imbrie, J., A theoretical framework for the Pleistocene ice ages, J. Geol. Soc. Lond., 142, 417-432, 1985. Imbrie, J., J. D. Hayes, D. G. Martinson, A. McIntyre, A. Mix, et al., The orbital theory of Pleistocene climate: support from a revised chronology of the marine δ18O record, in Milankovitch and Climate, Part 1, edited by A. L. Berger, J. Imbrie, J. Hayes, G. Kukla and B. Saltzman, p. 269-305, D. Reidel, Dordrecht, 1984. Jouzel, C., C. Lorius, J. R. Perfit, C. Genthon, N.I. Barkov, V. M. Kotlyakov, and V. N. Petrov, Vostok ice core: a continuous isotope temperature record over the last climatic cycle (160,000 years), Nature, 329: 403-403, 1987.

311 4/22/09 Geol. 656 Isotope Geochemistry

Lecture 25 Spring 2009

Jouzel, J., C. Waelbroeck, B. Malaizé, M. Bender, J. R. Petit, N. I. Barkov, J. M. Barnola, T. King, V. M. Kotlyakov, V. Lipenkov, et al., Climatic interpretation of the recently extended Vostok ice records, Clim. Dyn., 12:513-521, 1996. Jouzel, J., N. I. Barkov, J. M. Barnola, M. Bender, J. Chappelaz, C. Genthon, V. M. Kotlyakov, V. Li- penkov, C. Lorius, R. Petit, et al., 1993. Extending the Vostok ice-core record of paleoclimate to the penultimate glacial period, Nature, 364:407-412. Lawrence, J. R. and H. P. Taylor, Hydrogen and oxygen isotope systematics in weathering profiles, Geo- chim. Cosmochim. Acta, 36, 1377-1393, 1972. Lawrence, J. R. and J. R. Meaux, The stable isotopic composition of ancient kaolinites of North America, in Climate Change in Continental Isotopic Records, Geophysical Monograph 78, edited by P. K. Swart, K. C. Lohmann, J. McKenzie and S. Savin, p. 249-261, AGU, Washington, 1993. Mayewski, P. A. and M. Bender. 1995. The GISP2 ice core record — paleoclimatic highlights. Rev. of Geophys. Suppliment US National Report to the IUGC 1991-1994. 33 Meese, D., Alley, R., Gow, T., Grootes, P.M., Mayewski, P., Ram, M., Taylor, K., Waddington, E., and Zielinski, G. 1994. Preliminary depth-age scale of the GISP2 ice core. CRREL Special Report 94-1. Oppo, D. W. and R. G. Fairbanks. 1987. Variability in the deep and intermediate water circulation of the Atlantic Ocean during the past 25,000 years: Northern Hemisphere modulation of the Southern Ocean. Earth Planet. Sci. Lett. 86: 1-15. Petit, J. R., J. Jouzel, D. Raynaud, N. I. Barkov, J. M. Barnola, I. Basile, M. Bender, J. Chappellaz, J. Davis, G. Delaygue, et al., Climate and Atmospheric History of the Past 420,000 years from the Vostok Ice Core, Nature, 399:429-436, 1999. Quade, J., T. E. Cerling and J. R. Bowman, Development of Asian monsoon revealed by marked ecologi- cal shift during the latest Miocene in northern Pakistan, Nature, 342, 163-166, 1989. Savin, S. M. and S. Epstein, The oxygen and hydrogen isotope geochemistry of clay minerals, Geochim. Cosmochim. Acta, 34, 25-42, 1970. Shackleton, N. J. and J. P. Kennett, Paleotemperature history of the Cenozoic and the initiation of Ant- arctic glaciation: oxygen and carbon isotope analyses in DSDP sites 277, 279, and 281, Initial Rep. Deep Sea Drill. Proj., 29, 743-755, 1975. Shackleton, N. J. and N. D. Opdyke, Oxygen isotope and paleomagnetic stratigraphy of an equatorial Pacific core V28-238: oxygen isotope temperatures and ice volumes on a 105 and 106 year time scale, Quat. Res., 3, 39-55, 1973. Sowers, T., Bender, M., Labeyrie, L., Martinson, D., Jouzel, J., Raynaud, D., Pichon, J.J., and Korot- kevich, Y., 1993, 135,000 Year Vostok-SPECMAP common temporal framework. Paleoceanography 8:737-766. Stuiver, M., P. M. Grootes and T. F. Braziunas. 1995. The GISP2 delta 18O climate record of the past 16,500 years and the role of the sun, ocean, and volcanoes. Quaternary Research. 44: Taylor, K. C., Lamorey, G. W., Doyle, G. A., Alley, R. B., Grootes, P. M., Mayewski, P. A., White, J. W. C., and L. K. Barlow, 1993. The flickering switch of late Pleistocene climate change. Nature, 361:432- 436. Urey, H. C., The thermodynamics of isotopic substances, J. Chem. Soc., 1947, 562-581, 1947. Winograd, I., T. B. Coplen, J. M. Landwher, A. C. Riggs, K. R. Ludwig, et al., Continuous 500,000 year climate record from vein calcite in Devils Hole, Nevada, Science, 258, 255-260, 1992.

312 4/22/09