Info

FIGURE 5.40 The fiiaOATM record in the GISP2 andVostok ice cores (top) derived from gas bubbles in each core.The strong correlation has enabled the SD and 8I80 records in the ice (lower two series) to be aligned chronologically, though the absolute timescale remains uncertain.The two records show that interstadi-als in the Greenland record can also be identified in Antarctica if they lasted more than -2 kyr (Bender et al., 1994).

BOO x-O

Calendar age (ka)

BOO x-O

Calendar age (ka)

FIGURE 5.41 Isotopic and gas records from Byrd and Vostok (Antarctica) and Summit, Greenland (GISPII and GRIP).The records have been aligned by finding the optimum fit between 5I80 in gas bubbles in the ice (taking into account the time-varying age of gases trapped in the ice). With the chronologies phase-locked in this way, it is apparent that C02 increased just as Bl80 in ice at Byrd increased, whereas 8I80 in Greenland ice did not increase until ~3300 yr later.The methane record has more in common with the Greenland record, as it is largely controlled by northern hemisphere continental sources (Sowers and Bender, 1995).

An important by-product of being able to align the Antarctic and Greenland ice-core records together in time, using their respective 818Oatm content, is the ability to then directly compare changes in other parameters such as C02 and CH4 with temperature changes, as recorded by §18Oice (Fig. 5.41). By forcing the records of GISP2, Greenland and Byrd, Antarctica to match, Sowers and Bender (1995) showed that C02 levels began to increase by -17 ka B.P., well before there were large increases in 818Oice in the Greenland record. This rise in C02 is more or less synchronous with an increase in 818Oice at Byrd, as well as with an increase in SSTs in the Southern Ocean and tropical Atlantic, and with eustatic sea-level rise. By contrast, Greenland 818Oice shows no comparable changes, with a rapid shift from late glacial conditions delayed until -14.7 ka B.P. Hence, it appears that warming in Greenland and the North Atlantic was delayed until long after the southern hemisphere and the Tropics had begun to emerge from the last glacial period. A possible explanation is that the North Atlantic polar front was locked in position around -45° N by the atmospheric circulation around the Laurentide Ice Sheet; only when the Laurentide was reduced in size to the point that it no longer had a major influence on circulation could the polar front and associated air masses shift northward (Keigwin et ai, 1991). This shift is documented by the pronounced increase in 8180,~ at GISP2 around 14.7 ka B.P.

5.4.6 Correlation Between Ice Cores and Marine Sediments

Because changes in the isotopic content of the ocean due to continental ice sheet growth affect atmospheric 818Oatm, variations in 180 in both marine sediments and ice cores represent a common denominator that can be used to link both types of record. There are several complicating factors that arise in such a comparison. Just as the time to pore close-off in ice cores acts as a time-varying lowpass filter on the ice bubble gas record, so bioturbation in marine sediments acts to smooth the changes that occurred in oceanic isotopic composition. Also, bottom water temperatures have changed over glacial-interglacial time, and such changes also influence the isotopic composition of benthic forams, requiring adjustments to be made to obtain an unbiased time series of S18Osw. The overall residence time of 02, with respect to processes of photosynthesis and respiration, is around 2-3 ka, so that changes in 818Oatm will lag those in the ocean by that amount of time. Finally, there are no compelling reasons to suppose that the "Dole effect" has remained constant over time; indeed it is likely that 818Oatm has not simply followed 818Osw variations in the same way at all times; changes in the relative primary productivity of the terrestrial biosphere versus the marine biosphere, changes in continental hydrology, etc. may have affected the 818Oatm-S18Osw relationship over time. Nevertheless, in spite of these uncertainties, it is possible to make reasonable assumptions about each of these complications and then to correlate marine and ice core records over the last -130,000 yr. Figure 5.42 shows such a comparison, using the SPECMAP

Lorius Age (ka)

Lorius Age (ka)

Specmap Benthic Foraminifera

SPECMAP Age (ka)

FIGURE 5.42 The 8I8Oatm (from air bubbles in Vostok ice) compared to Sl8Osw (from benthic foraminifera) plotted on the SPECMAP timescale (Martinson, et al., 1987) (see Section 6.3.3).The BI8Oatm record has been forced to match the marine 8lsO record (Sowers et al., 1993).The upper timescale is the ice core chronology estimated by Lorius et al. (1985).

SPECMAP Age (ka)

FIGURE 5.42 The 8I8Oatm (from air bubbles in Vostok ice) compared to Sl8Osw (from benthic foraminifera) plotted on the SPECMAP timescale (Martinson, et al., 1987) (see Section 6.3.3).The BI8Oatm record has been forced to match the marine 8lsO record (Sowers et al., 1993).The upper timescale is the ice core chronology estimated by Lorius et al. (1985).

timescale for the marine 818Osw record, and optimizing the correlation with the ice core 818Oatm. Note that this comparison is between the age of the gas bubbles in the ice and the marine record; the age of the enclosing ice is older, the age difference being larger during low accumulation glacial periods than warmer interglacials. The overall correlation between these two records is excellent, though the fit is better at some times than at others, probably reflecting the fact that the Dole effect has not been constant over time. There is also very good agreement with the original Lorius et al. (1985) time-scale, which was based on fairly simple assumptions about past changes in accumulation rate at Vostok. However, the comparison suggests that the original ice-core chronology is "too old" at the penultimate glacial-interglacial transition (Termination 2) by 5-6 ka relative to the estimated chronology of the marine record. This argument has also been made by several other investigators who have compared marine records with the Vostok chronology (Pichon et al., 1992; Shack-leton et al., 1992). This does not resolve the question of whether the absolute chronology is correct and it is conceivable that both records are still incorrect in absolute terms. Nevertheless, the alignment of the ice core and marine records is extremely valuable because it then enables other records to be compared, providing insight into how different parts of the climate system are related in time. Figure 5.43 shows such a comparison between the Vostok 8D record and SSTs from a sub-Antarctic oceanic site (46° S). The 8D at Vostok is a function of temperatures in the air above the surface inversion and can be expected to vary with air mass temperatures over the surrounding oceans. Clearly, when aligned on a common temporal

Lorius ice age (Ka) 25 50 75 100 125

Lorius ice age (Ka) 25 50 75 100 125

0 25 50 75 100 125 150

0 25 50 75 100 125 150

SPECMAP age (Ka)

FIGURE 5.43 Temperature above the inversion layer at Vostok (relative to present, derived from 5D) and SSTs at a sub-Antarctic site (46.5°S) plotted on the SPECMAP timescale after the Vostok 8I8Oatm record was adjusted to produce an optimum fit with the SPECMAP 8l8Osw record (see Fig. 5.42) (Sowers et al., l993).The upper timescale is the ice-core chronology estimated by Lorius et al. (1985).

basis, there is a strong correlation between these series, providing support for the many assumptions made in using the &18Oatm-518Osw relationship to relate the two records (Sowers et al., 1993).

5.4.7 Ice-core Records from Low Latitudes

Ice caps are not confined to polar regions; they also occur at very high elevations in many mountainous regions, even near the Equator (Thompson et al., 1985b). High elevation ice cores provide invaluable paleoenvironmental information to supplement and expand upon that obtained from polar regions. By 1998 six high altitude sites had yielded ice cores to bedrock — Quelccaya and Huascaran in Peru, Sajama in Bolivia, and Dunde, Guliya, and Dasuopu Ice Caps in western China (see Fig. 5.1). In the Dunde and Huascaran ice cores the glacial stage ice is thin and close to the base, making a detailed interpretation very difficult (Thompson et al., 1988b, 1989, 1990, 1995). Nevertheless, even these short glacial sections can yield important information. For example, in ice cores from the col of Huascaran, Peru (6048 m) the lowest few meters contain ice from the last glacial maximum, with 8180 ~8%o lower than Holocene levels, and a much higher dust content (Thompson et al., 1995). The lower SlsO suggests that tropical temperatures were significantly reduced in the LGM (by -8-12 °C), which supports arguments that changes in tropical SSTs were much lower than those indicated by the reconstructions of CLIMAP (1981), which have guided thinking on this matter for many years (see Section 6.6).

In the Guliya ice core, from the western Qinghai-Tibetan plateau, much of the ice appears to date from the last glaciation,14 so details of conditions in this area at that time can be resolved in considerable detail (Thompson et al., 1997). The long-term record reveals several stadial-interstadial oscillations since the last interglacial (Fig. 5.44) with 8lsO values in the "interstadials" reaching Holocene

50 75

FIGURE 5.44 The 8'80 in an ice core from Guliya Ice Cap, western Tibet (-35 °N, 8l°E).The 308 m record was dated by correlating the CH4 record fromVostok and GISP2 with the oscillations of 8l8O.This assumes that the factors causing changes in 8lsO at Guliya would increase and decrease in phase with CH, (which is likely to be driven by low-latitude climatic changes) (Thompson et at., 1997).

50 75

FIGURE 5.44 The 8'80 in an ice core from Guliya Ice Cap, western Tibet (-35 °N, 8l°E).The 308 m record was dated by correlating the CH4 record fromVostok and GISP2 with the oscillations of 8l8O.This assumes that the factors causing changes in 8lsO at Guliya would increase and decrease in phase with CH, (which is likely to be driven by low-latitude climatic changes) (Thompson et at., 1997).

14 36C1 in the core indicates that the very oldest section of the record (below 300 m) may be >500,000 yr old, making it the oldest ice ever recovered (Thompson et al., 1997).

Climap Biosphere

FIGURE 5.45 A detailed section of the Guli/a ice core, centered around 25,000 yr B.P., showing large amplitude, rapid changes in 8lsO, with a period averaging ~200 yr in length. For the central section, details of variations in 5I80, micropartides, NH4, and N03 are shown; high values of 5I80 are associated with an increase in the percentage of dust particles > I |j,m, and higher levels of ammonium and nitrate ions (Thompson et at., 1997).

FIGURE 5.45 A detailed section of the Guli/a ice core, centered around 25,000 yr B.P., showing large amplitude, rapid changes in 8lsO, with a period averaging ~200 yr in length. For the central section, details of variations in 5I80, micropartides, NH4, and N03 are shown; high values of 5I80 are associated with an increase in the percentage of dust particles > I |j,m, and higher levels of ammonium and nitrate ions (Thompson et at., 1997).

levels, around -13%o. Abrupt, high amplitude changes in 8lsO occurred from -15-33 ka B.P., with an average length of -200 yr (Fig. 5.45). These oscillations are much shorter than the Dansgaard-Oeschger oscillations seen in GISP2 and are associated with higher levels of dust, NH4, and nitrate during warm (higher 8lsO) episodes (the opposite of what is seen in Greenland). This may indicate that warmer episodes occurred throughout the glacial stade, associated with less snow cover and more vegetation on the plateau. Whether the apparent periodicity is related to that seen in proxies of solar activity (-210 yr; Stuiver and Braziunas, 1992) remains to be seen.

Because of the high accumulation rates on mountain ice caps, high elevation ice cores can provide a high resolution record of the recent past, with considerable detail on how climate has varied over the last 1000-2000 yr, in particular (Thompson, 1991, 1992). The Quelccaya ice cores have been studied in most detail over this interval (Thompson et al., 1985, 1986; Thompson and Mosley-Thompson, 1987). Two cores extend back -1500 yr (though only one can be reliably interpreted before -A.D. 1200). These reveal a fairly consistent seasonal cycle of micropartides,

Accumulation Standard Total Conductivity

8180%o Deviation Particles* (nS/cm"1)

Paleoclimatologist Pangaea

FIGURE 5.46 Annual variations in 8leO, microparticles, conductivity, and accumulation at the Quelccaya Ice Cap, Peru. A distinctive period of low 8lsO values and relatively high micropartide concentrations is appar-

jfc. Huaynaputina (Peru) * > 0.63 to < 16.0 ^m per ml of eruption of 19 Feb.- 6 Mar. sample (10)5

FIGURE 5.46 Annual variations in 8leO, microparticles, conductivity, and accumulation at the Quelccaya Ice Cap, Peru. A distinctive period of low 8lsO values and relatively high micropartide concentrations is appar-

conductivity and 8lsO, which (collectively) have been used to identify and date each annual layer. Dust levels increase in the dry season (June-September) when 8lsO values and conductivity levels are highest, providing a strong annual signal. A prominent conductivity peak in A.D. 1600 (associated with a major eruption of the Peruvian volcano Huaynaputina in February-March, 1600) provides an excellent chronostratigraphic check on the annual layer counts.

Over the last 1000 yr 8180 shows distinct variations in the Quelccaya core, with the lowest values from 1530-1900 (Fig. 5.46). This corresponds to the so-called "Little Ice Age" observed in many other parts of the world. The longer record from nearby Huascarán (9°S) provides a longer perspective on this episode; it had the lowest 8180 values of the entire Holocene (Thompson et al., 1995b). Accumulation was well above average for part of this time (1530-1700) but then fell to levels more typical of the preceding 500 yr (Fig. 5.47). Accumulation was also higher from A.D. -600 to 1000. Archeological evidence shows that there was an expansion of highland cultural groups at that time. By contrast, during the subsequent dry episode in the mountains (A.D. -1040-1490) highland groups declined while cultural groups in coastal Peru and Ecuador expanded (Thompson et al., 1988a). This may reflect longer-term evidence for conditions that are common in El Niño years, when coastal areas are wet at the same time as the highlands of southern Peru are dry. Indeed, the Quelccaya record shows that El Niños are generally associated with low accumulation years, though there is no unique set of conditions observed in the ice core that permits unequivocal identification of an ENSO event (Thompson et al., 1984a). Nevertheless, by incorporating ice-core data with other types of proxy record it may be possible to constrain long-term reconstructions of ENSO events (Baumgartner et al., 1989).

High-altitude ice cores have experienced significant increases in temperature over the last few decades, resulting in glaciers and ice caps disappearing altogether in some places (Schubert, 1992). This is quite different from polar regions where temperatures have declined in many regions during the same period. At Quelccaya, temperatures in the last 20 yr have increased to the point that by the early 1990s melting had reached the Summit core site (5670 m), obscuring the detailed 8lsO profile that was clearly visible in cores recovered in 1976 and 1983 (Thompson et al., 1993). In the entire 1500-yr record from Quelccaya, there is no comparable evidence for such melting at the Summit site. Similarly, at Huascarán, in northern Peru, 8180 values increased markedly, from a "Little Ice Age" minimum in the seventeenth and eighteenth centuries, reaching a level for the last century that was higher than at any time in the last 3000 yr (Thompson et al., 1995b). Ice cores from the Gregoriev Ice Cap (in the Pamirs) Guliya and Dunde, China also show evidence of recent warming (Lin et al., 1995; Yao et al., 1995);

ent,from -A.D. 1490 to A.D. 1880. Accumulation in this period was initially high (to -1700) then fell to low levels. A large peak in conductivity in A.D. 1600 was the result of the eruption of nearby Huaynaputina.The historical record of El Niños is also shown, based on Quinn et al. (1987). 8lsO is in %» relative to SMOW; accumulation is in units of the standard deviation over the last 500 yr (Icr = 34 cm of accumulation); total particulates are xl0s ml"1; conductivity is in microSiemens cm"1 (Thompson, 1992).

Net Accumulation (m ice eq.)

Guliya

Quelccaya

1600

1200

1000

1600

1200

1000

Quelccaya Dust
0 0

Post a comment