Page 277 - Earth's Climate Past and Future

CHAPTER 14 • Millennial Oscillations of Climate 253

layers. The estimated counting errors increase from a
18
TABLE 14-1 Causes of δ O Changes few decades for ice 10,000 years old to several thousand
Recorded in Ice Cores years for ice 50,000 years old and to considerably more
18
Change in δ O for older layers.
Negative values Positive Both ice core sequences yielded nearly identical
long-term climate records to within 200 m of the
Colder Air temperature Warmer underlying bedrock. A portion of one of the ice core
over ice records is shown in Figure 14–2 (right). Because the
Distant Proximity of source Close two cores recorded nearly identical climatic signals over
region that entire interval, scientists had no doubt that both
18
18
18
Low δ O δ O composition High δ O were reliable records of climate.
of source
14-2 Oscillations Recorded in North
High Elevation of ice Low
Atlantic Sediments
Winter Primary season Summer
of precipitation During the 1980s and 1990s, millennial oscillations
were also being discovered in North Atlantic sediments.
Ocean sediments are normally not a promising archive
for monitoring short-term climatic fluctuations because
glacial intervals oscillated rapidly between extremely cold deposition rates are usually no greater than 1 or
intervals called stadials and relatively mild intervals called 2 cm/1000 yr, and small burrowing animals stir and mix
interstadials. These oscillations are often referred to as the sediments to depths of 5–10 cm (Chapter 2). As a
Dansgaard-Oeschger oscillations in honor of the geo- result, mixing usually obliterates climate oscillations
chemists Willi Dansgaard and Hans Oeschger, who first shorter than 2500 to 5000 years.
found and studied them. Their work suggested that the Fortunately, places exist in the North Atlantic Ocean
oscillations were spaced at intervals that ranged from as where deposition rates can be as high as 10–20 cm per
little as 1000 years to almost 9,000 years in length. 1000 years. Bottom currents carry fine sediments away
Each fluctuation toward more negative (glacial) from locations that are subject to swift flow and deposit
18
δ O values is matched by an abrupt increase in dust them as large lens-shaped piles (called sediment drifts)
concentrations in the ice. Again, the range of variation in regions where the currents slow. The process is the
in dust concentrations in these oscillations is a large same as the one that creates snowdrifts by scouring snow
fraction of the total difference between glacial and from exposed regions where the winds are strongest
interglacial values. Geochemical analysis of the dust and piling it in regions where the wind speed slows.
shows that most of it comes from distant source regions In this case, the coarser sand-sized sediments such as
in Asia, not from nearby North America. The size of foraminifera and ice-rafted debris are not so easily
the dust particles is larger in the cold intervals than in moved by bottom currents as are the silts and clays.
the milder ones, indicating that strong winds lifted and They tend to stay in place as a reliable record of climate
transported the dust when climate was very cold. The changes even while bottom currents are delivering fine
colder intervals also contain larger amounts of sea salt sediments that rapidly bury and preserve their climatic
–2
+1
(Na and Cl ions) plucked from salty sea spray above information.
the turbulent ocean during cold and windy intervals and In the mid-1980s, studies of these rapidly deposited
carried to the ice. sediments in the North Atlantic Ocean first detected
In the late 1980s, two long sequences were drilled shorter climate oscillations. The marine geologist Hart-
on the summit of the Greenland ice sheet at sites named mut Heinrich found episodes of unusually abundant ice
GISP and GRIP. These sites were carefully positioned rafting separated by as little as 5000 years to as much as
over areas of smooth underlying bedrock to minimize 15,000 years or more. These episodes are often referred
the impact of changes in ice flow that can disturb deeper to as Heinrich events. Later, the geologist Gerard
ice layers, and they were drilled less than 30 km apart to Bond discovered even shorter-term variations in two
see whether or not they would reveal similar climate climatic indices: (1) the percentage of the single polar
histories. The annual layering in these new cores species of foraminifera compared with the total popula-
extended to a greater depth and was used to date them tion and (2) the relative amounts of the shells of
further back in time, although stretching and thinning foraminifera compared with the sand-sized grains of
still introduced dating uncertainties deeper in the ice. ice-rafted sand. As in the case of orbital-scale changes,
The use of several annually deposited signals (dust, he used higher percentages of cold-water foraminifera
18
δ O, and others) lessened the chance of miscounting and larger concentrations of ice-rafted debris as an