Past Climate Variability and Change in the Arctic and at High Latitudes

50
The U.S. Climate Science Program Chapter 3
A B
Figure 3.14. A) One-meter section of Greenland Ice Core Project-2
core from 1,837 m depth showing annual layers. (Photograph courtesy
of Eric Cravens, Assistant Curator, U.S. National Ice Core Laboratory.)
B) Field site of Summit Station on top of the Greenland Ice Sheet
(Photograph by Mark Twickler, GISP2 SMO, University of New
Hampshire; NOAA Paleoslide Set.)
The center of the
Greenland Ice Sheet has
not finished warming
from the ice age.
the mean annual temperature at the core site,
and the use of both measures together offers
additional information about conditions at
the source of the water vapor (e.g., Dansgaard
et al., 1989). Recent work by Werner et al.
(2000), however, demonstrates that changes
in the seasonal cycle of precipitation over the
ice sheets can affect measurements of ice-core
temperature.
The underlying idea is that an air mass loses
water vapor by condensation as it travels from
a warm source to a cold (polar) site. This point
is easily shown by the nearly linear relationship
between precipitation and temperature over
modern ice sheets (Figure 3.15). Water that
contains the heavy isotopes has a lower vapor
pressure, so the heavy isotope preferentially
condenses into rain or snow, and the air mass
becomes progressively depleted of the heavy
isotope it moves to colder sites. It can easily
be shown from spatial surveys (Johnsen et al.,
1989) and, indeed, from modeling studies using
models enabled with water isotopes (e.g., Hoffmann
et al., 1998; Mathieu et al., 2002) that a
good spatial relationship between temperature
and water isotope ratio exists. The relationship
is
δ=aT+b
where T is mean annual surface temperature,
and δ is annual mean δ 18O or δD value in precipitation
in the polar regions, and the slope, a, has
values typically around 0.6 for GreenlAnd δ 18O.
Temperature is not the only factor that can affect
isotopic ratios. Changes in the season when
snow falls, in the source of the water vapor,
and other things are potentially important as
well (Jouzel et al., 1997; Werner et al., 2000)
(Figure 3.16). For this reason, it is common
whenever possible to calibrate the isotopic
ratios using additional paleothermometers.
For short intervals, instrumental records of
temperature can be compared with isotopic
ratios (e.g., Shuman et al., 1995). The few comparisons
that have been done (summarized in
Jouzel et al., 1997) tend to show δ/T gradients
that are slightly lower than the spatial gradient.
Accurate reconstructions of past temperature,
but with low time resolution, are obtained from
the use of borehole thermometry. The center
of the GreenlAnd ice Sheet has not finished
warming from the ice age, and the remaining
cold temperatures reveal how cold the ice age
was (Cuffey et al., 1995; Johnsen et al., 1995).
Additional paleothermometers are available
that use a thermal diffusion effect. In this effect,
gas isotopes are separated slightly when
an abrupt temperature change at the surface
creates a temperature difference between the
surface and the region a few tens of meters
down, where bubbles are pinched off from the
interconnected pore spaces in old snow (called
firn). The size of the gas-isotope shift reveals
the size of an abrupt warming, and the number
of years between the indicators of an abrupt
change in the ice and in the bubbles trapped in
ice reveals the temperature before the abrupt
change—if the snowfall rate before the abrupt
change is known (Severinghaus et al., 1998;
Severinghaus and Brook, 1999; Huber et al.,
2006). These methods show that the value of