Ice core methodology - Ice Ages and Interglacials: Measurements, Interpretation, and Models - page 114

during the spring/summer period and atmospheric circulation changes that
enable the stratospheric load to reach the troposphere. Dust particles in solid
ice and ice melt-water scatter incident light, and the intensity of light scattered at
90 to the incident light direction is proportional to the mass of suspended
particulates'' (Meese et al., 1997).
Many of these bands can be seen visually. However, a more sensitive method
for observing the bands utilizes laser light scattering (LLS). LLS allows extension
of visual stratigraphy further down in the core. The LLS method is based on the
fact that dust particles in the water or ice will scatter the light from a laser beam,
and the amount of scattered light orthogonal to the beam is proportional to the
amount of dust. However, the system works better if the ice has melted, because
bubbles in the ice also scatter the light in the upper portion of the core. The
bubbles disappear about 1,400m down the core as they change into air/ice clath-
rates. However, the LLS method applied to water is time consuming and destroys
the ice, so the method is not as widely used as solid LLS (Oard, 2005).
According to Meese et al. (1997) the LLS method can be used as an annual
layer indicator even though the signal changes from one of depth hoar to layers of
increased dust concentration as one goes from the Holocene to the Ice Age (hoar
layers are porous low-density snow layers). It was also claimed that the LLS meas-
urements were consistent with visible stratigraphy down to considerable depths.
LLS is a very valuable dating tool throughout almost the entire length of the core,
particularly in the deeper ice at GISP2 where the other techniques either fail or
become increasingly unreliable. However, an increased particulate concentration
may not be restricted to the spring or summer and additional influxes of dust may
occur during any part of the year, creating additional peaks of a non-annual
nature. The LLS signal can also be used as an indicator of major climate changes
and of some volcanic events. Meese et al. (1997) claimed that 110,000 annual
layers have been detected down to 2,800m. In fact, they believe they were able to
count 50,600 more annual layers at an accuracy of 24% from 2,800 to 3,000m.
The age at 3,000m was about 161,000 ybp . However, this lower layer was
probably distorted by ice flow, and the extension to 3,000m is likely to be invalid.
3.2.4 Layers determined by measurement
Layers may also be counted by making physical measurements on the core (rather
than by visual observation). Seasonal d 18 O cycles make exact dating possible by
counting layered variations in this quantity downward from the surface. Figure
3.7a shows d 18 O for a core increment that was deposited from ad 1210 to 1240.
When corrected for thinning, the distance between two minima is a measure of
precipitation in the year of deposition. The decade from ad 1225 to 1235 was drier
than the preceding decade and, since the d 18 O mean values were lower, it was
apparently cooler (assuming an unchanged summer to winter precipitation ratio)
(Dansgaard, 2005). However Oard (2005) has argued that this procedure is not