Challenges in the Use of Cosmogenic Exposure Dating of Moraine Boulders to Trace the Geographic Extents of Abrupt Climate Changes: The Younger Dryas Example - Abrupt Climate Change: Mechanisms, Patterns, and Impacts - page 115

Figure 3. (top) Fits of the moraine degradation model to the beryllium-10 exposure dates from the inner Titcomb Lakes
and Waiho Loop moraines [Gosse et al., 1995b; Barrows et al., 2007]. In each graph, the points with error bars represent
the exposure dates with their 1 σ measurement uncertainties; the curves are the best fit modeled distributions for each data
set. These fits presume an erosion rate for exposed boulders of 1.0 mm kyr 1 , consistent with earlier exposure dating
studies [e.g., Gosse et al., 1995b]. The initial height of the Titcomb Lakes moraine was prescribed at 25 m [cf. Gosse et al.,
1995a], and the initial height of the Waiho Loop moraine was prescribed to be 50 m [cf. Denton and Hendy, 1994].
(bottom) Effects of different sampling strategies on the resulting distributions of cosmogenic exposure dates. The
distributions are shown as box plots [Chambers et al., 1983]. Sampling very tall boulders (height greater than or equal
to 1.5 m) produces exposure dates that are within a few thousand years of the true age of the moraine (heavy line);
sampling boulders with minimal surface relief (relief less than or equal to 2.5 mm) produces exposure dates that are
thousands of years younger than the true age of the moraine.

4. IMPLICATIONS FOR FIELD SAMPLING CRITERIA gies could interfere with our ability to identify moraines

associated with abrupt climate changes.
Field workers use a variety of criteria to select samples for
cosmogenic nuclide measurements, but the two most com-
mon criteria are boulder height and surface freshness [e.g.,
Nishiizumi et al., 1989; Phillips et al., 1990; Cerling and
Craig, 1994; Gosse et al., 1995b; Fabel and Harbor, 1999;
Licciardi et al., 2001; Laabs et al., 2009]. Most field geo-
morphologists avoid sampling boulders below some mini-
mum height, often 1 m. Fresh boulders retain polish or
striations, if they were transported at the glacier bed. Few
moraine boulders have polish or striations when they are
sampled, so field geomorphologists estimate the thickness
of material eroded from each boulder by measuring the relief
on the boulders ' upper surfaces. For a single boulder, relief is
the distance between the lowest point and the highest point
on the boulder ' s upper surface, measured at right angles to
the sampled surface. The style of weathering varies with
lithology, but the low points on boulder surfaces are often
weathering pits, and the high points are often veins of

After identifying a moraine, the first step in cosmogenic
exposure dating is deciding which boulders to sample. De-
pending on which boulders are sampled, the measured con-
centrations either will or will not be representative of the
moraine ' s true age. Thus, imperfect boulder selection strate-

Table 2. Best Fits of the Degradation Model to the Waiho Loop
and Titcomb Lakes Data Sets a

Initial
Slope
(deg)

Kolmogorov-
Smirnov
Statistic

Moraine
Age (ka)

Diffusivity
(m 2 yr 1 )

Data Set

2.909 10 3 Waiho Loop 11.59 32.77 0.1400

6.883 10 4

Titcomb
Lakes

14.27 36.80 0.3113

a Data sets are from Barrows et al. [2007] and Gosse et al.
[1995b]. The parameter estimates here are probably good to two or
three significant figures. Four figures are reported to allow checking
of the model fits.