Geoscience Reference
In-Depth Information
Table 1.
Beryllium-10 Exposure Dates Recalculated Following Barrows et al. [2007]
a
Nucleon
10
Be
Production Rate
(atoms g
1
yr
1
)
Muon
10
Be
Production Rate
(atoms g
1
yr
1
)
Boulder Height
(m)
Apparent Age
(years)
1
σ
Uncertainty
(years)
Sample ID
Waiho Loop Moraine, Western New Zealand [Barrows et al., 2007]
WH-01B
5.554
0.157
1.24
0.42
WH-02
5.580
0.157
10.33
0.70
WH-03
5.576
0.157
10.70
0.38
WH-04B
5.575
0.157
9.65
0.47
WH-05
5.472
0.156
9.10
0.41
WH-08A
5.473
0.156
6.85
0.91
WH-09
5.473
0.156
5.30
0.33
WH-10
5.475
0.156
11.33
1.55
Inner Titcomb Lakes Moraine, Wind River Range [Gosse et al., 1995a]
WY-92-138
1.0
51.892
0.568
12.80
0.38
WY-92-139
1.0
51.892
0.568
12.06
0.36
WY-92-140
1.0
51.892
0.568
9.93
0.30
WY-93-333
1.0
51.892
0.568
12.08
0.36
WY-93-334
1.5
51.892
0.568
14.00
0.42
WY-93-335
0.6
51.892
0.568
12.97
0.39
WY-93-336
2.0
51.892
0.568
13.30
0.40
WY-93-337
1.5
51.892
0.568
13.26
0.40
WY-93-338
0.8
51.892
0.568
12.86
0.39
WY-93-339
0.1
51.892
0.568
12.73
0.38
WY-92-138
1.0
51.892
0.568
12.80
0.38
WY-92-139
1.0
51.892
0.568
12.06
0.36
WY-92-140
b
1.0 51.892 0.568 9.93 0.30
a
Production rates and exposure dates are recalculated following Barrows et al. [2007], using the scaling model of Stone [2000] for both
nucleon and muon production. The 1
σ
uncertainties reflect measurement uncertainty only.
beryllium-10 exposure dates; our recalculated dates agree
with those reported by Barrows et al. [2007] to within
0.6%, suggesting that our calculation method is consistent
with theirs. We did not use the CRONUS online calculator
[Balco et al., 2008] because the calibration of the online
calculator depends in part on the concentration measure-
ments from the inner Titcomb Lakes moraine (see below).
Thus, using the online calculator would introduce circularity
into our results. In any case, the choice of scaling model has
little influence on the scatter among exposure dates from
individual moraines [Balco et al., 2008; Applegate, 2009],
even at midlatitude sites where the effects of geomagnetic
field changes are greatest.
Both of these data sets are likely influenced by geomorphic
processes. The reduced
v
tent with moraine degradation than either measurement error
alone or inheritance [Applegate et al., 2010].
Explicit fitting of the degradation model to these data sets
also suggests that moraine degradation is responsible for
most of the scatter in each data set (Figure 3, top, and Table
2), although the model fit to the inner Titcomb Lakes data set
is poor. For the purposes of these fits, we prescribed the initial
height of each moraine and the erosion rate of the exposed
boulders (1.0 mm kyr
1
[Gosse et al., 1995a, 1995b]). The
constant-erosion-rate assumption is consistent with prior ex-
posure dating studies that correct for the effects of boulder
erosion [e.g., Gosse et al., 1995b; Kelly et al., 2008]. We then
used the differential evolution genetic algorithm to search for
the minimum value of the Kolmogorov-Smirnov test statistic
[Press et al., 2005; Clauset et al., data set, 2007]. The model
evaluation with the minimum KS statistic indicates the va-
lues of moraine age, initial moraine slope, and topographic
diffusivity that are most consistent with each data set. We
specified the initial moraine heights for these model inver-
sions because the distributions of cosmogenic exposure dates
are insensitive to the initial height of the moraine above some
minimum value [Applegate et al., 2010].
2
scores of these data sets are much
greater than 1 (Figure 2), indicating that the data sets contain
more scatter than can be explained by measurement error
alone. We cannot rule out the possibility that these data sets
are drawn from normal distributions because the observa-
tions fall reasonably close to a line on normal probability
plots (Figure 2). However, both data sets have skewnesses
less than
0.5, and these skewness values are more consis-
Search WWH ::
Custom Search