lable at ScienceDirect
ARTICLE IN PRESS
Quaternary Geochronology xxx (2008) 1–12
QUAGEO136_proof � 3 July 2008 � 1/12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Contents lists avai
Quaternary Geochronology
journal homepage: www.elsevier .com/locate/quageo
55
56
57
58
59
60
61
62
63
64
65
66
67
68
OOF
Research Paper
Determination of both exposure time and denudation rate from an insitu-produced 10Be depth profile: A mathematical proofof uniqueness. Model sensitivity and applications to natural cases
R. Braucher a,*, P. Del Castillo b, L. Siame a, A.J. Hidy c, D.L. Bourles a
a CEREGE UMR6635 Universite Paul Cezanne CNRS BP80 13545 Aix en Provence, Franceb Universite de Picardie Jules Verne, Laboratoire Amienois de mathematiques fondamentales et appliquees, CNRS UMR 6140, 33 r. Saint-Leu, 80039 Amiens Cedex 1, Francec Department of Earth Sciences, Dalhousie University, Edzell Castle Circle, Halifax NS, B3H 4J1 Canada
69
70
71
72
73
74
a r t i c l e i n f o
Article history:Received 21 June 2007Received in revised form 27 March 2008Accepted 4 June 2008Available online xxx
75
76
77
78
79
80
Keywords:Cosmogenic nuclide10BeDenudationTime
* Corresponding author. Tel.: þ33 4 42 97 15 09; faE-mail addresses: [email protected] (R. Bra
picardie.fr (P. Del Castillo), [email protected] (A.J. Hidy
1871-1014/$ – see front matter � 2008 Published bydoi:10.1016/j.quageo.2008.06.001
Please cite this article in press as: Braucher,depth profile: A mathematical proof..., Quat
81
82
83
CTEDPRa b s t r a c tMeasurements of radioactive in situ-produced cosmogenic nuclide concentrations in surficial materialexposed to cosmic rays allow either determining the long-term denudation rate assuming that thesurface studied has reached steady-state (where production and losses by denudation and radioactivedecay are in equilibrium) (infinite exposure time), or dating the initiation of exposure to cosmic rays,assuming that the denudation and post-depositional processes are negligible. Criteria for determiningwhether a surface is eroding or undergoing burial as well as quantitative information on denudation orburial rates may be obtained from cosmogenic nuclide depth profiles. With the refinement of thephysical parameters involved in the production of in situ-produced cosmogenic nuclides, a unique well-constrained depth profile now permits determination of both the exposure time and the denudation rateaffecting a surface. In this paper, we first mathematically demonstrate that the exponential decrease ofthe in situ-produced 10Be concentrations observed along a depth profile constrains a unique exposuretime and denudation rate when considering both neutrons and muons. In the second part, an improvedchi-square inversion model is described and tested in the third part with actual measured profiles.
� 2008 Published by Elsevier Ltd.
8485
86
E87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
UNCORR1. Introduction
Recently, the use of in situ cosmogenic nuclides has revolu-tionized Earth surface studies. They indeed not only allow surficialfeatures such as moraines, ‘‘stoss-and-lee’’ topography and alluvialfans to be dated for long periods as long as several million years, butalso allow distinguishing between both major dynamic processesaffecting surfaces (denudation and burial) and estimating theirrates (Siame et al., 2006). In the Earth’s environment, 10Be is mainlyproduced by interactions of primary cosmic ray particles (a parti-cles and protons) and their secondary particles (neutrons andmuons) with atmospheric nuclei of 14N and 16O. Although most ofthe cosmic ray’s energy is dissipated within the atmosphere,reducing cosmic rays intensity by almost 1000 from the top of theatmosphere to sea level, 10Be is also produced in the lithosphere byspallation of, for example, 16O, 27Al, 28Si, and 56Fe (Lal and Arnold,1995). The flux of the nuclear active particles efficiently dissipatestheir energy through cosmogenic nuclide producing nuclear
x: þ33 4 42 97 15 40.ucher), pierre.delcastillo@u-).
Elsevier Ltd.
R., et al., Determination of boernary Geochronology (2008
106
107
108
109
reactions in the lithosphere. Therefore cosmogenic nuclideproduction rates decrease exponentially with the mass of overlyingmaterial with, for each production mechanism, a characteristicattenuation length L (g/cm2).
The evolution of the cosmogenic nuclide production rate (P(x))as a function of depth x (expressed in g/cm2 in order to beindependent from the material density) is given by Eq. (1):
PðxÞ ¼ P eð�x=LÞ; (1)
where P is the surface production rate. However, two maintypes of secondary particles with significantly different atten-uation lengths, neutrons and muons, are involved in litho-spheric in situ-production. The effective production attenuationlength of neutrons, Ln, is indeed shorter (approximately 150 g/cm2) than that of muons: 1500 g/cm2 for negative muons Lm1
and 5300 g/cm2 for fast muons Lm2 (Braucher et al., 2003). Thisimplies that although neutron-induced production is dominantin the near-surface (Brown et al., 1995), muon-induced reactions become dominant below a few meters. Fora surface undergoing denudation (3 in g/cm2/yr) the evolutionof the 10Be concentrations (f) with time (t) and depth (x) is
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
E
Q1
Table 1Theoretical 10Be concentrations for a 2 m depth profile exposed for 20 ka un-dergoing a denudation rate of 10 m/Ma
Depth (g/cm2) 10Be (at/g)
0 262,88850 190,006100 137,742200 73,352300 40,164400 23,004
These concentrations come from Eq. (3) using a production rate of 15 at/g, a densityof 2 g/cm3; the relative contributions of neutrons and muons (negative and fast) tothe total 10Be production were n¼ 97.85%, m1¼1.5% and m2¼ 0.65% , the relativeattenuation lengths were Ln¼ 150 g/cm2, Lm1¼1500 g/cm2and Lm2¼ 5300 g/cm2
(Braucher et al., 2003).
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–122
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 2/12
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
UNCORRECT
commonly described by the following partial differentialequation (Lal, 1991):
df ¼ P dt eð�x=LÞ þ 3 dtdfdx� lf dt; (2)
with l the radioactive decay constant.Since the half-live of 10Be is short compared to the Earth’s age,
its primordial component has vanished and, in addition, if weassume that the rock studied has undergone a single cosmic rayexposure episode and had no cosmogenic nuclides at the beginningof the present exposure, then the initial concentration of cosmo-genic nuclides equals zero. Thus Eq. (2) may be solved to yield:
fxð3; tÞ ¼nP e�x=Ln
3Lnþ l
�1� e�tð3=LnþlÞ
�þm1P e�x=Lm1
3Lm1þ l
��
1� e�tð3=Lm1þlÞ�þm2P e�x=Lm2
3Lm2þ l
�1� e�tð3=Lm2þlÞ
�;
ð3Þ
where n, m1 and m2 are the relative contributions of neutrons andmuons (negative and fast) to the total 10Be production (m1¼1.5%,m2¼ 0.65% and n¼ 97.85%; Braucher et al., 2003), Ln, Lm1 and Lm2
are the relative attenuation lengths of neutrons and muons (neg-ative and fast).
1.1. Former approach
During the 1980s and early 1990s, published works using insitu-produced cosmogenic 10Be (Brown et al., 1994; Lal and Arnold,1995; Nishiizumi et al., 1986, 1991) were only based on the neutronproduced 10Be as physical parameters for muons were not wellconstrained (this was a reasonable approach as the surface pro-duction is dominated by neutrons). As a consequence, assumptionswere required to interpret the measurements. If field observationsor other evidence support the assumption of simple exposurehistory and negligible denudation, then a minimum exposure agemay be calculated using Eq. (4):
tmin ¼ �1l
ln�
1� lf ð0; tÞP
�; (4)
derived from Eq. (3) for 3¼ 0, n¼ 1 and m1¼m2¼ 0. By contrast, iffield evidence indicated an exposure time long enough to reach thesteady-state balance concentration, for example at the surfaces ofstable cratons, maximum denudation rates may be computed usingEq. (5):
3max ¼
Pfð0;NÞ
� l
!Ln; (5)
derived from Eq. (3) for t[½ð3=LnÞ þ l��1 and n¼ 1 andm1¼m2¼ 0.
Provided the steady-state concentration has been reached, thedenudation rate can also be deduced from Eq. (3) using the bestfit curve technique along a 10Be concentration depth profile; thevariable thus being the denudation rate. If the steady-state con-centration has not been reached, there is an infinity of exposuretime–denudation rate pairs that will fit Eq. (3), the 10Be concen-trations measured along a depth profile, when only neutrons areconsidered.
1.2. Use of depth profile
Nowadays, refinement of the physical parameters describing themuon and neutron interactions, in particular their significantlydifferent attenuation lengths (Heisinger et al., 2002a,b; Kim and
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
DPROOF
Englert, 2004) allow determination of the proportion of 10Be pro-duced by each type of particle. It becomes theoretically possible,using a single cosmogenic nuclide, to estimate both the exposureage and the denudation rate of surfaces affected by relativelyconstant denudation rates (Braucher et al., 2003; Siame et al.,2004). Because 10Be concentrations resulting from reactions withrapidly attenuated neutrons reach steady-state with respect todenudational loss much more rapidly than those resulting fromreactions with the more penetrating muons, 10Be produced at thesurface, which mainly results from interactions with neutrons,might be used to estimate the denudation rate and that produced atseveral meters depth, which mainly results from interactions withmuons to estimate the exposure time.
To illustrate why users have to consider all the particles impliedin the total production of 10Be when studying a depth profile, let usenvisage a 2 m depth profile (six sampling depths) exposed duringthe last 20 ka and undergoing a denudation rate of 10 m/Ma. Theexpected 10Be concentrations are then calculated using Eq. (3),involving both neutrons and muons, with a production rateof 15 at/g/a (Table 1). To model the thus theoretical 10Be depth-produced profile concentrations, the exposure ages implied bydifferent denudation rates (0, 4, 8, 10, 12, 15 m/Ma) were calculatedusing Eq. (3). This was done first considering only neutrons (formerapproach), and, second, considering both neutrons and the muons.For each scenario (time–denudation pair), the best fit wasdetermined by minimizing a chi-square value that is the sum ofeach individual chi-square determined at each depth:
chi-square ¼Xn
i¼1
�Ci � Cðxi; 3; tÞ
si
�2
; (6)
where Ci is the measured (or here the expected) 10Be concentrationat depth xi, C(xi, 3, t) is the modelled 10Be concentration determinedusing Eq. (3), si is the analytical uncertainty at depth i, and n is thetotal number of samples in the profile.
Results of the best fit curve are presented in Fig. 1.Considering that neutrons only yield to a constant chi-square for
any determined exposure age–denudation rate pair, implying thatthey all are a possible solution. In that case, there is no uniquesolution. Considering that both neutrons and muons yield signifi-cantly different chi-square, the minimum chi-square being associ-ated with the exposure age–denudation rate pair (20 ka–10 m/Ma)which corresponds to the input parameters used to construct the10Be concentrations depth profile. In that case, a unique solution isevident. But is this solution unique from a mathematical point ofview? This is what we intend to demonstrate in the following part;then an improved statistical model based on that of Siame et al.(2004) will be presented and tested on natural data sets in a lastsection.
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
CTEDPROOF
0.10
0.08
0.06
0.00
Ch
i-sq
uare
NeutronsNeutrons + Muons
NeutronsNeutrons + Muons
0 2 4 6 8 10 12 14 16
24000
23000
22000
21000
20000
19000
18000
17000
A
B
Ag
e (a
)
Denudation (m/Ma)
Fig. 1. Case study illustrating the solutions for a theoretically produced exponential down-concentrations profile (exposure age: 20 ka; denudation rate: 10 m/Ma). (A) Chi-squareversus denudation; (B) exposure ages versus denudation. Considering that neutrons only yield to a constant chi-square for any determined exposure age–denudation rate pair,implying that they all are a possible solution. In that case, there is no unique solution. Considering that both neutrons and muons yield to significantly different chi-square, theminimum chi-square being associated with the exposure age-denudation rate pair (20 ka–10 m/Ma; black arrow) which corresponds to the input parameters used to construct the10Be concentrations depth profile. In that case, a unique solution is evident.
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–12 3
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 3/12
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
UNCORRE2. Mathematical approach
Using Eq. (3) we suppose depth parameter x ˛ [0, 1000], and(e, t) ˛ [0, 10�1]� ]0, þN[ and we assume that P ˛; [5, 50]. Tosimplify the notations, we set
ax ¼ nP e�x=Ln ; bx ¼ m1P e�x=Lm1 ;gx ¼ m2P e�x=Lm2 ; (7)
and we introduce the functions
31ðeÞd1
eLnþ l
; 32ðeÞd1
eLm1þ l
; 33ðeÞd1
eLm2þ l
: (8)
Let us remark that the extrema of fx are given by
mðxÞdinf ðe;tÞ˛½0;10�1���0;þN½fxðe; tÞ ¼ 0 ; (9)
and
MðxÞd supðe;tÞ˛½0;10�1���0;þN½
fxðe; tÞ ¼ax þ bx þ gx
l: (10)
(this is achieved for the zero denudation steady-state equilibrium)According to Eqs. (9) and (10), we get
fxð3; tÞ˛�0;MðxÞ½; for all ð3; tÞ˛h0;10�1
i��0;þN½:
We consider two depths x1 and x2 with (x1, x2) ˛ [0, 1000]2. Then,for two 10Be concentrations N1 and N2 with N1 ˛ ]0, M(x1)[ and
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
N2 ˛ ]0, M(x2)[, N1 s N2 we can introduce the following nonlinearsystem in two unknowns (e, t) given by(
fx1ðe; tÞ ¼ N1;fx2ðe; tÞ ¼ N2:
(11)
Remark 2.1. From Eq. (3), we remark that x 1 fx(e, t) is decreasing.It results that for N1<N2 and x1� x2, system (11) does not admitsolution.
In the following, we establish the proposition:
Proposition 2.2. Let (x1, x2) ˛ [0, 1000]2, x1< x2, N1 ˛ ]0, M(x1)[ and N2 ˛ ]0, M(x2)[. Then, there exists at most a pair (e, t) ˛ [0,10�1]� ]0, þN[ satisfying system (11).
To prove this proposition, we proceed as follows. First, we studythe function fx defined in Eq. (3). We compute vfx=vt and vfx=ve andwe show that for all (e, t) ˛ [0, 10�1]� ]0, þN[, we haveðvfx=vtÞðe; tÞ > 0 and ðvfx=veÞðe; tÞ < 0. Then, we introduce e0(N1)defined by
e0ðN1Þdmaxn
e˛h0;10�1
i���dt > 0 satisfying fx1ðe; tÞ ¼ N1
o:
We establish that, for all e ˛ [0, e0(N1)], there is a unique t(e) suchthat fx1 ðe; tðeÞÞ ¼ N1. To obtain the uniqueness of the solution ofsystem (11), we prove that e1fx2 ðe; tðeÞÞ is strictly increasing.
We have fx ˛ CN([0, 10�1]� ]0, þN[). From Eq. (3), for all(e, t) ˛ [0, 10�1]� ]0, þN[, it derives that
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
E
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–124
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 4/12
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
UNCORRECT
vfxvtðe; tÞ ¼ ax e�t=e1ðeÞ þ bx e�t=e2ðeÞ þ gx e�t=e3ðeÞ; (12)
and
vfxveðe; tÞ ¼ �ax31ðeÞ2
Ln
�1� e�t=31ðeÞ
�þ ax31ðeÞ
Lnt e�t=31ðeÞ
� bx32ðeÞ2
Lm1
�1� e�t=31ðeÞ
�þ bx32ðeÞ
Lm1t e�t=31ðeÞ
� gx33ðeÞ2
Lm2
�1� e�t=33ðeÞ
�þ gx33ðeÞ
Lm2t e�t=33ðeÞ: (13)
From Eq. (12), as ax, bx and gx are positive, we have
vfxvtðe; tÞ > 0 for all ðe; tÞ˛
h0;10�1
i��0;þN
: (14)
On the other hand, we have on [0, 10�1]� ]0, þN[
v2fxvevtðe; tÞ ¼ �t
�ax
Lne�t=31ðeÞ þ bx
Lm1e�t=32ðeÞ þ gx
Lm2e�t=33ðeÞ
�< 0:
As ðvfx=veÞðe;0Þ ¼ 0 for all e ˛ [0, 10�1], it results that
vfxveðe; tÞ < 0; for all ðe; tÞ˛
h0;10�1
i��
0;þN
�: (15)
In the proof of Proposition 2.2, we use the following lemma:
Lemma 2.3. Let
ða1;a2;a3Þ�ðLn;Lm1;Lm2Þ; and ði;jÞ˛ð1;2Þ;ð1;3Þ;ð2;3Þ
�:
(16)
Let us consider the functions defined on [0, 10�1]� [0, þN[ by
gi;jðe;tÞ ¼ e�t=3iðeÞ �
3jðeÞ2
aj
�1�e�t=3jðeÞ
�þ
3jðeÞaj
t e�t=3jðeÞ!
�e�t=3jðeÞ �3iðeÞ2
ai
�1�e�t=3iðeÞ
�þ3iðeÞ
ait e�t=3iðeÞ
!: (17)
For all (i, j) satisfying Eq. (16), we have
gi;jðe; tÞ ¼ 0; for all ðe; tÞ˛h0;10�1
i�h0;þN
h:
The proof of Lemma 2.3 follows.
Proof of Lemma 2.3. Let us recall that
Ln < Lm1 < Lm2 : (18)
We set
ui;jðe; tÞ ¼ �e�t=3iðeÞ 3jðeÞ2
ajþ e�t=3jðeÞ 3iðeÞ2
ai; (19)
vi;jðe; tÞ ¼ e�tð1=3iðeÞþ1=3jðeÞÞ
3jðeÞ2
aj� 3iðeÞ2
ai
!; (20)
wi;jðe; tÞ ¼ e�tð1=3iðeÞþ1=3jðeÞÞ
t3jðeÞaj� t3iðeÞ
ai
!: (21)
Thus, gi, j¼ ui, jþ vi, jþwi, j. From Eq. (8), we have
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
DPROOF
3jðeÞ2
aj� 3iðeÞ2
ai¼�aj � ai
e2 � aiaj
�aj � ai
l2�
eþ ajl 2ðeþ ailÞ
2: (22)
According to Eq. (8), we have
ejðeÞaj� eiðeÞ
ai¼ �
�aj � ai
l�
eþ ajl ðeþ ailÞ
: (23)
From Eqs. (19)–(23), we get
ðeþailÞ2�eþajl
2�ui;jðe;tÞþvi;jðe;tÞþwi;jðe;tÞ
¼ e�t=3jðeÞ��etð1=3jðeÞ�1=3iðeÞÞajðeþailÞ
2þai�eþajl
2�
þe�t=3iðeÞ��l�aj�ai
tðeþailÞ
�eþajl
þ��
aj�ai e2
�aiaj�ai�aj
l2��: (24)
To show that the right-hand-side of Eq. (24) is positive, weintroduce the function hi, j, defined on [0, 10�1]� [0, 107] by
hi;jðe; tÞ ¼ � aj etð1=3jðeÞ�1=3iðeÞÞðeþ ailÞ2þai
�eþ ajl
2
þ e�t=3iðeÞ�� lt�aj � ai
ðeþ ailÞ
�eþ ajl
þ��
aj � ai e2 � aiaj
�ai � aj
l2�: ð25Þ
For (e, t) ˛ [0, 10�1]� [0, 107], we have
vhi;j
vtðe;tÞ�
aj�ai
aiee�ðaj�aiÞte=aiaj
�eþail
�2
þe�t=3iðeÞ�
l�aj�ai
ðeþailÞ
�eþajl
��1þt
�e
aiþl
���
�e�t=3iðeÞ��
aj�ai e2�aiaj
�aj�ai
l2�
e
aiþl
��: ð26Þ
Thus, as tðe=aiþlÞ>0, we get
vhi;j
vtðe;tÞ�e�ðaj�aiÞte=aiaj
�aj�ai
aie3þ2
�aj�ai
le2þ
�aj�ai
ail
2e
�
þe�t=3iðeÞ���aj�ai
e3
ai�2�aj�ai
le2�ai
�aj�ai
l2e
�:
We deduce easily that, for x ˛ [0, 10�1]� [0, þN[, we have
vhi;j
vtðe; tÞ �
�aj � ai
aie3 þ 2
�aj � ai
le2
þ ai�aj � ai
l2e
�e�t=3iðeÞ
�et=3jðeÞ � 1
�:
According to Eqs. (16) and (18), aj> ai. Thus, for t� 0, we get that
vhi;j
vtðe; tÞ � 0; on
h0;10�1
i��
0;þN
�:
It results that hi, j(e, t) is positive if hi, j(e, 0)� 0. From Eq. (25), wehave
hi;jðe;0Þ ¼ 0; for all e˛h0;10�1
i:
The proof of Proposition 2.2 follows.
Proof of Proposition 2.2. Let N1 ˛ ]0, M(x1)[. Let us introducee0(N1) defined by
e0ðN1Þdmaxn
e˛h0;10�1
i��� dt > 0 satisfying fx1ðe; tÞ ¼ N1
o:
(27)
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
E
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–12 5
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 5/12
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
UNCORRECT
In the following, we denote simply e0(N1) by e0. Let e ˛ [0, e0].From Eq. (15), e1fx1 ðe; tÞ is continuous and decreasing, then thereexists t1>0 such that fx1 ðe; t1Þ > N1 (we can choose t1¼ t(e0)).Moreover, fx1 ðe;0Þ ¼ 0, and t1fx1 ðe; tÞ is continuous and strictlyincreasing on [0, þN[. It results that there exists a unique t suchthat fx1 ðe; tÞ ¼ N1. We can conclude that there exists a function f
defined on [0, e0] by f(e)¼ t and satisfying
fx1ðe;fðeÞÞ ¼ N1; for all e˛½0; e0�: (28)
We have f ˛ C1([0, e0]). For e ˛ [0, e0(N1)], we consider the map
j : e1fx2ðe;fðeÞÞ � N2:
We want to prove that the solution of fx2 ðe;fðeÞÞ � N2 ¼ 0 isunique if it exists. The solution is unique if j is strictly increasing ordecreasing. For all e ˛ [0, e0], we have
j0ðeÞ ¼ vfx2
vtðe;fðeÞÞf0ðeÞ þ vfx2
veðe;fðeÞÞ:
From Eq. (28), we have
f0ðeÞ ¼ �vfx1
veðe;fðeÞÞ
vfx1
vtðe;fðeÞÞ
: (29)
According to Eq. (29), it results that the sign of j0 depends on thesign of
vfx1
vtðe;fðeÞÞj0ðeÞ ¼ �vfx2
vtðe;fðeÞÞvfx1
veðe;fðeÞÞ
þ vfx1
vtðe;fðeÞÞvfx2
veðe;fðeÞÞ:
To simplify the computations, for i ˛ {1, 2, 3}, we introduce
jiðe; tÞ ¼ e�t=3iðeÞ; ziðe; tÞ ¼ � 3iðeÞ2
ai
�1� e�t=3iðeÞ
�
þ3iðeÞai
t e�t=3iðeÞ; ð30Þ
where ai is defined in Eq. (16). From Eqs. (12), (13) and (30), we get
�vfx2
vtðe;fðeÞÞvfx1
veðe;fðeÞÞ þ vfx1
vtðe;fðeÞÞvfx2
veðe;fðeÞÞ
¼ ��ax2 j1 þ bx2
j2 þ gx2j3 ðe;fðeÞÞ
�ax1 z1 þ bx1
z2 þ gx1z3
�ðe;fðeÞÞ þ�ax1 j1 þ bx1
j2 þ gx1j3 ðe;fðeÞÞ
�ax2 z1 þ bx2
z2
þ gx2z3 ðe;fðeÞÞ
¼�ax1 bx2
� ax2 bx1
ðj1z2 � j2z1Þðe;fðeÞÞ þ
�ax1 gx2
� ax2 gx1
�ðj1z3 � j3z2Þðe;fðeÞÞ þ
�bx1
gx2� gx1
bx2
ðj2z3 � j3z2Þ
� ðe;fðeÞÞ:(31)
As x1< x2, from Eq. (7), we obtain
ax1 bx2� ax2 bx1
¼ nP2m1 e�x1=Ln e�x2=Ln
�e9x2=Lm1 � e9x1=Lm1
�> 0:
Similarly, we get
ax1 gx2� ax2 gx1
> 0; bx1gx2� gx1
bx2> 0: (32)
To get the sign of the right-hand-side of Eq. (31), we shouldstudy the sign of j1z2� j2z1, j1z3� j3z1, j2z3� j3z2. From Eqs. (17)and (30), we have
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
F
j1z2 � j2z1 ¼ g1;2; j1z3 � j3z1 ¼ g1;3; j2z3 � j3z2 ¼ g2;3
Applying Lemma 2.3 on [0, e0]� ]0, þN[, we deduce that
ðj1z2 � j2z1Þðe; tÞ > 0; ðj1z3 � j3z1Þðe; tÞ> 0; ðj2z3 � j3z2Þðe; tÞ > 0: (33)
According to Eqs. (31)–(33), it results that for all e ˛ [0, e0]
�vfx2
vtðe;fððeÞÞvfx1
veðe;fððeÞÞ þ vfx1
vtðe;fððeÞÞvfx2
veðe;fððeÞÞi0:
Thus
j0ðeÞ > 0; for all e˛½0; e0�: (34)
DPROO3. Model approach
Recently, Siame et al. (2004) showed that measurement of 10Beconcentrations along a depth profile allows estimating both ex-posure time and denudation rate using a chi-square inversionmodel based on Eq. (3). In the proposed model, however, the an-alytical uncertainties (1s) were not accounted for. Assuming thatthe analytical uncertainties have normal distribution centered onthe measured concentrations, we thus propose to improve the chi-square inversion model using Monte Carlo simulations. This wasperformed for 10Be but can be easily adapted for others nuclidesproviding that their production at depth can be easily modeled. (a)At least, 100 depth profiles are generated by randomly selectinga concentration within the concentration ranges defined by themeasured uncertainties (1s) at each sampling depth. (b) Loops onexposure time and denudation rate are performed and for eachtime – denudation pair a chi-square value is determined. Fora given profile and a time – denudation pair, this value is the sumof each individual chi-square determined at each depth (Eq. (6)).At the end, results can be stored in a (102� X) matrix, 102 columnscome from 100 simulated depth profiles plus two columns fordenudation and time, respectively, and X is the total number ofiterations (X ¼ ððemax � eminÞ=deÞððTmax � TminÞ=dTÞ with emax,emin and Tmax, Tmin as the ranges for denudation and time, re-spectively, and dT and de as the increments for time and de-nudation, respectively; these parameters are set by the user).Because the time–denudation pair that yields to the smallest chi-square value, corresponding theoretically to the best fit, may notbe statistically acceptable for all the simulated profiles, we preferto evaluate the median value for each time–denudation pair(contrary to the arithmetic mean, the median value makes itpossible to attenuate the disturbing influence of extreme values).This median value corresponds to the median of the 100 chi-square values determined for a given time–denudation pair. Theminimum median value gives the time–denudation pair solution.When the measured depth profile is close to an ideal exponentialdecrease, the absolute chi-square and median minimum valuesyield to the same time–denudation pair solution. All the pro-cedures described above are summarized in Fig. 2. Moreover, thisprocedure can be applied for different inherited 10Be concentra-tions. In that case, as previously, the solution corresponds to thelowest median value for a given triplet (denudation–exposureage–inheritance).
4. Model sensitivity
To test this model, theoretical depth profiles have been pro-duced along four increasing lengths (2, 4, 9 and 12 m). Thickness of
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
EDPROOF
Q2
Denudation
εmin
εmin
Tmin χ2
(εmin;Tmin)
Tmin+
dT
εmin+dε
Tmin+
(x+1).dT
+x.dT
Tmin
εmin+dε
χ2
(εmin;Tmin+dT)
χ2
(εmin+dε;Tmin+(x+1)dT)
χ2
(εmin+dε;Tmin+(x+1)dT)
χ2
(εmin+dε;Tmin+(x+1)dT)
(εmin+dε;Tmin+(x+1)dT)
Median
χ2
(εmin+dε;Tmin+xdT)
χ2
(εmin+dε;Tmin+xdT)
χ2
(εmin+dε;Tmin+xdT)
Tmax
Tmax - dTεmax
εmax χ2
(εmax;Tmax)χ2
(εmax;Tmax)χ2
(εmax;Tmax)
χ2
(εmax;Tmax - dT)
χ2
(εmax;Tmax - dT)
χ2
(εmax;Tmax - dT)
(εmax;Tmax)Median
Median(εmax;
Tmax - dT)
(εmin+dε;Tmin+xdT)
Median
Time Profile 1 Profile 100
χ2
(εmin;Tmin)χ2
(εmin;Tmin)
χ2
(εmin;Tmin+dT)χ2
(εmin;Tmin+dT)
Median(εmin;Tmin)
(εmin;Tmin+dT)Median
Profile 99 Median
Fig. 2. Matrix output of the proposed model. Columns 1 and 2 correspond to denudation rate and exposure time, respectively. Columns 3–102 correspond to depth profilesrandomly generated (here 100 profiles); the last column received the median value of the 100 chi-square values determined for a given time–denudation pair. The number of linesdepends on the chosen incremental steps for both denudation Q4(d3) and time (dT).
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–126
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 6/12
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
UNCORRECT
samples is ignored in this model and is considered to be part ofa post-treatment of the field data as well as all the productioncorrections. Each profile affected by a denudation rate of 10 m/Mahas been virtually exposed 20 and 800 ka, the selected productionrate being 15 at/g/a. Inheritance has not been considered here but isthoroughly discussed in Section 5. 10Be concentrations are sum-marized in Table 2.
To simulate the 1s measurement uncertainties, three Gaussiannoise values were applied to these concentrations: 3, 5 and 10%,corresponding, assuming a 2% variability on accelerator massspectrometry, to 2000, 455 and 104 10Be counts, respectively.Because sample preparation is time consuming and acceleratormass spectrometry measurement is costly, one has to consider theappropriate number of samples to be collected in the field. In thisstudy six samples per profile have been considered. Generatingtheoretical profiles with fewer samples will not alter significantlythe conclusions but in natural studies, because nature doesnot always ‘‘behave’’ as theory, sampling too few samples perprofile can be hazardous (possible outliers, several exposurehistories .). Field experience lets us think that sampling a mini-mum of six samples per profile is a reasonable compromise thatgenerally allows constraining the expected exponential decreaseof production with depth for an unperturbed natural marker thathas undergone a simple exposure history. Model outputs for theprofiles exposed 20 ka and 800 ka with a denudation rateof 10 m/Ma are presented in Table 3 and Table 4, respectively(Figs. 3 and 4).
4.1. Denudation rates
For each individual test, the median values for both de-nudation rate and exposure time agree with the inputs. However,
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
and obviously, as shown in Fig. 5 and by the minimum andmaximum values in Tables 3 and 4, denudation estimates arecloser to the input values when both exposure time and profilelength increase. Because, as demonstrated in Section 2 there isa unique pair (exposure time–denudation) when consideringboth neutrons and muons, minimization of the range of de-nudation rate (minimum and maximum values) is directly linkedto the muon effect. When profile lengths are not deep enough,the production by neutrons in the total 10Be production is pre-dominant and even if muon production is considered in themodels, this yields to a wide range of acceptable solutions. Thepart of muon production increasing with depth yields to a re-duction of acceptable solutions representing more accurately thelong-term denudation rate. The question is now: how deep tosample? The answer has to deal with theory and practical feasi-bility and first depends on the aim of the sampling. To study olderoding surfaces, a 2 m depth profile may be sufficient because inthat case, exposure time, long enough to reach steady-state bal-ance concentrations, can be neglected. On the contrary, whenworking on young surfaces, time usually cannot be neglected(unless denudation rates are so high that steady-state is reachedin few hundred years). In that case, the question of depth iscrucial. Theoretically speaking the deepest is the best. However,from Fig. 5 it appears that denudation rate estimates derivedfrom 9 and 12 m long depth profiles are not significantly differ-ent. But from a practical point of view, access to such a deepprofile may be very difficult. From Fig. 5 and from Table 3, one candeduce that a 4 m depth profile may be suitable and physicallyreachable, providing that the precision of the measurements isbetter than 5% which roughly corresponds to 500 counts in the10Be detector, assuming a 2% variability on accelerator massspectrometry.
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
E
OF
Table 2Depth profile 10Be concentrations used as input in models comprising six samples per profile
Depth (cm) 200 cm 400 cm 900 cm 1200 cm
20 ka 800 ka 20 ka 800 ka 20 ka 800 ka 20 ka 800 ka
0 262,888 1,216,246 262,888 1,216,246 262,888 1,216,246 262,888 1,216,24625 190,006 910,99350 137,742 691,338 137,742 691,338100 73,352 418,744 73,352 418,744150 40,164 275,870 40,164 275,870200 23,004 199,762 23,004 199,762 23,004 199,762300 9388 134,240 9388 134,240400 5494 110,019500 4197 97,433 4197 97,433750 30,95 77,991800 2957 74,902900 2710 69,2241000 2491 64,1301200 2122 55,404
The material density has been set at 2 g/cm3, the denudation rate at 10 m/Ma, the production rate at 15 at/g/a and exposure times at 20 and 800 ka.
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–12 7
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 7/12
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
RRECT
5. Application to natural data sets
To consider our work to be of real value, it must be now shownthat the model is in fact robust with real data. Three publisheddepth profiles have been used: one from a quartz sandstone corecollected in Antarctica (Brown et al., 1992) to illustrate long-termexposure history on an autochthonous bedrock and two othersfrom alluvial terraces studies (Brocard et al., 2003; Hidy et al.,2005).
5.1. Long-term exposure history
In their paper, Brown et al. (1992) have measured cosmogenic10Be and 16Al as a function of depth in a core of quartz bedrockcollected in South Victoria Land, Antarctica. Although this studywas dedicated to calculate the effective attenuation length ofcosmic rays producing 10Be and 16Al, Brown et al. also presentedthe dependence of calculated ages on assumed denudation ratesusing production rate of 24.7 at/g/a for 10Be based on Lal (1991).An updated production rate of 27.4 at/g/a based on Stone (2000)(using Antarctic air pressure) is used here to correct Brown et al.ages. These ages are ranging from 1.02 to 1.38 Ma with denudationrates ranging from 0 to 0.22 m/Ma. Interestingly, the authorsconcluded that the assumption of steady-state denudation maynot be valid for their samples. Using our model, the exposure agesrange from 1.07 to 1.12 Ma with a denudation rate of 0 m/Ma. This
UNCO
Table 3Model results for the profile exposed 20 ka with a denudation rate of 10 m/Ma
Depth(m)
Uncertainty (%) Mediandenudation(m/Ma)
Min and maxdenudation(m/Ma)
Mediantime (a)
Min and maxtime (a)
2 3 9.52 0; 19.32 19,885 17,692; 22,8555 10.33 0; 27.12 20,296 17,392; 26,869
10 11.97 0; 27.23 20,401 17,312; 26,436
4 3 9.98 6.43; 14.64 19,999 19,160; 21,0235 9.83 0.51; 17.54 19,874 18,311; 21,749
10 9.82 0; 24.27 19,886 18,130; 22,985
9 3 9.83 6.74; 13.68 19,965 19,515; 20,5065 9.8 3.4; 16.8 19,945 19,262; 20,811
10 7.63 0; 23.56 19,632 18,092; 21,587
12 3 10.11 5.89; 12.59 20,002 19,401; 20,4595 9.66 4.39; 15.06 19,950 19,189; 20,456
10 11.04 1.19; 20.52 19,991 18,213; 21,544
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
DPRO
result is in excellent agreement with those of Brown et al. andleads us to conclude that the samples of this core are not yet atsteady-state, which at this locality would require 6 Ma of expo-sure to be reached. This implies that the negligible erosion as-sumption by Brown et al. is valid and that the calculated ages arenot minimum ages but correspond to the effective exposureduration.
5.2. Mid-term exposure history: alluvial terraces dating
Antarctic depth profiles are ideal to assess long-term exposurehistory but most of the applications on alluvial terraces or alluvialfans concern younger exposure ages. In that case, a precise agedetermination of deposits using cosmogenic nuclides may bea problem because of inheritance, bioturbation or anthropogenicactivities. The work of Brocard et al. (2003) is used to test ourmodel on young alluvial terraces and details of the geomorphologyand soils at the site are described in their article. This article isa study of river incision as a response to tectonic and climaticcontrols. The authors estimate terrace ages using a Monte Carlotechnique applied on pebble depth profiles by considering as un-known values, the inheritance, exposure time and soil density.Because field observations such as terrace structures provide evi-dence for the pristine nature of the surface, no denudation factorwas applied in this article. The main difference between the Bro-card et al. model and ours is that in their model Brocard et al.
Table 4Model results for the profiles exposed 800 ka with a denudation rate of 10 m/Ma
Depth(m)
Uncertainty(%)
Mediandenudation(m/Ma)
Min and maxdenudation(m/Ma)
Mediantime (a)
Min and maxtime (a)
2 3 9.98 9.59; 10.43 793,834 711,751; 947,9985 10.06 9.43; 10.6 796,059 681,155; 1,037,320
10 10.01 9.03; 11.21 790,376 547,918; 1,208,807
4 3 10.04 9.63; 10.34 804,771 745,188; 866,0965 10.01 9.4; 10.54 794,978 728,642; 889,965
10 9.95 8.86; 10.93 775,400 672,978; 1,009,216
9 3 10.01 9.62; 10.36 798,295 762,729; 842,0635 10.1 9.43; 10.72 800,089 736,920; 873,999
10 9.91 8.91; 11.42 778,310 689,429; 920,510
12 3 10.02 9.64; 10.38 802,195 771,146; 835,7115 9.99 9.37; 10.57 796,283 746,137; 854,197
10 9.92 8.63; 11.44 793,717 678,399; 897,794
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
RRECTEDPROOF
3
4m
Denudation rate (m/Ma)
0 5 10 15 20 25 30
Tim
e (a)
10%5%3%
2m
Denudation rate (m/Ma)
0 5 10 15 20 25 30
Tim
e (a)
16 000
20 000
24 000
28 000
9m
Denudation rate (m/Ma)
0 5 10 15 20 25 30
12m
Denudation rate (m/Ma)
0 105 15 20 25 30
Tim
e (a)
Tim
e (a)
16 000
20 000
24 000
28 000
16 000
20 000
24 000
28 000
16 000
20 000
24 000
28 000
Fig. 3. Model outputs for the profiles exposed 20 ka with a denudation rate of 10 m/Ma. Four profile lengths have been simulated (2, 4, 9 and 12 m depth). Each symbol representsthe best pair (denudation–exposure time) associated to the minimum chi-square for one profile (100 profiles per uncertainty level). Grey dots, black squares, and open trianglesrefer to 3, 5 and 10% uncertainty in measurement, respectively. Median values are reported in Table 3.
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–128
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 8/12
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
UNCOconsider as a possible solution for each profile the triplet (time,
inheritance and soil density) that yields to an absolute minimum.That is the reason why they mention that more than one bestsolution can be found (one by random profile). On the contrary,our model approach is based, for a given inheritance and soildensity, on the best pair (denudation–time) that yields to theminimum median value applied on all the random profiles. Ourmodel thus includes the Brocard et al. model to which one post-statistical treatment (the median value) is applied. When data arewell distributed along an exponential decrease and when analyt-ical uncertainties are low, the solution deduced by the minimummedian value and by the absolute minimum value given by eachprofile are the same within uncertainties. Because the Drac 1 siteof Brocard et al. (2003) presents the longest depth profile(292 cm), it has been chosen to compare the two models. Fromthat site, Brocard et al. estimate an exposure age ranging from 7.2to 14 ka with inheritance ranging from 10 to 24 kat/g for a soildensity of 2 g/cm3. The median model applied on Drac 1 sitepredicts an estimated denudation rate that ranges from 11.4 to14.4 m/Ma, an exposure age ranging from 9 to 11 ka, for an
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
inheritance of 12 kat/g and a soil density of 1.85 g/cm3. All un-certainties are circumscribed as described in Granger (2006)which Qfollows Bevington and Robinson (2003) by increasing byone the minimum chi-square median value. Our results are in goodagreement with those of Brocard et al. except of course for thedenudation rate that had been set to zero by Brocard.
To test our model on older fill terraces, 10Be concentrationsmeasured within a depth profile from Lees Ferry terraces (Arizona)are used (Hidy et al., 2005). Table 5 presents the available data forthat site. At that location, the terrace top is extremely flat, notdissected by post-depositional gulleying. Desert pavement is ex-tremely well developed at the location of the pit, meaning thatthere are no large clasts that remained unfragmented (into smallpebbles), the pebbles are well interlocked, their tops are extremelyvarnished, and their bottoms are deeply rubefied. The site was wellaway from the edge of the terrace which is visibly influenced bydiffusion along the escarpment rim. This terrace fragment is iso-lated from any flooding or fluvial stripping because there is a deeplyincised tributary just upstream and there is a deep gulley sepa-rating the terrace from the bank of the Grand Canyon. Rainfall is so
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
ORRECTEDPROOF
2m
Denudation rate (m/Ma)
8 9 10 11 12
Tim
e (a)
600 000
800 000
1 000 000
1 200 0004m
Denudation rate (m/Ma)
8 9 10 11 12
9m
Denudation rate (m/Ma)
8 9 10 11 12
12m
Denudation rate (m/Ma)
8 9 10 11 12
5%10%3%
Tim
e (a)
600 000
800 000
1 000 000
1 200 000T
im
e (a)
600 000
800 000
1 000 000
1 200 000
Tim
e (a)
600 000
800 000
1 000 000
1 200 000
Fig. 4. Model outputs for the profile exposed 800 ka with a denudation rate of 10 m/Ma. Four profile length have been simulated (2, 4, 9 and 12 m depth). Each symbol representsthe best pair (denudation–exposure time) associated to the minimum chi-square for one profile (100 profiles per uncertainty level). Grey dots, black squares, and open trianglesrefer to 3, 5 and 10% uncertainty in measurement, respectively. Median values are reported in Table 4.
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–12 9
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 9/12
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
UNCrare and soil moisture so low that there is no opportunity for sur-
face runoff (the surface area of the terrace is too small and thegradient is too low). Pedogenic carbonate can be observed in thesoil but there are no fragments of the carbonate on the soil surface.From their data, Hidy et al. obtain a minimum exposure age of87.3� 2.1 ka and an inheritance of 109 kat/g, assuming no de-nudation. An average soil density of 2.5 g/cm3 was used based onfield measurements (although seemingly high, this value may beexplained by the presence of boulder-sized quartz-rich clasts in thesoil pit). The surface production rate of 10.01 at/g/a is deduced fromStone (2000) and is corrected from topographic shielding. The samedepth profile concentrations were modeled using our model withthe same soil density. The best pair (denudation–exposure age) was6.8 m/Ma and roughly 300 ka with an inherited 10Be concentrationof 105 kat/g (Table 6). This means that the steady-state is almostreached for this surface. However, this amount of denudation is notsupported by field observations and the chronology by OSL
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
(independent OSL data have constrained the abandonment of theLees Ferry M4 terrace to 77–98 ka). Such a denudation rate over300 ka would have implied a total of 2.1 m of erosion which isimpossible considering fieldwork evidences. Furthermore, U-seriesand OSL chronologies imply that this terrace cannot be three timesolder than the age they measured. Considering the measured 10Beconcentrations, the only way to lower the age and denudation is toconsider that the samples are closer to the surface than they arewhen applying a density of 2.5 g/cm3. The only way to reduce thesample depth is to consider a lower soil density. A second modelforcing the density to 2 g/cm3 was performed yielding to a de-nudation rate of 1 m/Ma an exposure age of 86 ka with an inherited10Be concentration of 91 kat/g. This is in a better agreement withfield observations but evidenced that the density is a crucial pa-rameter when one wants to deduce from the analysis of the 10Beconcentration distribution along a depth profile the exposure timeand the denudation rate affecting a surface deposit. Then, because
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
RRECTEDPROOF
2m
Denudation rate (m/Ma)
0 5 10 15 20
Tim
e (a)
9m
Denudation rate (m/Ma)
0 5 10 15 20
Tim
e (a)
4m
Denudation rate (m/Ma)
0 5 10 15 20
12m
Denudation rate (m/Ma)
0 5 10 15 20
20 ka800 ka
600000
800000
1000000
15000
20000
25000
15000
20000
25000
600000
800000
1000000
Fig. 5. Effect of exposure time (20 ka (black dots and left axis) and 800 ka (grey dots and right axis)) and depth on the data modeled using 3% uncertainty. See text for explanation.
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–1210
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 10/12
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
NCOthe evolution of measured 10Be concentrations with depth is asexpected by the theory, the last choice was to consider the densityas a ‘‘free’’ parameter as denudation, time and inheritance are.Doing this the best fit of all from a statistical point of view wasobtained for a density of 2.2 g/cm3, a denudation rate of 6.1 m/Ma,an exposure age of 153 ka with an inherited 10Be concentration of99 kat/g (Table 6). This model yields an exposure duration that is
UTable 5Lees Ferry sample identification and 10Be concentrations
Sample name Depth (cm) 10Be (at/g)
GC-04-LF-401 0 949,302� 24,097GC-04-LF-404.30s 27.5 629,031� 15,813GC-04-LF-404.60s 57.5 449,824� 11,295GC-04-LF-404.100s 97.5 323,221� 8283GC-04-LF-404.140s 137.5 224,598� 6024GC-04-LF-404.180s 177.5 173,873� 4518GC-04-LF-404.220s 217.5 148,423� 4142
AMS measurement has been performed at Laurence Livermore National Laboratory;corrected production rate deduced from Stone (2000) is 10.0 at/g/a.
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
double that of Hidy et al. and would have implied a 93 cm loss ofsoil by erosion. Fig. 6 is the representation of Lees Ferry data set andthe 10Be concentrations modeled using parameters presented inTable 6.
Table 6Comparison of model outputs for the Lees Ferry terrace depth profile
Hidy et al. Our model
Density (g/cm3) 2.5 2 2.5 2.2Denudation (m/Ma) 0 0.95 6.8 6.1Exposure time 87,300 86,000 299,000 153,000Inher. 10Be at/g 109,000 91,000 105,000 99,000Associated minimum
chi-square52.37 14.87 12.09 8.12
Hidy et al. (2005) obtained a minimum exposure age as they assume no de-nudation. In our two first attempts, density was forced to be 2 g/cm3 and 2.5 g/cm3.The last 2.2 g/cm3 density is the one determined by the model and for which theminimum median chi-square value is obtained overall. It is important to note thatHidy et al. did not use the surficial desert pavement sample (GC-04-LF-401) in theirmodel.
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
EPROOF
10Be (at/g)
0 200000 400000 600000 800000 10000000
50
100
150
250
200
Dep
th
(cm
)
Lees Ferry datasetHidy et al. modelOur model (d=2.5 g/cm3)Our model (d=2 g/cm3)Our model (d=2.2 g/cm3)
Fig. 6. Lees Ferry (Arizona) 10Be concentrations at depth. Black dots refer to measured concentrations with their associated error bars; grey dots refer to Hidy et al. model (adenudation of 0 m/Ma; an exposure time of 87.3 ka, an inheritance of 109 kat/g and a density of 2.5 g/cm3); open dots refer to our model using a density of 2.5 g/cm3, an exposureage of 86 ka, an inhertance of 91 kat/g and a denudation rate of 0.95 m/Ma; open triangle refer to our model using a density of 2 g/cm3, an exposure age of 299 ka, an inheritance of105 kat/g and a denudation rate of 6.8 m/Ma and open squares refer to our best model using a density of 2.2 g/cm3, an exposure age of 153 ka, an inheritance of 99 kat/g anda denudation rate of 6.1 m/Ma. In this last case, density, exposure age, denudation rate and inheritance were considered as free parameters whereas in the other models density wasa fixed value. It is important to note that Hidy et al. did not use the surficial desert pavement sample (GC-04-LF-401) in their model but in this Figure, we put the surficialconcentration implied by Hidy et al. parameters.
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–12 11
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 11/12
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
UNCORRECT
6. Conclusion
In this paper, we have mathematically proved that for two mea-sured 10Be concentrations N1 at depth x1 and N2 at depth x2, froma surface undergoing denudation and a single exposure to cosmicrays, only one pair (exposure time–denudation rate) is necessary todefine the system. This is not true considering neutrons only. Thedemonstration has been made using all the particles (neutrons, fastand stop muons) and their associated physical parameters involvedin the production of in situ-produced 10Be. Providing that the profilehas not been perturbed, since all the samples will have the sameexposure history at their specific depth, this demonstration based ontwo concentrations N1 at depth x1 and N2 at depth x2 randomlychosen, has been extended and tested to more than two concentra-tions. To do so, an improved model of the chi-square inversion modelof Siame et al. (2004) has been developed. This Monte Carlo modelrandomly generates a large number of depth profiles. For a givenprofile and a quadruplet (time–denudation–density and in-heritance), the sum of each individual chi-square determined at eachdepth is calculated. Then, considering all the generated depth profilesfor a given quadruplet (time–denudation–density and inheritance),a median value corresponding to the median of all the chi-squarevalues determined with the same quadruplet is calculated. Theminimum median value gives the time–denudation pair solution.When the measured depth profile is close to an ideal exponentialdecrease and analytical uncertainties are better than 5%, the absolutechi-square minimum and median minimum values yield to the samesolution. Although the model can help provide a unique solution forthe quadruplet, it will always be necessary for the user to considerthe geological evidence to evaluate the model output.
Acknowledgements
The authors thank J. Pederson and J. Gosse for having providedunpublished data set from Lees Ferry terraces as well as twoanonymous reviewers. Editorial handiling by (John C. Gosse).
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
D
References
Bevington, P., Robinson, K., 2003. Data Reduction and Error Analysis for the PhysicalSciences. Mc Graw-Hill Higher Education, 336 pp.
Braucher, R., Brown, E.T., Bourles, D.L., Colin, F., 2003. In situ produced 10Be mea-surements at great depths: implications for production rates by fast muons.Earth and Planetary Science Letters 211, 251–258.
Brocard, G.Y., van der Beek, P.A., Bourles, D.L., Siame, L.L., Mugnier, J.-L., 2003. Long-term fluvial incision rates and postglacial river relaxation time in the FrenchWestern Alps from 10Be dating of alluvial terraces with assessment of in-heritance, soil development and wind ablation effects. Earth and PlanetaryScience Letters 209, 197–214.
Brown, E.T., Bourles, D.L., Colin, F., Raisbeck, G.M., Yiou, F., Desgarceaux, S., 1995.Evidence for muon-induced production of 10Be in near-surface rocks from theCongo. Geophysical Research Letters 22 (6), 703–706.
Brown, E.T., Bourles, D.L., Colin, F., Sanfo, Z., Raisbeck, G.M., Yiou, F., 1994. The de-velopment of iron crust lateritic systems in Burkina Faso, West Africa examinedwith in-situ-produced cosmogenic nuclides. Earth and Planetary Science Letters124, 19–33.
Brown, E.T., Brook, E.J., Raisbeck, G.M., Yiou, F., Kurz, M.D., 1992. Effectiveattenuation lengths of cosmic rays producing 10Be and 26Al in quartz:implications for exposure age dating. Geophysical Research Letters 19 (4), 369–372.
Granger, E., 2006. A review of burial dating methods using 10Be and 26Al: in situ-produced cosmogenic nuclides and quantification of geological processes.Geological Society of America Special Paper 415, 1–16.
Heisinger, B., Lal, D., Jull, A.J.T., Kubik, P., Ivy-Ochs, S., Neumaier, S., Knie, K.,Lazarev, V., Nolte, E., 2002a. Production of selected cosmogenic radionuclidesby muons; 1. Fast muons. Earth and Planetary Science Letters 200, 345–355.
Heisinger, B., Lal, D., Jull, A.J.T., Kubik, P., Ivy-Ochs, S., Knie, K., Nolte, E.,2002b. Production of selected cosmogenic radionuclides by muons; 2.Capture of negative muons. Earth and Planetary Science Letters 200, 357–369.
Hidy, A.J., Pederson, J.L., Cragun, W.S., Gosse, J.C., 2005. Cosmogenic 10Be exposuredating of Colorado river terraces at Lees Ferry, Arizona. Geological Society ofAmerica Abstracts 37–7, 296.
Kim, K.J., Englert, P.A.J., 2004. Profiles of in situ 10Be and 26Al at great depths at theMacraes Flat, East Otago, New Zealand. Earth and Planetary Science Letters 223,113–126.
Lal, D., 1991. Cosmic ray labeling of erosion surfaces: in situ nuclide productionrates and erosion models. Earth and Planetary Science Letters 104, 424–439.
Lal, D., Arnold, J.R., 1995. Tracing quartz through environment. Earth and PlanetaryScience Letters 91, 1–5.
Nishiizumi, K., Kohl, C.P., Arnold, J.R., Klein, J., Fink, D., Middleton, R., 1991. Cosmicray produced 10Be and 26Al in Antarctic rocks: exposure and erosion history.Earth and planetary Science Letters 104, 440–454.
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001
1429
R. Braucher et al. / Quaternary Geochronology xxx (2008) 1–1212
ARTICLE IN PRESS QUAGEO136_proof � 3 July 2008 � 12/12
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
Nishiizumi, K., Lal, D., Klein, J., Middleton, R., Arnold, J.R., 1986. Production of 10Beand 26Al by cosmic rays in terrestrial quartz in situ and implications for erosionrates. Nature 319, 134–136.
Siame, L., Bourles, D.L., Brown, E.T., 2006. In situ-produced cosmogenic nuclides andquantification of geological processes. Geological Society of America SpecialPaper 415, 146.
UNCORRECTE
Please cite this article in press as: Braucher, R., et al., Determination of bodepth profile: A mathematical proof..., Quaternary Geochronology (2008
Siame, L., Bellier, O., Braucher, R., Sebrier, M., Cushing, M., Bourles, D.L., Hamelin, B.,Baroux, E., de Voogd, B., Raisbeck, G., Yiou, F., 2004. Local erosion rates versusactive tectonics: cosmic ray exposure modelling in Provence (South-EastFrance). Earth and Planetary Science Letters 220 (3–4), 345–364.
Stone, J.O., 2000. Air pressure and cosmogenic isotope production. Journal ofGeophysical Research 105, 23753–23759.
DPROOF
th exposure time and denudation rate from an in situ-produced 10Be), doi:10.1016/j.quageo.2008.06.001