Sediment burial dating using terrestrial cosmogenic nuclides

Burial dating using in situ produced terrestrial cosmogenic nuclides is a relatively new method to date sediments and quantify geomorphological processes such as erosion, accumulation and river incision. Burial dating utilises the decay of previously in situ produced cosmogenic nuclides and can be applied to sedimentary deposits such as cave fillings, alluvial fans, river terraces, delta deposits, and dunes. Using the established 10Be/26Al nuclide pair allows numerical dating of quartz bearing material from ~100 ka to 5 Ma, where other dateable material is often unavailable. To date, a number of studies have demonstrated the successful application of in situ produced cosmogenic nuclides in various scientific disciplines, such as Quaternary geology, geomorphology and palaeoanthropology. However, insufficiently defined physical properties such as nuclide half lives and complex depth dependent nuclide production rates result in relatively large uncertainties. Nevertheless, burial dating represents a promising method for determining numerical ages. [Datierung des Überdeckungsalters mit Hilfe von terrestrischen kosmogenen Nukliden] Kurzfassung: Die Methode der Bestimmung des Überdeckungsalters mit Hilfe von in situ produzierten terrestrischen kosmogenen Nukliden stellt ein verhältnismäßig neues Datierverfahren dar. Es ermöglicht die Altersbestimmung von Sedimenten und damit die Quantifizierung von geomorphologischen Prozessen, wie Erosion, Akkumulation und Flusseintiefung. Das Verfahren bedient sich dabei des Zerfalls von zuvor in situ produzierten kosmogenen Nukliden und kann auf sedimentäre Ablagerungen wie Höhlenfüllungen, Schwemmfächer, Flussterrassen, Deltaschüttungen und Dünen angewendet werden. Durch die Verwendung des erprobten 10Be/26Al Nuklidpaares erlaubt die Methode die Bestimmung eines numerischen Alters von quarzführendem Material über einen Zeitbereich von ~100 ka bis 5 Ma. In diesem Zeitabschnitt ist datierfähiges Material für andere Methoden oftmals nicht oder nur unzureichend vorhanden. Viele Studien konnten bereits die erfolgreiche Anwendung von in situ produzierten kosmogenen Nukliden in den verschiedensten wissenschaftlichen Bereichen, darunter zum Beispiel in der (Quartär)Geologie, Geomorphologie und Paläoanthropologie, belegen. Dennoch können die zur Zeit nur ungenügend genau bestimmten physikalischen Größen, wie zum Beispiel die Nuklidhalbwertszeiten oder die tiefenabhängigen Nuklidproduktionsraten zu vergleichsweise großen Unsicherheiten führen. Trotz dieser Nachteile stellt die Methode eine vielversprechende Möglichkeit der numerischen Altersbestimmung dar.


Introduction
Sedimentary archives, for example marine, terrestrial and glacial deposits, provide information regarding the climatic and environmental history as well as the tectonic development of a given area.In this context, the youngest part of the Earth' history, the Neogene and the Quaternary, are of particular interest.This time have been characterised by massive mountain forming processes, large changes in the temperature of oceans and the atmosphere as well as by important biological evolution, particularly with the appearance of early and modern hominids (ASFAW et al. 1999;CLARK et al. 2003;DÈ-ZES et al. 2004;RAVELO et al. 2004;GIBBARD et al. 2005).All this information, however, is only of minor value if it can not be integrated into a global chronological framework.Only with reliable dating can such comparisons be made and the dating of sediments of the past million years is, therefore, one of the most important tasks in modern Quaternary research.Dating of terrestrial sediments over long time periods, however, can often be highly complex, imprecise or even impossible.Available methods, for example radiocarbon and luminescence dating, cover up to 50 ka and a few 100 ka, respectively, only a relatively short part of the Quaternary (PREUSSER et al. 2008, HAJDAS 2008).We are presenting here a brief summary on recent developments in using terrestrial cosmogenic nuclides for dating the burial age of sediments.Burial dating is based on the assumption that rocks that have been exposed to cosmic radiation for a given time are enriched with various cosmogenic radionuclides (Fig. 1).When these rocks, or their erosional products, are shielded from cosmic radiation, nuclide production ceases and the radionuclides decay according to their individual half lives (Fig. 2).Shielding may be due to transportation of sediment into a cave, by covering with overburden or by deposition in a deep water body (Fig. 3).By measuring the concentration of two nuclides (at least one radionuclide), it is possible to date the time elapsed since shielding.
Cosmogenic nuclides are produced by nuclear reactions in the Earth's atmosphere as well as in the upper parts of the Earth's crust.These nuclear reactions are initiated by high energy secondary cosmic radiation (LAL & PETERS 1967).The development of accelerator mass spectrometry (AMS) over the past few decades has enabled the detection of very small amounts of cosmogenic nuclides (FINKEL & SUTER 1993) and is fundamental to the effective use of cosmogenic nuclides in modern geosciences.The radionuclides 10 Be and 26 Al are good candidates for burial dating: both are produced in situ in quartz and possess relatively long half lives.Quartz is one of the most common minerals in the Earth's crust and so is present in almost all sedimentary deposits.The long half lives of both radionuclides enable dating over a time period of 5 Ma and, in addition to this, the nuclide production ratio has been quantified and is independent of latitude, altitude, depth below surface and time (GOSSE & PHILLIPS 2001).For these reasons, the method of burial dating is independent of production rate changes in time and space (LAL & ARNOLD 1985;KLEIN et al. 1986;LAL 1991), which causes most of the inaccuracies in surface exposure dating (GOSSE & PHILLIPS 2001;IVY-OCHS & KOBER 2008).The principle of burial dating was proposed by LAL & ARNOLD (1985) and underwent a major revision by LAL (1991).The first application of burial dating published (KLEIN et al. 1986) was able to show, with the analysis of the 26 Al/ 10 Beratio, that the Libyan Desert Glass Field was occasionally covered by shifting sand dunes, although it was not possible at this time to give discrete burial ages for the sampled glass.GRANGER et al. (1997) were the first to date a deposition event by using the 26 Al/ 10 Be-ratio.They determined river down-cutting rates for the past 1.5 Ma using relocated sediments from cave fillings.Since these first applications a few studies have demonstrated the successful application of burial dating.In addition to the more application focused studies, reviews by GRANGER & MUZIKAR (2001) and more recently by GRANGER (2006) provide the basic knowledge needed for age calculations in various burial scenarios.
Sediment burial dating using terrestrial cosmogenic nuclides As already mentioned, the method of burial dating with cosmogenic nuclides is based on the decay of radionuclides.This requires dependable accuracy in half life determination and precision of the nuclide measurement itself.The half life of 26 Al of 0.702 ± 0.056 Ma (MIDD- LETON et al. 1983), 0.705 ± 0.024 Ma (NORRIS et al. 1983), or 0.716 ± 0.032 Ma (RIGHTMIRE et al. 1958) is well defined and widely accepted in the cosmogenic nuclide community.However, the traditionally recognised half life of 10 Be of 1.51 ± 0.06 Ma (HOFMANN et al. 1987) is still undergoing discussion.Some authors (e.g.PARTRIDGE et al. 2003;HÄUSELMANN & GRANGER 2005;GRANGER et al. 2006) believe that this figure is too high and suggest a shorter 10 Be half life of ~1.34 Ma (MIDDLETON et al. 1993).NISHIIZUMI et al. (2007) support this, having recently re-evaluated the commonly used ICN (ICN Chemical & Radioisotope Division) and NIST (National Institute of Standards and Technology) reference material, suggesting that the 10 Be half life should be lowered to 1.36 ± 0.07 Ma.On the other hand, FINK & SMITH (2007) also re-evaluated the same material but hesitated to lower the 10 Be half life, stating that the direct and accurate specific activity measurement of the parent solution of both standards is needed to calculate the 10 Be half life, but this is not yet available.
2 In situ production of cosmogenic 10

Be and 26 Al
The Earth's atmosphere is undergoing permanent bombardment by primary cosmic radiation.This high energy nucleon radiation originates mainly in the Milky Way (with E ≈ 1 -10 10 GeV), with a much smaller fraction descended from beyond our galaxy (E < 10 20 eV) (GOSSE & PHILLIPS ANDREAS DEHNERT & CHRISTIAN SCHLÜCHTER 2001).Interactions of incoming primary radiation with atoms of the Earth's atmosphere result in low-energy particles, traditionally known as "secondary cosmic radiation" (E ≈ 100 MeV).
For the in situ production of 10 Be and 26 Al, both at the rock surface and subsurface level, there are two kinds of relevant particles: nucleons (neutrons and protons) and muons (fast and negative).2001), among others.Nuclide production in near surface material is dominated by nucleons (~97.5 %)(HEISINGER 1998) and changes with increasing depth, as muons display greater penetration depth as a result of their lower reactivity.The nucleonic production as a function of depth can be shown as a simple exponential law (LAL 1991;BROWN et al. 1992;GOSSE & PHILLIPS 2001): with production rate in depth in atoms g -1 a -1 (P(z)), scaled surface production rate in atoms g -1 a -1 (P(0)), depth in cm (z), attenuation length in g cm -2 (Λ) and density of the overburden in g cm -3 (ρ).The index of 10 and 26 represents the nuclides 10 Be and 26 Al, respectively, and N nucleonic production.Constants for all equations are listed in Tab. 1.The nuclide production by muons is valid within a depth range of 200 -5000 g cm -2 and is well described by the sum of three exponential functions (GRANGER & SMITH 2000).At shallower depths nuclide production by muons is negligible, as nucleonic production is predominant.Nuclide production by negative muon capture is described in the first two terms.The last one represents production by Coulomb-reactions (GRANGER & SMITH 2000;GRANGER & MUZIKAR 2001): The index μ represents muonic production, + and -show fast and negative muons respectively.The scaled surface production rate, sometimes called local production rate, is dependent on altitude above sea level and latitude, due to the effects of the Earth's magnetic field on the se-tion of cosmogenic 10 Be and 26 Al sphere is undergoing permanent bombardment by primary cosmic radiation.nucleon radiation originates mainly in the Milky Way (with E � 1 a much smaller fraction descended from beyond our galaxy (E < 10 20 eV) IPS 2001).Interactions of incoming primary radiation with atoms of the re result in low-energy particles, traditionally known as secondary cosmic 00 MeV).For the in situ production of 10 Be and 26 Al, both at the rock surface vel, there are two kinds of relevant particles: nucleons (neutrons and ns (fast and negative).According to strict nomenclature rules, muons belong rate in depth in atoms g -1 a -1 (P(z)), scaled surface production rate in atoms g - in cm (z), attenuation length in g cm -2 (�) and density of the overburden in dex of 10 and 26 represents the nuclides 10 Be and 26 Al, respectively, and N tion.Constants for all equations are listed in Tab. 1.
uction by muons is valid within a depth range of 200 5000 g cm -2 and is the sum of three exponential functions (GRANGER & SMITH 2000).At nuclide production by muons is negligible, as nucleonic production is clide production by negative muon capture is described in the first two terms.

� � �
The index � represents muonic production, + and show fast and n respectively.
The scaled surface production rate, sometimes called local product altitude above sea level and latitude, due to the effects of the Earth secondary cosmic rays and the attenuation effect of the atmosphere for the sampling position is also under debate at the moment, and i in LAL (1991), DUNAI (2000), STONE (2000), WAGNER et al. (2000WAGNER et al. ( (2001)), and references therein.
The nuclides 10 Be and 26 Al are also produced in situ by non-cosmo reactions.SHARMA & MIDDLETON (1989) stated that only �-induce provide a significant portion of 10 Be and 26 Al.Lithium as target nu ( 7 Li(�,p) 10 Be) is normally present at trace levels (�g g -1 ) but sodium ( 23 Na(�,p) 26 Al), is nearly always present at concentrations of one o 10 Be production is quite small compared to the cosmogenic compo negligible, while that of 26 Al may be significant.BROWN et al. (199 the steady-state concentrations of radiogenically produced 26 Al in a ~6.0 x10 4 at g -1 .This low concentration, with respect to steady-stat cosmogenically produced 26 Al of ~3.2 x10 7 at g -1 (Fig. 1 & 2), will samples with a short exposure history.Beside this, no relevant ra contamination in quartz has ever been observed and reported so far The index � represents muonic production, + and show fast and negative muons respectively.
The scaled surface production rate, sometimes called local production rate, is dependent on altitude above sea level and latitude, due to the effects of the Earths magnetic field on the secondary cosmic rays and the attenuation effect of the atmosphere itself.Its value and scaling for the sampling position is also under debate at the moment, and is discussed in more detail in LAL (1991), DUNAI (2000), STONE (2000), WAGNER et al. (2000), DESILETS & ZREDA (2001), and references therein.
The nuclides 10 Be and 26 Al are also produced in situ by non-cosmogenic such as radiogenic reactions.SHARMA & MIDDLETON (1989) stated that only �-induced nuclear reactions could provide a significant portion of 10 Be and 26 Al.Lithium as target nuclei for production of 10 Be ( 7 Li(�,p) 10 Be) is normally present at trace levels (�g g -1 ) but sodium, as a source of 26 Al ( 23 Na(�,p) 26 Al), is nearly always present at concentrations of one or more percent.Hence, 10 Be production is quite small compared to the cosmogenic compound and therefore negligible, while that of 26 Al may be significant.BROWN et al. (1991), however, argued that the steady-state concentrations of radiogenically produced 26 Al in average sandstones is ~6.0 x10 4 at g -1 .This low concentration, with respect to steady-state concentrations of cosmogenically produced 26 Al of ~3.2 x10 7 at g -1 (Fig. 1 & 2), will only be significant in samples with a short exposure history.Beside this, no relevant radiogenic 26 Al contamination in quartz has ever been observed and reported so far.
The index � represents muonic production, + and show fast and negative muons respectively.
The scaled surface production rate, sometimes called local production rate, is dependent on altitude above sea level and latitude, due to the effects of the Earths magnetic field on the secondary cosmic rays and the attenuation effect of the atmosphere itself.Its value and scaling for the sampling position is also under debate at the moment, and is discussed in more detail The nuclides 10 Be and 26 Al are also produced in situ by non-cosmogenic such as radiogenic reactions.SHARMA & MIDDLETON (1989) stated that only �-induced nuclear reactions could provide a significant portion of 10 Be and 26 Al.Lithium as target nuclei for production of 10 Be ( 7 Li(�,p) 10 Be) is normally present at trace levels (�g g -1 ) but sodium, as a source of 26 Al ( 23 Na(�,p) 26 Al), is nearly always present at concentrations of one or more percent.Hence, 10 Be production is quite small compared to the cosmogenic compound and therefore negligible, while that of 26 Al may be significant.BROWN et al. (1991), however, argued that the steady-state concentrations of radiogenically produced 26 Al in average sandstones is ~6.0 x10 4 at g -1 .This low concentration, with respect to steady-state concentrations of cosmogenically produced 26 Al of ~3.2 x10 7 at g -1 (Fig. 1 & 2), will only be significant in samples with a short exposure history.Beside this, no relevant radiogenic 26 Al contamination in quartz has ever been observed and reported so far.Sediment burial dating using terrestrial cosmogenic nuclides condary cosmic rays and the attenuation effect of the atmosphere itself.Its value and scaling for the sampling position is also under debate at the moment, and is discussed in more detail in LAL (1991), DUNAI (2000), STONE (2000), WAG-NER et al. (2000), DESILETS & ZREDA (2001), and references therein.The nuclides 10 Be and 26 Al are also produced in situ by non-cosmogenic such as radiogenic reactions.SHARMA & MIDDLETON (1989) stated that only α-induced nuclear reactions could provide a significant portion of 10 Be and 26 Al.Lithium as target nuclei for production of 10 Be ( 7 Li(α,p) 10 Be) is normally present at trace levels (μg g -1 ) but sodium, as a source of 26 Al ( 23 Na(α,p) 26 Al), is nearly always present at concentrations of one or more percent.Hence, 10 Be production is quite small compared to the cosmogenic compound and therefore negligible, while that of 26 Al may be significant.BROWN et al. (1991), however, argued that the steady-state concentrations of radiogenically produced 26 Al in average sandstones is ~6.0 x10 4 at g -1 .This low concentration, with respect to steady-state concentrations of cosmogenically produced 26 Al of ~3.2 x10 7 at g -1 (Fig. 1 & 2), will only be significant in samples with a "short" exposure history.Beside this, no relevant radiogenic 26 Al contamination in quartz has ever been observed and reported so far.

Application of burial dating
The following chapter looks at different kinds of burial situation and is designed to give an overview of the required formulas, followed by selected examples.For more detailed information see GRANGER & SMITH (2000) and GRANGER & MUZIKAR (2001).

Single stage exposure history -cave sediments
The simplest case of burial is provided by cave sediments (Fig. 4), in that shielding from cosmic rays occurs quickly and effectively to inwashed sediments.The great advantage of these deposits is that they remain almost unaffected by cosmic radiation and thus only negligible post burial production results.Assuming steady-state erosion conditions for the sediments, the burial age is directly dependent on nuclide half-lives.Thus today's 26 Al/ 10 Be ratio gives us the burial age (t in years) solving equation 5 and 6 for 10 Be and 26 Al. [5] [6] Equation 5 and 6 can be combined to equation 7

Application of burial dating
The following chapter looks at different kinds of burial situation and is designed to give overview of the required formulas, followed by selected examples.For more detailed information see GRANGER & SMITH (2000) and GRANGER & MUZIKAR (2001).

Single stage exposure history cave sediments
The simplest case of burial is provided by cave sediments (Fig. 4), in that shielding from cosmic rays occurs quickly and effectively to inwashed sediments.

3 Application of burial dating
The following chapter looks at different kinds of burial situation and is designed to give overview of the required formulas, followed by selected examples.For more detailed information see GRANGER & SMITH (2000) and GRANGER & MUZIKAR (2001).

Single stage exposure history cave sediments
The simplest case of burial is provided by cave sediments (Fig. 4), in that shielding from cosmic rays occurs quickly and effectively to inwashed sediments.Cave infill sediments were used to identify accelerated incision rates (up to 1.2 km Ma -1 ) in the River Aare valley, between 0.8 to 1.0 Ma, in the Siebenhengste-Hohgant cave system, Switzerland.Archaeological, as well as geological, questions can be solved using burial dating with 10 Be and 26 Al (see also AKÇAR et al. 2008).
Dating of embedded bones is mostly done using U/Th dating of surrounding calcite layers (PICKERING et al. 2007).In the absence of speleothems or with unsuitable U/Th con- The following chapter looks at different kinds of burial situation and is designed to give an overview of the required formulas, followed by selected examples.For more detailed information see GRANGER & SMITH (2000) and GRANGER & MUZIKAR (2001).

Single stage exposure history cave sediments
The simplest case of burial is provided by cave sediments (Fig. 4), in that shielding from cosmic rays occurs quickly and effectively to inwashed sediments.The great advantage of these deposits is that they remain almost unaffected by cosmic radiation and thus only negligible post burial production results.Assuming steady-state erosion conditions for the sediments, the burial age is directly dependent on nuclide half-lives.Thus todays 26 Al/ 10 Be ratio gives us the burial age (t in years) solving equation 5 and 6 for 10 Be and 26 Al.GRANGER et al. (2001) conducted similar studies at the Mammoth Cave multi level system in Kentucky, USA.By analysing the 26 Al/ 10 Be ratio of 29 quartz gravel and sand deposits throughout the cave system and from its surface, they were able to reconstruct a water table history for the nearby Green and Ohio Rivers over the past 3.5 Ma.Furthermore, they erent kinds of burial situation and is designed to give an 1 N (t) st to use the 26 Al-10 Be pair to date burial events of les found in various caves within dolomites along the New isburg, Virginia, USA.The pebbles, quartz vein remnants Rivers headwaters, were sampled in numerous cave levels . By dating the inwash events of these pebbles from 0.  Sediment burial dating using terrestrial cosmogenic nuclides ditions, dating can also be done with the help of inwash sediments containing quartz.

Multiple stage exposure history -Nuclide profiling
Post burial nuclide production can not be ignored where sediment has only few metres of overburden, as can be done in the cave scenario described above.The post burial production is depth-and time-dependent (see chapter 2) and thus the relevant equations for both nuclides become: [8] [9] Following GRANGER & SMITH (2000) the above integrals can be expressed as equation 10: [10] The index i represents either the nuclide 10 Be or 26 Al.Each of equation 8 and 9 contains three unknowns (burial age, inherited nuclide concentration and erosion rate).Thus, a combination of [8] and [9] can not be solved uniquely by only analysing 10 Be and 26 Al in one sample at a specific depth.The solution for this problem lies in sampling a vertical profile within the deposit.By analysing two samples from the same profile and considering suitable constraints for the inherited nuclide concentration [8] and [9] can be solved uniquely.These assumptions are site-specific and may comprise of, for example, a constant inherited nuclide concentration 4.02 ± 0.27 Ma.Although, these burial ages are in good agreement t debate (see WALKER et al. 2006).

Multiple stage exposure history Nuclide profiling
Post burial nuclide production can not be ignored where sediment h overburden, as can be done in the cave scenario described above.2000) the above integrals can be exp flowstones dated the sediment layer containing hominid fossils into a timeframe ranging from 3.22 to 3.58 Ma. PARTRIDGE et al. (2003) dated this layer using the 26 Al/ 10 Be ratio to 4.17 ± 0.14 Ma in the Silberberg Grotto and the previously undated fossil layer in Jacovec Cavern to 4.02 ± 0.27 Ma.Although, these burial ages are in good agreement they are currently under debate (see WALKER et al. 2006).

Multiple stage exposure history Nuclide profiling
Post burial nuclide production can not be ignored where sediment has only few metres of overburden, as can be done in the cave scenario described above.The post burial production is depth-and time-dependent (see chapter 2) and thus the relevant equations for both nuclides become: Following GRANGER & SMITH (2000) the above integrals can be expressed as equation 10: flowstones dated the sediment layer containing hominid fossils into a 3.22 to 3.58 Ma. PARTRIDGE et al. (2003) dated this layer using the 26 0.14 Ma in the Silberberg Grotto and the previously undated fossil lay 4.02 ± 0.27 Ma.Although, these burial ages are in good agreement th debate (see WALKER et al. 2006).

Multiple stage exposure history Nuclide profiling
Post burial nuclide production can not be ignored where sediment has overburden, as can be done in the cave scenario described above.2000) the above integrals can be expr (2003).The fossils are encased in a breccia of dolomite, chert and surface soil that accumulated as debris dropped into the cave.PARTRIDGE et al. (2003) utilised this fact to obtain burial ages from the quartz bearing material.Palaeomagnetic signals in calcitic flowstones dated the sediment layer containing hominid fossils into a timeframe ranging from 3.22 to 3.58 Ma. PARTRIDGE et al. (2003) dated this layer using the 26 Al/ 10 Be ratio to 4.17 ± 0.14 Ma in the Silberberg Grotto and the previously undated fossil layer in Jacovec Cavern to 4.02 ± 0.27 Ma.Although, these burial ages are in good agreement they are currently under debate (see WALKER et al. 2006).

Multiple stage exposure history Nuclide profiling
Post burial nuclide production can not be ignored where sediment has only few metres of overburden, as can be done in the cave scenario described above.The post burial production is depth-and time-dependent (see chapter 2) and thus the relevant equations for both nuclides become: Following GRANGER & SMITH (2000) the above integrals can be expressed as equation 10: The index i represents either the nuclide 10 Be or 26   & SMITH (2000).They measured the 10 Be and 26 Al concentrations of nine samples in a ~10 m profile of a river terrace of Old Kentucky River at Rice Station, Kentucky, USA.The sandy sediment was rapidly deposited, perhaps at the shores of a rising proglacial lake, and has remained exposed since deposition.By solving an equation similar to [8] and [9] and leastsquare-fitting of the solutions, they dated the terrace formation to 1.50 +0.32 / -0.25 Ma with a post-depositional terrace erosion rate of 6.2 ± 0.2 m Ma -1 .This is in good agreement with the nearby Green River incision around 1.5 Ma (GRANGER et al. 2001, see above).This initial study was followed by dating of alluvial deposits above the San Juan River near Bluff and Mexican Hat, Utah, USA by WOLKO-WINSKY & GRANGER (2004).Their cosmogenic 10 Be and 26 Al data suggest a deposition age of 1.36 +0.20 / -0.15 Ma and an erosion rate of 14 ± 4 m Ma -1 for the ~12 m thick Bluff site.Additional to t and ε, they allowed a fit within the model for bulk density and the addition of, at most, 20 cm material on top of the sampled profile.They were not able to realise dates for the Mexican Hat site due to insufficient burial depth of only 5.5 m following GRANGER & MU-ZIKAR (2001).The more recent work of HÄUSELMANN et al. (2007) utilises the nuclide profiling technique Sediment burial dating using terrestrial cosmogenic nuclides to date glacially related gravel successions in the Bavarian foreland of the Alps, Germany, at an abandoned gravel quarry near Bad Grönenbach, and in a second site, Böhener Feld.Both sites are made up of a mixture of limestone and quartz gravels in a carbonate matrix.The pioneer work on alpine Quaternary geology by PENCK & BRÜCKNER (1909) named these gravels "Deckenschotter" (cover gravels) and placed them within the classical Mindelian (Bad Grönbach) and the Günzian (Böhner Feld) glaciation, respectively.The age and terrace erosion rate determination of HÄUSELMANN et al. (2007) follows the nuclide profiling approach of GRANGER & SMITH (2000), as described above, followed by a χ 2 minimisation.This resulted in a fitted burial age of 0.68 +0.23 / -0.24 Ma and a terrace erosion rate of 123 +139 / -32 m Ma -1 at Bad Grönenbach, and 2.35 +1.08 / -0.88 Ma and an erosion rate of 18 +10 / -4 m Ma -1 for the section at Böhener Feld.This work was the first attempt to absolutely date these gravel deposits and demonstrates the suitability of the burial dating approach using 10 Be and 26 Al within the time span of the Quaternary, although relatively high measuring uncertainties, due to the abundance of common Al impurities in the processed quartz, result in large model uncertainties and subsequently in large errors.This work also demonstrated that sample preparation and nuclide measurements are still delicate tasks.
The three examples described above, can explain the build-up of terraces in a single depositional event, which may occur over hundreds or even thousands of years.Cosmogenic burial dating of more complex deposit evolution is also possible, as long as a well constrained stratigraphy is available.In such a sedimentary sequence, each package will have its own deposition age and burial history that is determined by the deposition age of the following layer.BALCO et al. (2005) describe the age determination of such a multiple package set-up.They dated a profile containing three palaeosols buried beneath Laurentide Ice Sheet sediments and a loess cover.At the Musgrove clay pit, Missouri, USA, two tills, the Atlanta and Moberly formations, overlie deeply weathered bedrock, as well as locally derived colluvium of the Whippoorwill formation (Fig. 5).BALCO et al. (2005) developed a three step burial history for the Musgrove section with rapid burial and moderate surface erosion (Fig. 5).Using Monte Carlo simulations in MATLAB's Optimization Toolbox enables minimisation of χ 2 discrepancies between model and measurements.In doing so, they determine a burial age of 2.41 ± 0.14 Ma for the lower till section (Atlanta formation) and, one between 1.6 -1.8 Ma for the upper till (Moberly formation).It should be noted, however, that it was necessary to assume some initial conditions in order to complete this well-developed model.BALCO et al. (2005) chose an age of 125 ka for the deposition of the covering loess layer based on regional stratigraphy (identified by t 3 in Fig. 5).This assumption consequently affects all other age calculations in the used stepwise model (age of Atlanta till = t 1 + t 2 + t 3 ).This approach is, therefore, not an independent age determination as is normally assumed with the burial dating method, although it should be noted that a stratigraphical age of 125 ka for the loess is convincing and that BALCO et al. (2005) subsequently correlated the age of 2.41 Ma with a (global) ice sheet build up from 2.7 to 2.4 Ma, as suggested by marine oxygen isotope data (JOYCE et al. 1993).

Complex sediment burial histories
The history of sediments in the natural environment is often highly complex.The material undergoes several kinds of transportation usually involving multiple stages of exposure and burial from its source until its modern day position.It is often impossible to describe the complete transportation history with confidence.In these situations the in situ cosmogenic nuclides 10 Be and 26 Al may provide useful information regarding, whether sediment has been buried or not.In some cases it is possible to evaluate cumulative burial and exposure durations.The first use of the 10 Be/ 26 Al ratio was done by KLEIN et al. (1986).They looked at 12 Libyan ANDREAS DEHNERT & CHRISTIAN SCHLÜCHTER Dessert Glass samples, subdivided into three groups.Fission-track dating placed the age at 28.5 Ma, but in situ cosmogenic 10 Be and 26 Al concentrations did not agree with this estimate.KLEIN et al. (1986) attempted to solve this discrepancy with regard to the history of the Libyan Dessert Glass and the distribution of nuclide concentrations within the individual groups.By analysing the ratio of 10 Be and 26 Al they could clearly show that burial by sand dunes is the most plausible explanation for the obtained nuclide concentrations.Furthermore they presented the first (minimum) burial ages based on the simplification of a single exposure followed by a single burial.The article of KLEIN et al. (1986) marks the advent of "burial dating with in situ produced cosmogenic nuclides".A second interesting application of the 10 Be/ 26 Al ratio in the early stage of burial dating development is documented by ALBRECHT et al. (1993), working on volcanic ash-flow tuffs at the Pajarito plateau of the Valles caldera, New Mexico, USA.These surfaces part of the Tshirege member of the Bandelier Tuff, which was deposited during a caldera eruption at ~1.14 Ma.The Tshirege member can be subdivided into four subunits, each with its own erosive character.The analysis of nuclide concentrations of 10 Be and 26 Al identified two to five times lower concentrations than a 1.14 Ma exposure history should yield, which was mainly due to erosion and previous burial of the investigated surfaces.ALBRECHT et al. (1993) suggested a burial period of 0.6 ± 0.3 Ma for most of the studied samples by using the 26 Al/ 10 Be ratio.The former shielding of the sampled surface is explained by an eroded tuff and soil cover consisting of the overlaying subunits of the Tshirege member with an average thickness of ~6 m.Burial by overlaying sediment is not the only explanation for a complex burial history.The shielding material does not necessarily have to be rock or sediment to result in the attenuation of secondary cosmic ray particle energy, which is dependent on the density of the overburden material.Thus any matter may shield cosmic radiation, with rock causing the most effective attenuation followed in order by sediment, water and glacial ice (Fig. 6).Any shielding history is clearly shown by the position of the sample on a two-isotope diagram, for example 26 Al/ 10 Be ratios vs. 10 Be concentrations (Fig. 7), referred to as a "steadystate erosion island plot" (LAL 1991), "simple exposure island plot" or "banana plot".The two-isotope diagram allows the detection of any shielding of a sample during its history.The "zero erosion line" and the "steady-state erosion line" frame a banana shaped area on a plot with logarithmic abscissa referred to Sediment burial dating using terrestrial cosmogenic nuclides as the steady-state erosion island (LAL 1991).All samples with a nuclide inventory resulting from simple exposure will plot within this area, while samples that have also undergone some shielding will fall below that.A determination of burial duration is only possible for a single burial event.A discrete observation of multiple shielding events is not possible with the steadystate erosion island plot, as the measured nuclide inventory is a composite of all exposure and shielding periods.Nuclide concentrations and production rates have to be normalised to sea level, high latitude (> 60°; SLHL) values to avoid any affects of production rate changes resulting from scaling to sample location (for more detailed discussion of two-isotope diagrams see LAL 1991;BIERMAN et al. 1999;GOSSE & PHILLIPS 2001).Shielding by material other than sediment or rock was studied by BIERMAN et al. (1999), who measured 10 Be and 26 Al concentrations at two locations at the northern and southern margin of the former Laurentide Ice Sheet.The northern site around Pangnirtung on Baffin Island, Canada consists of in situ deeply weathered Precambrian gneissic bedrock.In the area around these tor samples neither bedrock nor boulders display striations, grooves or chattermarks, although erratic cobbles and boulders confirm the presence of glacial ice.The southern part near Pipestone, Minnesota, USA is made up of very erosion resistant Sioux Quartzite.Its surface Fig. 7: Two-isotope diagram for in situ produced cosmogenic 10 Be and 26 Al in quartz as proposed by LAL (1991) with selected burial isochrones (dotted) and decay lines (dashed).Production rates are normalised to SLHL values with a 26 Al/ 10 Be production ratio of 6.1.Decay lines are given for complete shielding after single exposure with selected erosion rates (in mm ka -1 ) until reaching saturation.Samples that have been shielded will plot below the steady-state erosion line in that the 26 Al/ 10 Be ratio will decrease resulting from the shorter half live of 26 Al with respect to 10 Be.Used parameters are listed in Table 1.
Abb. 7: Zwei-Isotopen-Diagram für in situ produziertes 10 Be und 26 Al in Quarz nach LAL (1991), mit ausgewählte Überlagerungsisochronen (gepunktet).Produktionsraten sind normalisiert auf SLHL-Bedingungen mit einem 26 Al/ 10 Be Produktionsverhältnis von 6,1.Zerfallsgeraden (gestrichelt) sind dargestellt für den Fall der vollständigen Abschirmung nach einmaliger, kontinuierlicher Bestrahlung für ausgewählte Erosionsraten (in mm ka -1 ).Proben die während oder nach ihrer Bestrahlung abgeschirmt wurden, plotten unter der Gleichgewichts-Erosionsgeraden, da das 26 Al/ 10 Be Verhältnis abnimmt, gemäß der kürzeren Halbwertszeit des 26 Al Nuklids.Zu Grunde liegende Parameter können Tabelle 1 entnommen werden.preserves ice-flow indicators which have subsequently been wind polished and cross-cut by shallower grooves, and earlier studies suggested that the areas were ice-free during the last glaciation.BIERMAN et al. (1999), however, determined 26 Al/ 10 Be ratios that are less than the production ratio of ~6.1 within the bedrock that do not support a single exposure history, and suggest the areas must have been shielded from cosmic radiation at some point during and/or after initial exposure.On the one hand, young exposure ages (see IVY-OCHS & KOBER 2008) of the sampled erratics, lying on the modern surface, suggest deposition by ice during the Last Glacial Maximum (LGM, 14 -30 ka).On the other hand, high nuclide concentrations within the bedrock do not support glacial cover during the LGM, due to the lack of glacial erosion and therefore the lowering of nuclide concentration.To explain these opposing observations BIERMAN et al. (1999) suggest cover by coldbased glaciers that would have not eroded the underlying bedrock and therefore would have not reset the "cosmogenic exposure clock".By modelling different exposure-burial scenarios they identified that the Sioux Quartzite samples have minimum total burial times that are more than twice as long as minimum exposure times, with an average minimum burial duration of 414 ± 29 ka.For the northern edge of the former Laurentide Ice Sheet on Baffin Island they conclude from 10 Be and 26 Al data, a non-erosive cold-based ice or deep snowfield cover for at least 400 ka.A similar approach was used by STROEVEN et al. (2002) andFABEL et al. (2002) who investigated ancient landscapes in Northern Sweden as relicts of the Fennoscandian Ice Sheet, to test the hypothesis of landscape preservation through multiple glacial cycles.STROEVEN et al. (2002) investigated three tor samples and a bedrock outcrop in a meltwater channel in the Parkajoki area.The meltwater channel data are an exposure age of 11 ± 3 ka for both 10 Be and 26 Al, as a reliable deglaciation age.The lowered 26 Al/ 10 Be ratios of the tors, however, suggest that the sampled relict landscapes did not undergo a single exposure history, as re-ported in the regional literature.FABEL et al. (2002) looked at bedrock outcrops and erratics on relict surfaces in the northern Swedish mountains.These yield deglaciation ages of ~8 -12 ka with 10 Be surface exposure dating of the erratics but much older surface ages of ~33 -60 ka for the bedrock, similar to STROE-VEN et al. (2002).These researchers refined the approach described by BIERMAN et al. (1999) by using the marine benthic foraminifer oxygen isotope record of global ice volume from DSDP 607, as a proxy for the duration of periods of ice sheet cover vs. ice free conditions.They postulated 11 exposure and 10 burial events with a combined duration of 128 and 477 ka respectively for the Parkajoki area and a mean exposure-burial duration of 45 and 800 ka for the northern Swedish mountains.

Summary and outlook
The recent advances in terrestrial cosmogenic nuclides in geology and geomorphology have only been possible following the development of accelerator mass spectrometry in the late 1970's and early 1980's (see FINKEL & SUTER 1993), although the first burial and exposure histories were too complex and so difficult to interpret and provide reliable ages.A 26 Al/ 10 Be ratio significant lower than the production ratio was believed to indicate that the samples had been buried before, and only minimum burial ages could be constrained.After identifying that cave environments provide the ideal shielding scenario (instantaneous and sufficient), which prevents post burial nuclide production, it was possible to produce absolute burial ages from 10 Be and 26 Al determinations (GRANGER et  al. 1997).This step, around 10 years ago, marked the beginning of "real" cosmogenic burial dating.In the last few years, burial dating has been applied using the nuclide profiling approach, to more complex environments such as alluvial fans (MATMON et al. 2005), fluvial terraces (WOLKOWINSKY & GRANGER 2004) or to relict surfaces beneath glacial ice that undergo partial shielding (STAIGER et al. 2005;BRINER et al. 2006;DAVIS et al. 2006).Nevertheless, this Sediment burial dating using terrestrial cosmogenic nuclides often produces relatively large uncertainties that result in scepticism outside the cosmogenic community.As already shown, however, dating with cosmogenic nuclides offers unique advantages: application to quartz, the most common mineral on Earth's surface, and a time range of 100 ka to 5 Ma, making it very attractive to geomorphologists, (Quaternary) geologists and palaeoanthropologists.Although fundamentals underlying its application have been recognised for almost 70 years, burial dating using cosmogenic nuclides is a relatively new technique and it is likely that in future nuclide pairs other than the 10 Be/ 26 Al one will be used.The shorter half-life of radiocarbon, 5730 ± 40 years (GODWIN 1962), would enable burial dating in the order of thousands to tens of thousands of years and allow the dating of more recent processes using the 10 Be/ 14 C pair in quartz.Additionally, slowly eroding rocks should have inherited 14 C concentrations close to saturation and would allow burial age determination without any other nuclides as initial nuclide concentration is reasonably well known.The 10 Be/ 36 Cl pair is also of interest, and would allow burial age determinations within carbonate rich and mafic environments.The first step towards the use of this nuclide pair has been made recently by BRAUCHER et al. (2006) who investigated the in situ 10 Be production rate and the chemical behaviour of 10 Be in carbonates as well as clinopyroxene samples.They identified a normalised 10 Be production rate of 37.9 ± 6.0 atoms g -1 a -1 in calcite and 3.1 ± 0.8 atoms g -1 a -1 in clinopyroxenes using the longer half life of ~1.5 Ma, and also presented laboratory protocols for the in situ 10 Be extraction from calcite and pyroxene samples.New cosmogenic nuclide systems or combinations will be of interest in the coming years and the ongoing evaluation of physical constraints such as nuclide half lives (see chapter 1), nuclide production rates, muon production effects in depth and general long term production effects will also improve the precision of the general cosmogenic nuclide approach.The burial dating method stands today at a reasonably wellunderstood experimental level, although model evaluations and sensitivity tests are often not considered, or only partially understood, as a result of lack available data, compared to other well established dating techniques, such as radiocarbon, uranium/thorium or luminescence dating.This situation will change in the future through further research and the discovery of new applications.
radiation, as they are a byproduct of decaying secondary cosmic radiation (GOSSE & PHILLIPS 2001).Cosmogenic 10 Be and 26 Al are generated in e main target for 10 Be and 28 Si transforms to 26 Al) by spallation reactions ive muon capture, and a cascade of reactions called Coulomb-reactions (fast rocesses are explained in detail in LAL & PETERS (1967) and GOSSE & among others.Nuclide production in near surface material is dominated by %)(HEISINGER 1998) and changes with increasing depth, as muons display n depth as a result of their lower reactivity.oduction as a function of depth can be shown as a simple exponential law N et al. 1992; GOSSE & PHILLIPS 2001): The index � represents muonic production, + and show fast and n respectively.The scaled surface production rate, sometimes called local producti altitude above sea level and latitude, due to the effects of the Earth secondary cosmic rays and the attenuation effect of the atmosphere for the sampling position is also under debate at the moment, and is inLAL (1991),DUNAI (2000), STONE (2000), WAGNER et al. (2000) (2001), and references therein.The nuclides 10 Be and 26 Al are also produced in situ by non-cosmog reactions.SHARMA & MIDDLETON (1989) stated that only �-inducedprovide a significant portion of 10 Be and26 Al.Lithium as target nuc ( 7 Li(�,p) 10 Be) is normally present at trace levels (�g g -1 ) but sodium ( 23 Na(�,p) 26 Al), is nearly always present at concentrations of one or 10 Be production is quite small compared to the cosmogenic compou negligible, while that of26 Al may be significant.BROWN et al. (199    the steady-state concentrations of radiogenically produced26 Al in a ~6.0 x10 4 at g -1 .This low concentration, with respect to steady-state cosmogenically produced 26 Al of ~3.2 x10 7 at g -1 (Fig.1& 2), will samples with a short exposure history.Beside this, no relevant ra contamination in quartz has ever been observed and reported so far.
. (1997)  were the first to use the26 Al-10 Be pair to date burial events of sediments.They dated quartz pebbles found in various caves within dolomites along the New River, between Eggleston and Pearisburg, Virginia, USA.The pebbles, quartz vein remnants of metamorphic host rocks in New Rivers headwaters, were sampled in numerous cave levels up to 35 m above the modern river.By dating the inwash events of these pebbles from 0.29 ± 0.18 Ma to 1.47 ± 0.22Ma, GRANGER et al. (1997)  were able to calculate river incisions rates for the New River of 30.2 ± 5.5 m Ma -1 at Pearisburg and 19.7 ± 3.2 m Ma -1 at Eggleston, with a mean rate of 27.3 ± 4.5 m Ma -1 .
followed by selected examples.For more detailed (2000) and GRANGER & MUZIKAR (2001).orycave sediments ded by cave sediments (Fig. 4), in that shielding from ectively to inwashed sediments.The great advantage of lmost unaffected by cosmic radiation and thus only sults.Assuming steady-state erosion conditions for the dependent on nuclide half-lives.Thus todays 26 Al/ 10 Be ars) solving equation 5 and 6 for 10 Be and 26 Al.

Table 1 :
NISHIIZUMI et al. (2007))^HOFMANN et al. (1987)  # NORRIS et al. (1983)Constants for production rate and burial age calculations with in situ produced 10 Be and26Al extracted from quartz minerals.SLHL = sea level, high latitude (> 60°)Tab.1:Zusammenstellungder Konstanten für die Produktionsraten-und Altersbestimmung mit in situ produziertem 10 Be und26Al aus Quarz.SLHL (sea level, high latitude) = Meeresspiegel, hohe Breiten (> 60°) According to strict nomenclature rules, muons belong to tertiary cosmic radiation, as they are a byproduct of decaying secondary cosmic radiation pions (π mesons) (GOSSE & PHILLIPS 2001).Cosmogenic 10 Be and 26 Al are generated in quartz ( 16 O as the main target for 10 Be and 28 Si transforms to 26 Al) by spallation reactions (neutrons), negative muon capture, and a cascade of reactions called Coulomb-reactions (fast muons).These processes are explained in detail in LAL & PETERS (1967) and GOSSE & PHIL-LIPS ( The great advantage o these deposits is that they remain almost unaffected by cosmic radiation and thus only negligible post burial production results.Assuming steady-state erosion conditions for th sediments, the burial age is directly dependent on nuclide half-lives.Thus todays26Al/ 10 ratio gives us the burial age (t in years) solving equation 5 and 6 for 10 Be and26Al.
The great advantage o these deposits is that they remain almost unaffected by cosmic radiation and thus only negligible post burial production results.Assuming steady-state erosion conditions for th sediments, the burial age is directly dependent on nuclide half-lives.Thus todays26Al/ 10 ratio gives us the burial age (t in years) solving equation 5 and 6 for 10 Be and26Al.
ANDREAS DEHNERT & CHRISTIAN SCHLÜCHTERacross the profile or a known inherited26Al/ 10 Be ratio.The model solutions from[8] and  [9]need to be fitted to the field observations (measured nuclide concentrations).This can be done by least squares regression or other suitable optimisation techniques.Collecting more than two samples overconstrains the model solution, resulting in reduced random errors and yields a more robust detection of systematic deviations.The first use of nuclide profiling to obtain sediment burial ages was done by GRANGER GRANGER & SMITH (2000)a known inherited26Al/ 10Be ratio.The model solutions from[8]and [ need to be fitted to the field observations (measured nuclide concentrations).This can be d by least squares regression or other suitable optimisation techniques.Collecting more than two samples overconstrains the model solution, resulting in reduced random errors and yie a more robust detection of systematic deviations.The first use of nuclide profiling to obtain sediment burial ages was done byGRANGER & SMITH (2000).They measured the 10 Be and26Al concentrations of nine samples in a ~10 m