Diagenesis of rockslide deposits
Diagenesis of rockslide deposits:
A discovery and its relevance for proxy-dating
sturzstrom events and their subsequent changes

During our investigation of the Fern Pass sturzstrom (Northern Calcareous Alps, Austria), we discovered that the deposit is locally lithified by carbonate cement that precipitated after the sturzstrom event (Fig. 2), and that this cement can be used to proxy-date the sturzstrom event by the 234U/230Th disequilibrium method (Ostermann et al., 2007). The 234U/230Th age obtained for the cement is in perfect agreement with results from 36Cl exposure age-dating and with 14C ages, but is considerably more precise than either of the methods mentioned (Ostermann et al. 2007; Prager et al., 2006). In combination with the 36Cl exposition age, which by its approach yields the ‘correct’ age of the rockslide event (see above) but has a large error range, the 234U/230Th age served to substantially improve the precision of age determination of the sturzstrom event (Ostermann et al., 2007; Prager et al., in press). To the best of our knowledge, this represented the first proxy-dating ever of a sturzstrom by U/Th dating of cements. Because the 234U/230Th method is applied to carbonate cement that formed after emplacement of the rockslide deposit, the age of the cements represents a ‘minimum age’ of the sturzstrom event. The 234U/230Th method is applicable to carbonate cements back to an age of 500.000 years (Geyh, 2005), and has recently been successfully applied to age-dating of cements of Alpine Quaternary ‘non-rockslide’ deposits such as talus breccias, fluviatile conglomerates, and spring tufas (Ostermann, 2006; Ostermann et al., 2006a-d; Sanders et al., submitted). For the Fern Pass event, the comparison of exposition ages of the rockslide detachment scar with the U/Th age of the cement revealed that cement precipitation started soon or immediately after the event, such that the numerical age of the cement represents a good proxy of event age (Ostermann et al., 2007). To explain this, we speculated that immediately after descent, rock slides are rich in chemically reactive carbonate-rock flour produced by dynamic disintegration. In contact with meteoric waters, part of the rock flour dissolves and re-precipitates as a cement (Ostermann et al., 2007).

The prime question emerging in view of the Fern Pass case study was whether other lithified rockslide deposits can be found or whether Fern Pass is an exception. A preliminary field survey during spring and summer 2007, however, shows that localized cementation of carbonate-lithic rockslide deposits is indeed widespread. By September 2007, we had identified lithification in eight of eight carbonate-lithic sturzstroms visited, including Flims, the largest rockslide of the Alps. One of the rockslides in which cementation was observed, the Pfitschertal rockslide (Southern Tyrol, Italy), is situated in a metamorphic terrain of the Central Alps, and is composed of clasts of calcschists and calcareous phyllites. This shows that, given a carbonate-rich lithological composition, rockslides are accessible to U/Th dating in metamorphic terrains, as well. Localized lithification of rockslide masses by carbonate cements was rarely mentioned - if considered worthy of mention at all - by a very few previous authors (Abele, 1974; Poschinger, 1995; Weidinger, 1998), but always in passing only and always as a kind of ‘strange’ feature. Our data, although still preliminary, clearly show that cementation of carbonate-lithic rockslides is by no means rare or the result of exceptional conditions, and that the cements bear high potential as a new unexploited source of age information on the rockslide event itself. U/Th dating of cements can be conducted at uranium concentrations down to an order of magnitude less than in those found on Fern Pass (Kramers et al., 2006). With respect to lithification of rockslides, we identified a whole spectrum of diagenetic products. Lithification of rockslides and formation of diagenetic products are present in different locations on different rockslides. In case of the rockslides of Flims, Tschirgant and Pletzachkopf, existing radiometric age data (mainly 14C ages and 36Cl exposition ages) would make it possible to check the deduced U/Th for their correctness as a proxy age of the event, or whether the U/Th ages record diagenetic changes well after the event. To this end, careful petrographic analysis of samples is necessary to distinguish different generations and types of cement. Preliminary U/Th-dating results from the Tschirgant rockslide resulted in an age of about 2900 years BP, which is in fine agreement with the 3000 years BP 14C-data (Erismann & Abele, 2001).

In the Fern Pass case study, it was only the petrographically oldest generation of cement that was used for U/Th dating (Ostermann et al., 2007). In principle, however, the petrographically subsequent generations of cement can also be age-dated. The results of such a ‘multi-generation dating’ of cements constrain the total age range during which diagenetic changes took place within the rockslide mass. Such changes may proceed over a considerable interval of time and, in some rockslides at least, may still be going on at present. For instance, along the walls of the deep gorge of Rabiusa river incised into the Flims rockslide mass, cemented breccias were found that contain limpid cements suitable for U/Th dating. Along the near-vertical bank of Rein Posteriur incised into the Tamins rockslide, underboulder breccias are present closely above the stream level; these breccias may not have formed closely after the rockslide event, but only subsequent to stream incision. In addition, in the case of the Tamins rockslide, cementation of conglomerates that overlie the rockslide deposit, and cements within talus slope breccias that formed after erosional incision of the rockslide mass by the Rhine river, may allow the establishment of a chronological sequence of geomorphic changes subsequent to the rockslide event. Given careful sampling under due consideration of sampling context, U/Th dating of cements in rockslide deposits thus additionally has the potential, if successful, to provide information on later changes that affected the rockslide mass. To date this type of information could not be provided, not even in principle by any other method of radiometric dating.

In summary, proxy-dating of rock slide events and potential dating of later changes affecting the rockslide mass by U/Th dating of cements most probably face a wide field of application. Relative to exposition dating, U/Th dating has the advantages that field sampling is quicker and easier, that ages can be produced more readily, that the precision of deduced U/Th ages (including 2 σ standard errors) tends to be higher than that of most exposition ages, and that sample preparation and measurement are significantly cheaper. Given the criteria where to search for lithification in rockslides, if cementation took place, it is commonly recognized within a short time. U/Th dating of cements in rockslides can be done with a few breccia samples each only about 10-20 cm in size, thus this seems an ideal method to proxy-date rockslides in remote, hard-to-reach locations (e. g. the Himalayas) and with limited time for field work. Yet exposition dating has the undisputed advantage that it is the only method that veers for the ‘real’ age of a sturzstrom event. Combining the precision of U/Th ages with the correctness of (often more blurred) exposition ages was, therefore, the ideal approach to determine most precise ages for selected Alpine landslides. Uranium-thorium dating of cements (and, where possible, of diagenetic successions) of Alpine sturzstroms combined with exposition dating would represent a worldwide pioneer approach to precisely date sturzstrom events and their subsequent diagenetic changes. This approach may represent a major step towards age determination of a large number of rockslides that hitherto could not be (precisely) dated by the 14C method and for improving, by cross-checking, both U/Th disequilibrium dating and exposition dating.



Phenomena of lithification

Examples of lithified rockslides

some observations