Search

The extent and distribution of glaciers on the Swiss Plateau during the Last Glacial Maximum (LGM) can be determined from the geological record. However, similar reconstructions for the glaciations that preceded the LGM are far more difficult to be made due to the inaccessibility of suitable sedimentary records. Here, we explored Quaternary sediments which were deposited during the MIS 8 glaciation at least 250 ka ago, and which were recovered in a drilling that was sunk into an overdeepening W of Bern (Switzerland). We analyzed the sediment-bulk chemical composition of the deposits to investigate the supply of the material to the area by either the Aare Glacier or the Valais Glacier. The potential confluence of these two glaciers in the Bern area makes this location ideal for such an analysis. We determined the sediment-bulk chemical signal of the various lithological units in the central Swiss Alps where the glaciers originated, which we used as endmembers for our provenance analysis. We then combined the results of this fingerprinting study with the existing information on the sedimentary succession and its deposition history. This sedimentary suite is composed of two sequences A (lower) and B (upper), both of which comprise a basal till that is overlain by lacustrine sediments. The till at the base of Sequence A was formed by the Aare Glacier. The overlying lacustrine deposits of an ice-contact lake were mainly supplied by the Aare Glacier. The basal till in Sequence B was also formed by the Aare Glacier. The provenance signal points towards a simultaneous material supply by both the Aare and the Valais Glaciers during the formation of the lacustrine sediments in Sequence B. We use these findings for a paleogeographic reconstruction. During the time when Sequence A and the basal till in Sequence B were deposited, the Aare Glacier dominated the area. This strongly contrasts with the situation during the LGM, when the Aare Glacier was deflected by the Valais Glacier towards the NE. Probably, the Valais Glacier was less extensive during MIS 8. However, part of the lacustrine sediments deposited within Sequence B could only have been supplied by the Valais Glacier, indicating that the glacier did not cover the study area, yet had been in close proximity to the study area. We thus postulate that during the deposition of Sequence B both the Aare Glacier and the Valais Glacier were connected to this lake that had formed at the foot of these glaciers. These glaciers potentially also dammed this lake. In conclusion, we could outline a detailed scenario of sediment supply to the investigated overdeepening during the MIS 8 glacial period based on the provenance and sedimentological data, and that glaciers were arranged in a different way than during the LGM.

We investigated the formation mechanism of tunnel valleys, by producing 3D models of bedrock topography using gravimetry. We obtained the cross-sectional geometry of tunnel valleys in the Swiss foreland, near Bern. The combination of information about the densities of the sedimentary fill and of the bedrock together with borehole data and gravity surveys along profiles across the valleys served as input for our 3D gravity modelling software, Prisma. This finally allowed us to model the gravity effect of the Quaternary fill of the overdeepenings and to produce cross-sectional geometries of the overdeepenings. We focused on two sections situated in the Gürbe valley and in the Aare valley. We determined a density of 2’500 kg/m3 for the Upper Marine Molasse bedrock, and with Prisma we obtained a bulk density of kg/m3 for the Quaternary infill. Our gravity surveys across the valleys yielded a maximum residual anomaly of -2.9 mGal for the Gürbe valley and -4.1 mGal for the Aare valley. The application of our Prisma model showed that these anomalies can be explained by Quaternary suites with a thickness of 160 m and 235 m for the infill of the Gürbe and Aare valleys, respectively. The high-resolution information about the cross-sectional geometry of the tunnel valley flanks, from the application of Prisma, allowed us to infer a two-step formation process of the overdeepened trough. A first glaciation, during MIS 6 or before, would have deepened the trough. And a second glaciation, during the Last Glacial Maximum (MIS 2), would have widened the valleys. We explain this pattern by the differences between the ice thicknesses of the LGM and MIS 6 glaciers and by the relatively low erodibility of the Upper Marine Molasse bedrock. The Molasse units indeed comprise tender and porous sandstones and offer a lower erosional resistance than the Quaternary infill, which consists of cohesive and thus competent glacio-lacustrine marls. This probably offered ideal conditions for the thick and thus erosive MIS 6 glaciers to erode deeply into the Molasse bedrock. In contrast, the lacustrine fill of this trough possibly prevented the thinner and thus less erosive LGM or MIS 2 glaciers to further incise the bedrock. The consequence was that erosion of the LGM glaciers mainly occurred on the lateral sides, thereby resulting in a widening of the tunnel valleys. Finally, we apply this approach to the remaining gravity profiles, to create a 3D model of the geometry of the overdeepening network near Bern.

A scientific drilling was conducted into a bedrock trough in Bern-Bümpliz, a branch of the Aare Valley overdeepening (Switzerland). It is the first scientific drilling in the Bern area that reached the bedrock. We analyzed the 208.5 m-thick succession of Quaternary sediments recovered in this scientific drilling and present the sedimentological results of the campaign. In the retrieved sediments 12 different lithofacies were identified, which were grouped into 5 facies assemblages, and 2 major sedimentary sequences (A = lower, B = upper), which transition into a minor sequence C. Generally, the sedimentary successions of sequences A and B are similar. The lowermost facies assemblage of each sequence consists of a till that was deposited during a period of ice cover. However, the two tills differ from each other. In particular, while the till at the base of sequence A is dominated by large clasts derived from the underlying Molasse bedrock, the till at the base of sequence B has no such Molasse components. Furthermore, the till in sequence A bears evidence for glaciotectonic deformations. Both tills are overlain by thick facies assemblages of subaqueous, most likely glaciolacustrine and lacustrine sediments. Sequence A is characterized by cross-bedded and steeply inclined sand, gravel and diamictic beds which we interpret as deposits of density currents in a subaqueous ice-contact fan system in a proglacial lake. In contrast, the lacustrine sediments in sequence B are considered to record a less energetic environment where the material was most likely deposited in a prodelta setting that gradually developed into a delta plain. Towards the top, sequence B evolves into the fluvial system of sequence C, where large sediment fluxes of a possibly advancing glacier resulted in a widespread cover of the region by a thick gravel unit. Additionally, feldspar luminescence dating was performed on two samples from a sand layer at the top of sequence B. The dating in combination with lithostratigraphic correlations with the sequences encountered in the neighboring scientific drillings to the north (Meikirch) and south of Bern (Thalgut) suggests that sequence B was deposited during Marine Isotope Stage 8 (MIS 8; 300–243 ka).

Interpretations of the processes leading to the formation of overdeepened valleys, where the bedrock lies well below sea level today, are contested as the overdeepenings have been filled by sediments or host lakes making observations difficult. Here, we combine gravimetric, GNSS (Global Navigation Satellite System) and borehole data within a 3D forward modelling framework (Gravi3D) to assess the 3-D subsurface geometry of such overdeepenings in the Swiss plateau, to the North of the Alps. Gravi3D has two components (PRISMA and BGPoly), which allow to obtain analytically the gravity effect of prisms and polygons (Nagy (1966) and Talwani & Ewing (1960)). PRISMA allows first to estimate the spatial extent of an overdeepening and the density contrast between the overdeepening fill and the bedrock. In contrast, BGPoly is designed to disclose the details of a complex 3-D geometry of an overdeepening fill through an approximation of its shape with polygons. Gravi3D will be open access and is designed for a larger scientific community. Here, we focus on overdeepenings beneath two valleys, the Aare valley and the Gürbe valley to the South of Bern. In this region, the occurrence of overdeepenings has already been disclosed through drilling, but the details about the geometry have not been elaborated yet. The study region is characterized by three mountain ranges oriented North-South and comprises Burdigalian Upper Marine Molasse bedrock. The Gürbe and Aare valleys in-between are c. 300 m-deep and c. 1 km-wide, where overdeepenings with a >100 m-thick Quaternary fill have already been identified by drilling. The gravity data collected along an 8 km-long profile with stations spaced between 100 and 300 m yield a Bouguer anomaly that ranges from c. -99 to -106 mGal. We relate this anomaly to the regional trend (c. 2 mGal over 8 km) and to the effect of the overdeepenings’ sedimentary fillings (2 – 4 mGal/km), disclosing a sharp anomaly pattern over the inferred tunnel valleys. The removal of the signal related to the regional trend results in a residual anomaly of c. 1 mGal for the bedrock ridge in-between the valleys (Belpberg mountain), and of -2.65 and -3.56 mGal for the Gürbe and Aare valley overdeepenings, respectively. We observe a steeper gravity gradient for the Eastern flank of both overdeepenings. The use of Nettleton method to model the residual gravity anomaly across Belpberg yields a density of 2.5 g/cm3 for the Molasse bedrock. In addition, the estimation of the largest gravity response through the overdeepening fill, calculated with Prisma yields a density value of c. 2.0 – 2.2 g/cm3 for the Quaternary sediments. As a further information, Prisma predicts a maximum thickness of 140 m for the Quaternary suite beneath the Gürbe valley and at least 200 m beneath the Aare valley. This yields a minimum slope of approximately 18° for the Gürbe overdeepening. REFERENCESNagy, D.: The gravitational attraction of a right rectangular prism. Geophysics 31, 362–371, 1966.Talwani, M., Ewing, M.: Rapid computation of gravitational attraction of three‐dimensional bodies of arbitrary shape. Geophysics 25, 203–225, 1960.

Pollen analysis has shown that at regional to subcontinental scales interglacial vegetation successions usually follow similar patterns throughout the Middle and Late Quaternary. Yet, relevant compositional and species-specific differences among interglacials exist and contribute to their characteristic biostratigraphic fingerprint. We use these divergences to indirectly date palaeolake deposits from Spiezberg (SW margin of overdeepened lake Thun, Switzerland), which have been assigned to either MIS 5e (last interglacial) or MIS 7a (penultimate interglacial) according to the average IRSL age of 164 ± 16 ka. For this purpose, we analyzed the fossil pollen record of the Spiezberg palaeolake deposits and applied optimal partitioning and the broken stick model for zonation. Furthermore, we re-assessed local pollen assemblage zones (LPAZ) of two physically dated reference records (Beerenmösli: MIS 5e; Meikirch: MIS 7a) from the study area by using the same zonation approaches. LPAZ of all three records were then compared optically and numerically (PCA and significance tests) to test which of the two reference records is more similar to Spiezberg. Pollen data show that the major part of the Spiezberg record was sedimented during full interglacial conditions. The irrelevance of Fagus points towards an MIS 5e age for the sediments. This is supported by PCA axis 2, which shows a statistically significant similarity of the Spiezberg record to the MIS 5e reference. This outcome is explained by Fagus playing an important role in the PCA axis 2 gradient, which is probably determined by competition for light. PCA axis 1 is not taken into consideration for the correlation since it represents a climate gradient (e.g. from boreal to temperate forests) and explains most of the intra-locality but less of the inter-locality variance. We thus assign the Spiezberg record to the last interglacial and show that climate-driven compositional differences between MIS 5e and MIS 7a are not diagnostic. In contrast, distinct differences in single-species abundances (e.g. Fagus) are strong enough to significantly distinguish between MIS 5e and MIS 7a records. We conclude that variability across various interglacial vegetation successions (e.g. MIS 5e vs. 7a, PCA axis 2) is smaller than local vegetation variability within a full interglacial succession at the same site (PCA axis 1). This implies a very high vegetation resilience to glacial-interglacial climatic variability, the biotic properties involved might be migration capacity, survival in refugia and adjustments to rapidly changing Quaternary environments.

Overdeepenings are common bedrock features that were formed within the Alps and their forelands during repeated glaciations. Here, we present a total of 208.5 m of sediment core recovered during a scientific drilling campaign in spring 2019. The drilling is located on the Swiss Plateau (CH) close to the terminus of the overdeepening that was formed by the Alpine Aare glacier. The drilling reached the Molasse bedrock at an elevation of 362.5 m a.s.l. This is the first time in the middle Aare Valley that a scientific drilling fully recovered the sedimentary fill of the overdeepening.The retrieved material comprises an alternation of silt- and sandbeds, and massive as well as cross-bedded gravel layers. We divide the suite into two packages reflecting sedimentation within a fluvio-glacial environment. The first package starts with a glacial till, c. 15 m thick, which is overlain by an 89 m-thick succession of cross-bedded sandy and gravelly layers and mud interbeds reflecting deposition at the toe of a prograding delta. At c. 104 m depth, the succession is interrupted by a till sequence, c. 20 m thick, which forms the base of the second package. This package records a fining-upward trend, starting with massive sand beds at the base and ends with massive silt layers at the top. Mapping has shown that the core section is overlain by a sand layer with soft-sediment deformation structures reflecting the occurrence of a mouth bar environment. The sequence ends with fluvial gravels. The entire sediment core thus records a shallowing-up sequence that is interrupted by a glacial till.Optical stimulated luminescence dating (IRSL; Buylaert et al., 2009) yielded a minimum age of c. 200 ka for the mouth bar deposits, which is in accordance with reported Late-Middle Pleistocene ages of corresponding sequences in the region (Preusser et al., 2005). We therefore assign the upper package in our drilling to the MIS 7/8 glacial-interglacial cycle, and to the MIS 9/10 for the basal package. The sediments in our drilling thus record a crucial time interval when glacial erosion in the Alps (Haeuselmann et al., 2007) and in the northern foreland (Schlüchter, 2004) occurred at the highest rates.Buylaert, J.P., Murray, A.S., Thomsen, K.J., Jain, M., 2009. Testing the potential of an elevated temperature IRSL signal from K-feldspar. Radiat. Meas. 44, 560–565.Haeuselmann, P., Granger, D.E., Jeannin, P.Y., Lauritzen, S.E., 2007. Abrupt glacial valley incision at 0.8 Ma dated from cave deposits in Switzerland. Geology 35, 143–146.Preusser, F., Drescher-Schneider, R., Fiebig, M., Schlüchter, C., 2005. Re-interpretation of the Meikirch pollen record, Swiss Alpine Foreland, and implications for Middle Pleistocene chronostratigraphy. J. Quat. Sci. 20, 607–620.Schlüchter, C., 2004. The Swiss glacial record - a schematic summary. Dev. Quat. Sci. 2, 413–418.