RESEARCH CONTRIBUTIONS - GEODYNAMICS GROUP

With the formation of the Centre for Advanced Data Inference (CADI) in Earth Physics, the Geodynamics Group’s research now focusses on two principal areas. (i) the geodetic monitoring of crustal deformation and the study of the underlying tectonic processes, and (ii) the interactions between ice sheets and sea level during recent glacial cycles.

The crustal deformation studies involve two long term observation projects, both partly funded by ARC Discovery Grants. One is in Papua New Guinea where the initial goal is to develop a comprehensive model of the kinematics of the crustal motion using GPS. 2003 fieldwork included the reobservation of a transect along the western border and results include estimates of shortening across the highlands. A substantial basic network has now been established and repeat observations at two or more epochs are now available from many sites that give a quite detailed picture of the deformation of PNG in response to the convergence of the Australia-Pacific plates (Figure 1). From the equator to the pole, the second observation program is the monitoring of the time dependence of geodetic rock-sites in Antarctica, in and around the Lambert Glacier, with the objective of determining the glacial rebound signal. This project is also partly funded by the Australian Antarctic Division. The fourth site, at Mt Komsomolsky, some 800 km inland, was installed in the 2002–2003 season and preliminary site-velocity estimates were obtained for 2002–2003.

Figure 1: Plot of the velocity field and tectonic setting of PNG. The velocity fields in the Gazelle Block and the Ramu-Markham Fault Zone have been omitted for the sake of clarity.

Another long-term observing program is the measurement of the time-dependence of gravity in a joint experiment with the National Astronomical Observatory of Japan at Mizusawa using a superconducting gravimeter. The instrument located at the ANU’s Mount Stromlo Observatory survived the destruction of the January 2003 fires but, despite the destruction of the building in whose basement it was housed and the loss of power, the instrument sensor continued to function throughout and loss of recorded data was minimal.

The sea level and isostatic rebound research focussed on three areas in 2003: the development of new ice models for the last European and North American glaciations, the establishment of the change in ocean volumes from Late Holocene to the present, and constraining sea levels and ice sheets during the period between the last two major glaciations, including recent time. This research is also partly funded by ARC Discovery Grants and by the Antarctic Climate and Ecosystems Cooperative Research Centre (ACE-CRC). This latter work includes the analysis of speleothems from now-submerged caves to establish the timing of marine and terrestrial growth. This work, using uranium series dating and trace element analyses, is being carried out by Dr Andrea Dutton and Mr Giovanni Scicchitano, in cooperation with the environmental Geochemistry Group in RSES. The samples worked on are from two caves in Italy and the project is a cooperative one with ENEA and the University of Catania.

Work on recent sea level change includes collaboration with scientists from the ACE-CRC in analysing and interpreting instrumental measurements (tide gauge and satellite altimetry). The RSES contribution includes evaluation of geological, glaciological and hydrological contributions to present ocean volume and is being carried out by Ms Gisela Estermann.

In 2003, Dr Douglas Christie returned to RSES from the Comprehensive Nuclear Test Ban Treaty Organisation, in Vienna, to continue his infrasonics research.

The following short reports represent a cross-section of the group’s activities for 2003.

Tectonic Studies in Papua New Guinea

P. Tregoning, R. Stanaway, H. McQueen and K. Lambeck

In 2003, a major GPS field campaign was undertaken in Papua New Guinea spanning most of the mainland. There were several objectives of the work:

1. To reobserve a transect of sites running along the western border of PNG from the coast, across the highlands to the southern flood plains. The objective is to measure the distribution of shortening through the highlands caused by the collision of the Australian Plate with the New Guinea land mass.
2. To reobserve a network of sites in the Schouten Islands and in the region of Wewak. Some of these sites were affected by co-seismic displacement after the September 2002 Wewak earthquake, while other sites had been observed only once before.
3. To continue the monitoring of the post-seismic motion in the New Ireland/New Britain regions
4. To reobserve a network in the Lae regions as part of monitoring deformation across the Ramu-Markham Fault.

There is ongoing relaxation occurring after the November 2000 New Ireland earthquakes. The co-seismic displacements estimated from GPS analyses have been inverted to model the slip distribution on the Weitin Fault strike-slip event and the two subsequent thrust events that occurred on the New Guinea Trench. The inversion for fault location and slip distribution is non-unique for the two thrust events because the GPS co-seismic offset data are not able to separate the contributions of each thrust event (both occurring south of the GPS network). Aftershocks were relocated using the Arrival Pattern method developed at RSES (Nicholson et al., 2002) in an attempt to better constrain the locations of the thrust earthquakes. The research is ongoing.

Antarctic Glacial Isostatic Rebound

P. Tregoning, K. Lambeck, H. McQueen and R. Decrevel

In the 2002/03 Antarctic summer season one new remote GPS site was installed at Komsomolsky Peak (75.2S 63.6E) and the equipment at Dalton Corner was upgraded. Both these sites were equipped with Iridium satellite modems provided under the US Iridium Project to permit data transmission back to RSES while the sites were unattended. This has worked successfully. The sites at Beaver Lake and Landing Bluff were serviced but remain essentially the same, with satellite phone communications continuing via the Inmarsat Satcom-B system.

All four sites recorded GPS data until early May 2003 and all four sites successfully awoke from hibernation in early spring. The two sites equipped with Ashtech microZ receivers recommenced recording GPS data in late September and were operating when revisited in the 2003/04 summer season. We have now demonstrated that it is possible to record GPS observations over 200 days per year with our solar-powered systems.

The analysis of the recorded data shows temporal height variations of up to 20 mm, with the signals at all sites showing significant correlation. Furthermore, it was found that receiver firmware changes at the GPS receivers at Mawson and Davis in 2000 (operated by Geoscience Australia) have increased the number of low elevation observations, which has subsequently introduced a bias in the height estimates (Tregoning et al., 2003). Relative heights calculated between sites have a much smaller scatter and are more likely to provide useful estimates for geophysical interpretation. At this stage it appears that there is less than 1 mm/yr relative vertical movement between Davis, Beaver Lake, Landing Bluff and Dalton Corner, but Mawson appears to be uplifting at around 2 mm/yr relative to all of these sites.

A Dynamic Geodetic Datum for Papua New Guinea

R. Stanaway and P. Tregoning

Significant advances have been made in the accessibility and accuracy of positioning technology in recent years, especially with GPS. Tectonic deformation impacts on the use of geodetic datums in tectonically active countries, such as PNG. We have developed a strategy whereby national geodetic datums and survey networks in tectonically active regions can include a geodetic velocity field, strain models and other non-secular offset data in order to maintain the integrity of the datum. A least-squares datum adjustment program has been developed which includes these dynamic elements, to enable geodetic surveyors to reduce geodetic measurements made in dynamic local networks to a reference epoch. The program has applications for the monitoring of geophysical hazards and localised crustal deformation.

RSES’s ongoing geodetic monitoring of crustal deformation in PNG has been closely allied with geodetic activities conducted by other PNG institutions, including the National Mapping Bureau, the PNG University of Technology, and the Rabaul Volcanological Observatory. GPS observations from a network of over 200 geodetic stations in PNG gathered by RSES and these institutions, have been collated and analysed in order to examine the effects of tectonic deformation on the PNG Geodetic Datum. The availability of geodetic observations from these institutions has proved to be highly beneficial in the understanding of the complex tectonic setting in PNG, particularly across the Highlands Fold and Thrust Belt, the Lae urban area, and the region encompassed by the Gazelle Peninsula and Southern New Ireland. The rugged and inaccessible nature of much of PNG supports a collaborative approach for the gathering of geodetic data, as the data are useful for both geophysical and practical geodetic applications, including land surveying, mapping, and navigation. RSES has funded several large GPS campaigns in PNG, and the physical infrastructure, data archive and analysis resulting from these campaigns forms an ideal basis for a Geodetic datum and geodynamic monitoring network for PNG (See Figure 1).

Improved ice sheet model for Scandinavia for the last glaciation

K. Lambeck and A. Purcell

The last ice age of Fennoscandinavia continues to have geological and geodetic repercussions across the region despite the ice having retreated almost 10,000 years ago. The most obvious consequences are the land uplift along both sides of the Gulf of Bothnia and the concomitant retreat of the sea. More subtle changes include the present day tilting of large inland lakes and sea shorelines and the vertical and horizontal displacements of the earth’s crust, measured with geodetic techniques, and the associated changes in crustal stress. The geological and geodetic signals provide constraints on two important classes of parameters that define (i) the Earth’s rheology, and (ii) the repeat history of the Scandinavian ice. Through a careful exploration of the combined parameter space, exploiting the spatial and temporal variability of the response signals (sea-level change, tilting, strain) it is possible to establish constraints on both groups of parameters and to develop physically plausible models that are not just descriptive but that also have a predictive capability.

Mantle rheology. From inversions of an improved geological observational data set (including the Baltic Ice Lake shorelines), the mantle rheology can be adequately represented by a three-layer model of lithospheric thickness Hl, upper mantle viscosity hum, lower mantle viscosity hlm (with realistic radial dependence of density and elastic moduli) with values of

Hl ¼80–90 km
hum (2.5–3)1020 Pa s
hlm ¼(7–30)1021 Pa s.

Curiously, the inversion of geodetic data (including tide gauges, lake tilting and GPS) yields somewhat higher values for hum (4.25 ± 0.25) 1020 Pa s. These values represent "effective" values and if substantiated by the next-iteration solution, points to possible non-linear behaviour of the mantle response.

Crustal Stress. The state of stress of the planet’s surface is the sum total of many processes but the one that, in areas subject to glaciation, changes on relatively short time scales is that induced by the changes in the surface loads of ice and water. The rebound model-development has reached a stage where it is possible to predict realistic regional stress patterns. We use as a measure of the incremental stress the change in the Fault Stability Margin (FSM), ÐFSM, resulting from the change in surface load. In the absence of other force fields, the crust beneath the ice is stabilized but it becomes less stable in regions outside of the loaded area. When the ice retreats the crust beneath the former ice becomes less stable and existing faults may be reactivated. Detailed modelling has been carried out for Finland and for much of the central area the deviatoric FSM reaches its maximum values at about 10,000 years ago but the magnitudes relax with time and today are only about 20% of their peak values (See Figure 2). Thus any failure within the crust triggered by glacial loading and unloading will have occurred preferentially when the region first became ice free and the potential for reactivating faults today must be considered as negligibly small.

Figure 2: The DFSM at the surface for four epochs (kaBP = 1000 years Before Present). The magnitudes are given by the contours and the style of faulting is given by the colour coding (orange-brown thrust; blue normal). Only for T = 12 ka BP are there regions where the style of faulting is different from thrusting.

 

The Revised Ice Sheet. Figure 3 illustrates the ice thickness at three epochs as inferred from the inversion of the geological data with the ice margin locations constrained by field data. The maximum ice, at the time of the Last Glacial Maximum (LGM) occurred over the northern part of the Gulf of Bothnia with a value of ~2700 m. A secondary maximum of ~2400 occurred over southern-central Sweden. By 18.8 ka BP considerable thinning of ice occurred over the Gulf of Bothnia and we speculate that this caused the late ice advances over the northern European Plain and Denmark. This rapid decrease in ice volume at ~19 ka BP appears to be a robust feature all the inversions.

Figure 3. Ice thickness maps for three epochs: (a) the time of the Last Glacial Maximum at 20 ka BP. (b) at ~19 ka after the occurrence of a substantial thinning of the ice, (c) at the time of the Younger Dryas between about 12.8 and 12ka BP.

Ancient Roman fish tanks and present-day sea level change

K. Lambeck, A. Purcell, M. Anzidei and F. Antonioli

An issue of some importance is whether the present-day observed sea level change is representative of the past century, the typical duration of the instrumental record, or whether it is representative of a longer interval. If the latter, then the causes could be human-induced. A recent re-evaluation of tide gauge records by Church et al. (J. Climate, in press) indicates that the sea-level rise globally has been ~1.8 ± 0.2 mm/year for the past 50 years. A few isolated tide gauge records suggest that this rise may have started towards the end of the 19th century but the data is limited. As part of our global sea level studies we have searched for sea level indicators for the first few millennia that may be used to constrain ocean volumes. Archaeological remains of fish tanks from the Roman epoch (100 BC to 100 AD) have provided particularly useful indicators because water exchange between the tanks and the sea is tidally controlled and because the tidal range in the Mediterranean is small (see Figure 4). A substantial number of fish tanks have been examined south and north of Rome and they provide a consistent estimate of local sea level for 2000 years ago. Control on any tectonic component is provided by the observed height of the Last interglacial Shoreline and the glacio-hydro isostatic contribution (the remnant signal from mainly the Northern European deglaciation) is constrained by geological data extending back to ~14,000 years ago in Italy. The result provides an estimate of eustatic sea level at 2000 BP.

Figure 4 : (A) underwater photo of the in-situ sluice gate at La Banca. The complete gate consist of: (i) a horizontal stone surface that defines the threshold with a groove to receive the gate; (ii) two vertical posts with grooves to guide the movement of the gate; (iii) an upper stone slab with horizontal slot to extract the gate; and (iv) the gate itself, ~65 cm high, with small holes for water. In this illustration the gate is partially covered by sand and the bright zone is that part of the gate that has been cleaned of sand and biological growth. The threshold, covered by sand, lies ~10 cm below the measuring rod that is calibrated at 10 cm intervals. (B) Sketch of the channel sluice gate with sliding posts, threshold and lowest level crepidinae as viewed from within the fish tank. The top of the sluice gates coincides with the elevation of the lowest level foot-walks and corresponds to a position above the highest tide.11 In this example, the lowest foot-walk is now 0.72 m below the sea surface at the time of measurement and the threshold is 1.32 m below present sea level. (C) Plan and cross-section of the Ventotene Roman harbour indicating structures that identify limiting values to sea level (s.l.), including the lowest level foot-walks and bollards carved into the rock. Roman time s.l. is ~1.40 m below present mean sea level (m.s.l.), the now-submerged bollard would have been at ~0.85 m above m.s.l. and the lowest level foot-walks are estimated to have been ~0.30 m above m.s.l. The channel through the breakwater at B for flushing the harbour remained operational at all tidal levels.

Four tide gauge records exist for the same section of coast from which a composite record of modern local sea level change has been constructed. Corrections for the isostatic and tectonic signals then lead to an estimate of the present-day eustatic change (the change in sea level due to changes in ocean volume, caused either by addition of water into the oceans or by changes in water temperature). Extrapolation of this rate back in time results in the Roman sea levels being reached only after about 100 years and the conclusion is that the present-day eustatic sea level change is a recent phenomenon and is not part of a longer-term "natural" background signal.

Infrasound and the Comprehensive Nuclear-Test-Ban Treaty

D.R. Christie

The infrasound component of the International Monitoring System (IMS) is used to detect and locate atmospheric nuclear explosions as part of the verification regime for the Comprehensive Nuclear-Test-Ban Treaty. In contrast with seismology and hydroacoustics, research in the field of atmospheric infrasound has largely been neglected during the last 30 years. This situation has changed dramatically with the on-going establishment of the 60-station IMS global infrasound-monitoring network. The prospect of data of unprecedented quality from a global network, coupled with the recognition that the field of infrasound can profit greatly from recent advances in electronics and computing, has led to a renaissance in the field of infrasound. New infrasound research programs have been established at a large number of universities and other institutions around the world. Fundamental problems in the field of infrasound that originally appeared to be intractable can now be tackled with confidence and the results can be applied to a wide variety of geophysical studies in the atmosphere.

Work on the 60-station IMS infrasound-monitoring network is proceeding rapidly. Approximately 40% of the stations in the network have been completed and are transmitting continuous data via satellite to the International Data Centre in Vienna Austria. Work has started at nearly all of the other stations in the global network and it can be anticipated that 60% of the stations will be in operation by the end of 2004.

The IMS infrasound-monitoring network is far larger and much more sensitive than any previous infrasound monitoring network. The global coverage of the IMS infrasound network provides a unique opportunity for high-resolution time-dependent tomographic studies of atmospheric dynamics and structure, studies of acoustic wave propagation on a global scale and detailed studies of atmospheric transport processes.

While there has been considerable progress during the last few years in the technology of infrasound monitoring, a number of important problems remain unresolved. All of these problems have a potentially serious detrimental influence on the performance of the IMS monitoring system. The detection and location capability of the infrasound monitoring system needs to be improved significantly. Work needs to be undertaken to develop optimal array designs and reliable detection algorithms. Many of the stations that will be installed in the next few years are located in high-wind environments. These stations will be subject to unacceptably high background noise levels. In addition, high levels of background noise during the daytime seriously limit the performance of a substantial number of existing stations in the IMS infrasound network. Further fundamental research needs to be carried out to identify all components in the infrasound noise spectrum and techniques need to be developed to reduce background noise levels to acceptable levels. The source of most infrasound signals observed at IMS stations is largely unknown. Research needs to undertaken to develop procedures that can be used for the reliable identification and discrimination of infrasonic signals. This research should include the development of propagation models that can accurately describe the complex evolution pattern of infrasonic waves at regional distances. The research on infrasound in Earth Sciences will be largely focussed on an attempt to resolve these problems in support of the CTBT monitoring program.

Data from the global infrasound monitoring system can also be used in a number of practical applications. Perhaps the most important application at the present time is the use of infrasonic data to provide a timely warning to the aviation community of the presence of volcanic ash clouds in the lower stratosphere. Ash deposited in the lower stratosphere by a volcanic explosion represents a very serious hazard to aviation. The early identification of areas with hazardous volcanic ash clouds remains an unresolved problem. Even after nearly a decade of work on this subject, it is recognized that complete avoidance of areas with potential ash clouds is the only guaranteed way of ensuring aircraft safety. Earlier work carried out in Earth Sciences on this subject indicated that data from the IMS infrasound network could be used to provide a timely advisory warning of all significant volcanic explosions that may inject volcanic ash into the lower stratosphere. Negotiations between the Comprehensive Nuclear Test-Ban-Treaty Organization and the International Civil Aviation Organization have been underway for some time to provide IMS infrasound data for a timely volcanic ash advisory warning system for aircraft. It is planned to extend the earlier work on the volcanic ash problem to establish a realistic measure of the effectiveness of the use of infrasound as a means for reliably monitoring global volcanic activity and also to provide a preliminary model that will allow a direct assessment of volcanic ash hazard as determined from the morphology of observed infrasound signals from volcanic explosions.

Estimating neotectonic movement in southern Victoria using cosmogenic burial dating

Fabel, D., Gardner, T., Webb, J., and Fink, D.

The aim of this project is to determine the age of several tectonically displaced sedimentary deposits in the Cape Liptrap area of southern Victoria using cosmogenic burial dating.

The geomorphology of this region testifies to the profound influence of faulting in shaping the landscape. While quantitative constraints on the displacement history of these faults are not yet available, dramatic evidence for recent movement is indicated by kink-bands in ~125,000 year old cemented dune limestone near Cape Liptrap. There are no numerical ages for tectonically offset older sedimentary deposits that could provide further constraints on the age and rate of neotectonic movement. We are using a relatively new chronological technique, cosmogenic burial dating, utilising the radioactive decay of cosmic-ray produced Be-10 and Al-26 in buried quartz, to estimate these ages. Based on the differential decay of the in-situ produced cosmogenic nuclides Be-10 (radioactive half-life = 1.51 ± 0.03 m.y.) and Al-26 (0.705 ± 0.02 m.y.) in quartz, the technique is widely applicable to dating up to approximately 5 million year old quartz-rich sediments. The ratio Al-26/Be-10 in quartz is dominated by the nuclide production rate ratio for all but the very slowest erosion rates where radioactive decay starts to become significant. Because Al-26 and Be-10 are produced at a constant ratio, independent of absolute production rates, the Al-26/Be-10 ratio in quartz is robust against temporal production rate variations. In a steadily eroding landscape, quartz grains within the soil and sediment contain Al-26 and Be-10 concentrations in this predictable ratio. If these quartz grains are subsequently buried, for example deep within a sedimentary deposit, then cosmogenic nuclide production within those grains is attenuated by the overburden and inherited Al-26 and Be-10 concentrations diminish by radioactive decay. Because Al-26 decays more rapidly than Be-10, the Al-26/Be-10 ratio decreases exponentially with time. By measuring the Be-10 and Al-26 concentrations using accelerator mass spectrometry, the current Al-26/Be-10 ratio in the sample can be determined, and the burial time calculated.

Preliminary results indicate that coarse sand and gravel overlying horizontally cut marine platforms at elevations ranging from 4 – 18 m above present mean sea level were deposited between 630 ka and 1.15 Ma. Although the results are intriguing, they are complicated by post burial cosmogenic nuclide production. We are currently evaluating methods of obtaining reliable ages for samples with complex burial histories.

Tracing the post-Younger Dryas retreat of the northern Fennoscandian Ice Sheet using cosmogenic radionuclide exposure ages

Fabel, D., Stroeven, A., Dahlgren, T., Harbor, J., Hättestrand, C., and Kleman, J.

The aim of this project is to determine the rate of retreat of the Fennoscandian ice sheet from the Younger Dryas limits in northern Norway to the terminal limits in the northern Swedish mountains (Figure 5). The north to south retreat history is poorly constrained due to a lack of datable material. We are working to provide new constraints on the timing and pattern of deglaciation using cosmogenic nuclide apparent exposure ages.

The work involves mapping and dating depositional and erosional geomorphological features related to the former ice sheet margin. Because the ice sheet initially had warm-based conditions close to its margin, the dominant morphology is one of eskers and aligned lineation systems, such as crag-and-tails. Abundant meltwater eroded bedrock locally to considerable depth and deposited fans or deltas perched above current local base levels (Figure 5). However, subglacial conditions during final deglaciation were generally cold-based, inhibiting the formation of eskers and lineation systems, although there are widespread (lateral) meltwater channel erosional imprints and occasional plucking scars.

Each geomorphological setting was examined for its value in providing deglaciation ages, testing the initial assumption that, (i) abundant erosion on crags of crag-and-tails, across transverse erosional scarps, and in meltwater channels has exposed bedrock surfaces without a prior exposure history and (ii) depositional features contain embedded boulders without a prior exposure history (on the surfaces of eskers and deltas, and erratics). Preliminary results indicate that meltwater channels, transverse erosional scarps, and erratics yield deglaciation ages that are consistent with the limited ages provided by other methods, but that crag-and-tails yield apparent exposure ages that are too old, presumably because of a prior exposure history that was not fully removed by glacial erosion.

Figure 5. (A) Crag and tail. Ice flow was from right to left. (B) Perched glaciofluvial delta dissected by meltwater channels. (C) Esker

Basal thermal regimes of former ice sheets constrained using cosmogenic nuclides

Fabel, D., Stroeven, A.P., Harbor, J., Kleman, J., Clarhäll, A., Elmore, D., Fink, D.

The motion of ice sheets occurs as a result of internal deformation and basal sliding, which includes direct sliding of ice over its substrate, internal deformation in the substrate, and enhanced deformation in basal ice (Paterson, 1994; Benn and Evans, 1998). Basal sliding is ineffective where basal ice is below the pressure melting point (cold based, or frozen bed conditions) and most effective where basal ice is at the pressure melting point (warm based, or thawed bed conditions), so that water is present (Cuffey et al., 2000). Reconstructions of the last glacial maximum Fennoscandian and Laurentide ice sheets are extremely sensitive to assumed basal thermal patterns (Kleman and Hättestrand, 1999; Marshall et al., 2000), resulting in estimates of ice thickness that vary on the order of 1 km (Clark et al., 1999).

Recognition of subglacial boundaries between sliding and frozen-bed areas for former ice sheets is typically based on distinct morphological contrasts between areas with glacial landform assemblages and relict areas showing little alteration of pre-existing features. However some of these boundaries, especially on continental shield areas, are clearly visible from air photos but have minimal topographic expression (Figuure 6). Understanding the chronology and erosional development of such boundaries is important to provide insight into the pattern and persistence of basal conditions under ice sheets.

Geomorphic evidence and cosmogenic radionuclide concentrations of bedrock outcrops on either side of sliding boundaries on low-relief upland plateaus in northern Sweden (Figure 6) are consistent with negligible erosion in relict landscape (frozen bed) areas due to the last glaciation, but also indicate very limited erosion in the sliding areas. This pattern and magnitude of landscape modification indicates that sliding was short lived in these areas, likely as a transient phase during deglaciation. These sites demonstrate that short periods of sliding are in some cases sufficient to produce landscapes that are recognized as ‘glacial’ from air photos. Thus regions of sliding identified on shield areas must be viewed as the cumulative total area that has experienced sliding at any time during a glaciation. The actual extent of sliding areas during most ice sheet phases is presumably considerably less than this cumulative total, which has important implications for establishing appropriate basal boundary conditions for ice sheet reconstructions.

Figure 5. Infra-red air photo (a), geomorphological map (b) and oblique airphoto (c) of Mt. Tjuolmma and Juovvakielas on Ultevis. The oblique air photo (c) is taken from the point shown by the eye symbol in (b). Pre-Late Weichselian lateral meltwater channels are crosscut by till lineations and the lateral sliding boundary. The channels can be traced on the glacially eroded side of the boundary as water-filled depressions aligned with the down-stream continuation of the lateral channels. The lateral sliding boundary itself appears as a sharp feature on air photos and terminates in transverse lee-side scarps. Sample locations are marked by open squares (erratic samples) and closed squares (bedrock samples). 10Be and 26Al (italics) apparent exposure ages with uncertainties are given in thousands of years. An asterisk indicates a weighted mean age.