We synthesized published data on the erosion of the Alpine foreland basin and apatite fission-track ages from the Alps to infer the erosional sediment budget history for the past 5 m.y. The data reveal that erosion of the Alpine foreland basin is highest in front of the western Alps (between 2 and 0.6 km) and decreases eastward over a distance of 700 km to the Austrian foreland basin (∼200 m). For the western Alps, erosion rates are >0.6 km/m.y., while erosion rates for the eastern foreland basin and the adjacent eastern Alps are <0.1 km/m.y., except for a small-scale signal in the Tauern Window. The results yield a large ellipsoidal, orogen-crossing pattern of erosion, centered along the western Alps. We suggest that accelerated erosion of the western Alps and their foreland basin occurred in response to regional-scale surface uplift, related to lithospheric unloading of the Eurasian slab along the Eurasian-Adriatic plate boundary. While we cannot rule out recent views that global climate change led to substantial erosion of the European Alps since 5 Ma, we postulate that regional-scale tectonic processes have driven erosion during this time, modulated by an increased erosional flux in response to Quaternary glaciations.

It has been proposed that mountainous erosion increased globally around 5 Ma in response to global climate change (e.g., Hay et al., 1988; Zhang et al., 2001; Herman et al., 2013), mainly because this increase coincides with a cooling trend indicated by global isotopic data (e.g., Zachos et al., 2001). Records of erosion rates are typically provided by low-temperature thermochronological analyses of crystalline rocks exposed in mountainous regions or by sedimentary budget studies of the sedimentary rocks of the adjacent sedimentary basin. However, Willenbring and von Blanckenburg (2010) have challenged the validity of such records, based on aliasing effects.

The Alps have played a prominent role in this debate. Kuhlemann (2000) and Kuhlemann et al. (2002) constructed sediment budgets for the western and eastern Alps for the past 35 m.y. Herman et al. (2013) inverted apatite fission-track (AFT) ages across the Alps to derive an erosional history of the orogen. Both data sets disclosed a substantial increase in the erosion of the Alps at 5 Ma (dashed line in Figs. 1B and 1C). This temporal coincidence was used to call for a climate driver (Cederbom et al., 2004; Vernon et al., 2008; Herman et al., 2013), mainly because this increase was not accompanied by tectonic convergence across the Alps during this time period (e.g., Schmid et al., 1996). However, several authors have emphasized the importance of lithospheric-scale processes beneath the Alps, (e.g., Lyon-Caen and Molnar, 1989; Andeweg and Cloetingh, 1998; Kissling, 1993; Lippitsch et al., 2003; Genser et al., 2007; Sue et al., 2007; Kuhlemann, 2007; Wagner et al., 2010), which could also explain the increase in erosion rates through surface uplift. However, these studies are based on data from the Alps or the foreland basin without considering either the contribution of erosional recycling of foreland basin material to the erosional budget or the shape and size of the region undergoing erosion and surface uplift.

To provide an additional perspective on this debate, we combined sediment budgets from the foreland plus AFT ages from the orogen to infer a spatial gradient map of erosion rates for the Alps and the Alpine foreland basin. Our database1 consists of published (1) AFT cooling ages for the Alps (Fig. 2; e.g., Vernon et al., 2008; Luth and Willingshofer, 2008; Wölfler et al., 2012), (2) AFT ages from wells from the Swiss foreland basin (e.g., Cederbom et al., 2011), and (3) stratigraphic data from industry wells in the German and Austrian foreland basin (e.g., Lemcke, 1974; Genser et al., 2007). We focus our analysis on the shape and scale of the area undergoing erosion since 5 Ma within the foreland and the orogen.

The western and eastern Alps formed as a result of convergence between the Eurasian and Adriatic plates. The most recent collisional episode occurred before 35 Ma, when the Adriatic plate collided with the already deformed southern margin of Eurasia, resulting in frontal collision in the eastern Alps but oblique collision in the western Alps (Schmid and Kissling, 2000). The postcollisional deformation history also differs across the orogen, reflecting counterclockwise rotation of the Adriatic plate (Fig. 3). Shortening continued to ca. 5 Ma in the western Alps, but at slow rates (e.g., Schmid et al., 1996), while at present, orogen-parallel transtension and surface uplift dominate the deformational style (e.g., Sue et al., 2007). Convergence, thrust-faulting, and lateral escape continue to present day in the east (e.g., Robl and Stüwe, 2005; Rosenberg and Berger, 2009).

The Alpine foreland basin formed on the Eurasian lithosphere in response to loading and flexural bending starting in late Eocene time (e.g., Lyon-Caen and Molnar, 1989; Schmid et al., 1996; Andeweg and Cloetingh, 1998; Genser et al., 2007). The basin subsided rapidly during the Oligocene, but slower during the Miocene, reaching maximum depths of >4 km (e.g., Lemcke, 1974). The end of deposition of sediments in the basin is difficult to reconstruct, because regional-scale erosion has removed the youngest sediments from the basin. The age of the youngest preserved sediments ranges from 6 Ma (east) to 16 Ma (west) (Pfiffner et al., 2002; Genser et al., 2007). Regional-scale erosion appears to have been active by 5 Ma at the latest, based on high-resolution AFT analyses from boreholes (Cederbom et al., 2004).

Erosion of the Po Basin on the south side of the Alps was limited to a brief episode of deep incision between 5.6 and 5.5 Ma, related to the Messinian salinity crisis. This was followed by renewed sedimentation that continues to present day (Bertotti et al., 2001). On the eastern margin of the Alps, the Styrian Basin and part of the Pannonian Basin experienced uplift at ca. 5–6 Ma due to lithospheric-scale processes, perhaps due to counterclockwise rotation of the Adriatic plate (e.g., Wagner et al., 2010).

Between 15 and 5 Ma, Kuhlemann et al. (2002) inferred a constant material flux of ∼20,000 km3/m.y. for the western Alps, and ∼10,000 km3/m.y. for the eastern Alps, respectively (Fig. 1), based on sediment budgets. At ca. 5 Ma, the sediment discharge gently increased in the eastern Alps by ∼5000 km3/m.y., while the sediment discharge in the western Alps increased significantly to ∼30,000 km3/m.y. These values represent the bulk material flux with sources in the Alps and the adjacent foreland basin. As outlined herein, we estimate the contribution of foreland basin recycling and glacial erosion (yellow and blue bars in Fig. 1, respectively), which we use to show that the size of the eroding system (i.e., the Alps and the foreland) has increased. We use this observation to invoke a driver situated in the Eurasian lithosphere underlying the western Alps.

Sediment Budgets, Recycling, and Glacial Carving

To determine the robustness of the sediment budget curve of Kuhlemann (2000), we first ruled out aliasing effects by recalculating the sediment budget in equally spaced 5 m.y. steps (Figs. 1B and 1C). We then considered that the younger Alpine-derived foreland basin sediments were eroded from the foreland basin within a few million years after their deposition, and we subtracted their contribution (calculated from borehole information; Appendix 1 [see footnote 1]) from the sediment flux curves in Figures 1B and 1C accordingly (see yellow boxes). The difference yields the flux from the orogen. We finally estimated the contribution of glacial erosion using a mean reduction in the Alpine elevation of ∼65 m due to Pleistocene glacial carving, as reported by Sternai et al. (2012), and extrapolated this value across the Alps (87,000 km2; blue boxes in Figs. 1B and 1C).

Synthesis of Erosion Rates for the Alps and the Foreland Basin

We compiled published data of erosion for the foreland basin and exhumation for the Alps, as summarized in Appendix 1 (see footnote 1). We plotted published estimates of erosion and erosion rates, where available, on Figure 2A. In cases where no erosion rates were provided, we estimated erosion or exhumation rates using basic assumptions about (1) the geothermal gradient (between 25 and 30 °C/km), (2) partial annealing temperatures for AFT (between 120 °C and 90 °C) studies, (3) the significance of the AFT cooling ages (exhumation = erosion), and (4) the timing of the onset of erosion in the foreland basin (5 Ma; Appendix 1 [see footnote 1]). We then drew contour lines of equal erosion rate by hand on Figure 2B, using geological, geomorphological, and geochronological constraints on the position of the contour lines. We note that erosion rate estimates from the basin often lack precise age control for the onset of erosion, and that exhumation data from mountain belts may be older than the time period under consideration and hence provide a longer-term average with no resolving power over the period considered here (ca. 5 Ma).

Sediment Budgets, Recycling, and Glacial Carving

Thicknesses of eroded sections (between 0.6 and 2 km; Fig. 2A) of the Swiss foreland (12,000 km2) yield ∼7000–24,000 km3 of eroded material. The relatively large errors in the erosion rate estimates are mainly due to uncertainties regarding the amount of basin inversion (Data Repository [see footnote 1]). The recycling of the Swiss foreland yields between ∼1400 and <5000 km3/m.y. of the sediment flux for the western sector during the past 5 m.y. This is one third of the increase in Kuhlemann’s original curve (Fig. 1B). The remaining two thirds of the remaining material were derived from the western Alps during the past 5 m.y. This implies that erosion rates in the western Alps must have increased at ca. 5 Ma, consistent with Kuhlemann (2000).

For the German and Austrian foreland basin (∼35,000 km2), thicknesses of eroded sections (∼600–700 m; Lemcke, 1974) amount to 21,000–24,500 km3 of recycled material (Figs. 1 and 2A). This is <5000 km3/m.y. of the sediment flux with sources in the eastern foreland during the past 5 m.y., implying a relatively constant erosional flux in the eastern Alps since the Oligocene. A potential increase appears nonmeasurable with the sediment budget approach (e.g., Wagner et al., 2011).

The contribution of <6000 km3 of material removed by glacial erosion in the Alps is significant, but its contribution to the late Miocene–Pliocene sediment budget is relatively small. Accordingly, while a cooling climate does contribute to a higher erosional flux (Herman et al., 2013), glacial erosion alone appears to be insufficient to explain the increase in erosion in the western Alps. Likewise, as already pointed out by Champagnac et al. (2007), erosional rebound induced by glacial erosion alone cannot explain the modern pattern of rock uplift in this part of the Alps. As outlined next, we invoke here an additional driving force rooted in deeper lithospheric levels beneath the western Alps.

Pattern of Late Miocene–Pliocene Erosion Rates

The synthesis shows a coherent, regional-scale pattern of erosion for the western Alps and the foreland basin, and a smaller-scale, fault-bounded pattern centered on the Tauern Window of the eastern Alps. The highest exhumation rates occur in the western Alps (>0.6 km/m.y. over 5 m.y.; cf. area of deeply exhumed crystalline massifs in Fig. 2B), while the lowest detectable erosion rates occur in the easternmost foreland basin, and in the eastern Alps (<0.06 km/m.y.) around the Tauern window. Erosion and erosion rates of the Alpine foreland basin are highest in front of the western Alps (>1.5 km, >0.3 km/m.y.; Figs. 2A and 2B) and decrease eastward over a distance of 700 km to the eastern portion of the Alpine foreland basin (∼200 m; <0.06 km/m.y.).

Our synthesis also reveals very low average erosion rates for the eastern Alps, the Northern Calcareous Alps, and the Alps east of the Tauern Window. This is consistent with the pre-Miocene paleosurfaces described by Hejl (1997) and Frisch et al. (2001) (Fig. 2). The regional-scale pattern of erosion rates is perturbed by an anomaly centered on the Tauern Window, which is characterized by the highest local spatial gradients in erosion rates (>0.3 km/m.y. over <10 km), reflecting young tectonic activity (e.g., Scharf et al., 2013; Frisch et al., 2000).

Erosion Rate Pattern and Tectonic Features

The center affected by high erosion rates is situated in the western Alps and the adjacent portion of the Alpine foreland basin, all of which overlie the Eurasian lithosphere (Figs. 3 and 4). While the southwestern margin of this pattern is poorly known (e.g., Sissingh, 1997), the southeastern extent has a well-defined spatial gradient and coincides largely, from southeast to the northeast, with (1) the Periadriatic Line, (2) the western limit of the Austroalpine nappes, and (3) the northern Alpine thrust front (Fig. 2B). High erosion does not affect any other circum-Alpine basin (e.g., the Po Basin), nor any portion of the southern Alps, which sit above Adriatic lithosphere.

The acceleration of erosion rate, by itself, cannot be used to rule out any tectonic or climatic processes, because both drivers may have occurred on a short time scale and at high rate, not detectable over a 5 m.y. period. However, our new synthesis provides a high spatial resolution of exhumation far outside the core of the Alps to the foreland, which allows us to eliminate models that do not result in the shape of the erosion rate pattern described herein. A viable climate model should result in erosion rates that are higher in the mountains than in neighboring low-relief basins. However, the erosion rate pattern shows the opposite behavior (Fig. 2B): The erosion contour lines cross the mountain-foreland border at an oblique angle. On this basis, we rule out global climate change as the main driving force.

The onset of accelerated erosion (Fig. 1) also coincided with the end of shortening in the Jura (Schmid et al., 1996; Sue et al., 2007), implying that horizontal convergence rates across this portion of the orogen were negligent by ca. 5 Ma. Instead, the erosion pattern reflects vertical, buoyancy-driven lithospheric processes through unbending and unloading of the downward-bent Eurasian slab (e.g., Genser et al., 2007), as already proposed by Lyon-Caen and Molnar (1989) and Sue et al. (2007). The eastward-decreasing gradient in foreland basin erosion may then reflect the increasing flexural plate strength (e.g., Stewart and Watts, 1997). Slab unloading is also consistent with high uplift rates beneath the area of Mont Blanc, where current erosion rates have been highest during the past few millions of years (Figs. 2B and 3). Lippitsch et al. (2003) explained this rebound through positive buoyancy resulting from possible tearing, breakage, and separation of the Eurasian slab (Figs. 3 and 4).

Effects from the counterclockwise rotation of Adria were invoked by Wagner et al. (2011) to explain the young uplift of the Styrian Basin in the eastern Alps, which has resulted in crustal shortening and underthrusting of the Pannonian fragment. The same mechanism can be used to explain active uplift and erosion in the vicinity of the Montello thrust (Fig. 2B) on the southern side of the Alps (Rosenberg and Berger, 2009). For the western Alps, Sue et al. (2007) attributed extensional deformation and dextral escape as consequences of a rotating Adriatic plate. In this context, it seems plausible that other tectonic events, such as reactivation of the Tauern Window (e.g., Frisch et al., 2001), or active buckling of the Aar massif, particularly at its eastern tip (labeled with Ch on Fig. 3), may be consequences of deformation along an unevenly shaped plate boundary geometry of the rotating Adriatic indenter. In particular, simultaneous extension and slab unloading in the western Alps would result if the Euler pole of rotation between the Eurasian and the Adriatic plate is located along the plate boundary, between the two regions (Fig. 3); changes in the motion of Adria then result in changes in the kinematics of the Alps (Figs. 3 and 4B).

It is possible that reactivation along the Periadriatic Line (accelerated exhumation of the Bergell area since 4 Ma; Fig. 2) decoupled most of this rock uplift from the southern Alps. The same mechanism can be invoked for the Longitudinal-Houiller-Penninic frontal fault (LF-HF-PF; Fig. 3) in the western Alps, which delineates a western block with late Miocene to Pliocene fission-track ages from an eastern block with Oligocene to Miocene fission-track ages (Malusà et al., 2005). In particular, the Periadriatic Line has juxtaposed the subsiding Po Basin next to the Alpine blocks where exhumation has persisted to the present (Carrapa and Garcia-Castellanos, 2005). Counterclockwise rotation of the Adriatic plate about an Euler pole located south of the central Alps (Fig. 3) could contribute to unloading of the Eurasian slab below the western Alps (and thus to surface uplift through positive buoyancy), and likewise to uplift and compressional escape in the eastern Alps due to indentation of the Adriatic plate (Fig. 4B).

The notion of lithospheric slab unloading is also consistent with the recent reorganization of the entire drainage pattern of the major rivers. By ca. 5 Ma or even later, cannibalization between the Rhine and Danube streams (Ziegler and Fraefel, 2009) could have occurred in response to the inferred slab tear in the west-central Alps (Mont Blanc area).

We suggest a model that is capable of explaining the following observations: (1) a coherent “banana-shaped” pattern of fast erosion during the past ∼5 m.y. over an area of ∼700 × 150 km from the western Alps to the eastern portion of the Alpine foreland basin, (2) an increase in erosion rates of >0.6 km/m.y. toward the center of the erosional region, (3) a switch from depositional to erosional processes over most of the Alpine foreland basin contemporaneous with acceleration of erosion in the Alps, (4) an erosional signal that is strongest in the western Alps, (5) and concurrent sedimentation in the Po Basin south of the Periadriatic Line. We propose that the erosion rates depicted here largely accommodate surface uplift due to unloading of the Eurasian lithospheric slab beneath the western Alps, from where it affected the adjacent mountains and the Alpine foreland basin to a distance of over 700 km along strike. The erosional signal is strongest above the Eurasian lithosphere and weakest above the Adriatic plate. Local tectonic exhumation of the Tauern Window and active shortening in the southeastern Alps interfere with this pattern. In contrast, subsidence continues in the Po Basin adjacent to the Periadriatic Line, implying decoupling between the Eurasian and the Adriatic plates. Counterclockwise rotation of Adria contributes to east-west gradients in shortening and erosion via lithospheric loading in the east and unloading in the west. Spatial patterns of erosion rate on a regional scale may be helpful in identifying lithospheric-scale tectonic processes.

R. Baran acknowledges a graduate student fellowship from the Bavarian Elite Network. Partial support was provided by the Swiss National Science Foundation (SNSF) (project No. 20T021–120525) and the TopoEurope initiative of the European Science Foundation (ESF) awarded to F. Schlunegger. We thank Science Editor Eric Kirby, Kurt Stüwe, and an anonymous reviewer for constructive comments.

1GSA Data Repository Item 2014127, Appendix 1, is available at www.geosociety.org/pubs/ft2014.htm, or on request from editing@geosociety.org, Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301-9140, USA.