The Long-Lasting Exhumation History of the Ötztal-Stubai Complex (Eastern European Alps): New Constraints from Zircon (U–Th)/He Age-Elevation Profiles and Thermokinematic Modeling

Abstract

The Eastern European Alps formed during two orogenic cycles, which took place in the Cretaceous and Cenozoic, respectively. In the Ötztal-Stubai Complex—a thrust sheet of Variscan basement and Permo-Mesozoic cover rocks—the record of the first (Eoalpine) orogeny is well preserved because during the second (Alpine) orogeny, the complex remained largely undeformed. Here, new zircon (U–Th)/He (ZHe) ages are presented, and thermokinematic modeling is applied to decipher the cooling and exhumation histories of the central part of the Ötztal-Stubai Complex since the Late Cretaceous. The ZHe ages from two elevation profiles increase over a vertical distance of 1500 m from 56 ± 3 to 69 ± 3 Ma (Stubaital) and from 50 ± 2 to 71 ± 4 Ma (Kaunertal), respectively. These ZHe ages and a few published zircon and apatite fission track ages were used for inverse thermokinematic modeling. The modeling results show that the age data are well reproduced with a three-phase exhumation history. The first phase with relatively fast exhumation (~250 m/Myr) during the Late Cretaceous ended at ~70 Ma and is interpreted to reflect the erosion of the Eoalpine mountain belt. As Late Cretaceous normal faults occur at the margins of the Ötztal-Stubai Complex, normal faulting may have also contributed to the exhumation of the study area. Subsequently, a long period with slow exhumation (<10 m/Myr) prevailed until ~16 Ma. This long-lasting phase of slow exhumation suggests a rather low topography with little relief in the Ötztal-Stubai Complex until the mid-Miocene, even though the Alpine orogeny had already begun in the Eocene with the subduction of the European continental margin. Accelerated exhumation since the mid-Miocene (~230 m/Myr) is interpreted to reflect the erosion of the mountain belt due to the development of high topography in front of the Adriatic indenter and repeated glaciations during the Quaternary.Mountain belts with thick continental crust, such as the European Alps, the Himalaya, or the North American Cordillera, are formed during long-lasting plate convergence with crustal shortening by nappe stacking and folding [1-3]. Due to the isostatic uplift of the thickened crust, the internal parts of such orogens become the locus of erosion, which removes material at the Earth’s surface and leads to the cooling and exhumation of metamorphic rocks [4, 5]. Apart from erosion, another important mechanism that may cause rock exhumation and cooling is normal faulting because tectonic slip along normal faults transports rocks in their footwalls toward the Earth’s surface [6-9].To quantify the cooling history of metamorphic rocks, it is necessary to determine the temperature conditions in rocks through time, which is possible by applying geochronological methods such as Sm/Nd, Rb/Sr, or Ar/Ar dating to minerals with different closure temperatures [10-12]. The final cooling in the upper crust from temperatures of ~250°C to ~60°C can be constrained with low-temperature thermochronology (i.e., fission track and [U–Th]/He dating of zircon and apatite) [13-15]. Although low-temperature thermochronology is often applied on spatially distributed samples to understand regional cooling patterns [16, 17], the sampling of elevation profiles together with thermokinematic modeling of the respective cooling ages has proven to be a particularly powerful tool for reconstructing multistage cooling histories [18-21]. Such histories may be expected if exhumation rates during or after mountain building change repeatedly. Potential reasons for repeated changes in exhumation rate may include climate variations, which may promote or decelerate erosion [22], variations in the slip rate of normal faults, [23] or changes in tectonic boundary conditions. The latter may, for example, be related to the onset of continental collision after a period of subduction, the waning of plate convergence, or renewed plate convergence following a phase of tectonic quiescence [24-27].A key example of a region that is known to have experienced different phases of tectonic activity and quiescence is the Eastern European Alps (Figure 1), where the Austroalpine nappes preserve a unique geological record of Cretaceous and Cenozoic orogenies [28, 29]. The Cretaceous and Cenozoic orogenies were separated by a phase of tectonic quiescence of unknown duration [3]. Although it is likely that the different phases of tectonic activity and quiescence resulted in considerable changes in exhumation rates, the exhumation history of the Austroalpine nappes since the Late Cretaceous is still poorly resolved because most thermochronological studies focused on the Cenozoic orogeny and the evolution of the Tauern Window in the Miocene-Pliocene (Figure 1). In particular, only a limited number of apatite and zircon fission track ages are available from the Ötztal-Stubai Complex west of the Tauern Window (Figure 1), and these ages were mainly obtained from spatially distributed samples taken at low elevations [30, 31]. Moreover, quantitative constraints on cooling and exhumation based on samples from elevation profiles in combination with thermokinematic modeling are still lacking.Here, we present the results of zircon (U–Th)/He dating on samples from two elevation profiles in the Ötztal-Stubai Complex, which we complement with existing fission track data. By thermokinematic modeling of these age data, we quantify the cooling and exhumation histories of the Ötztal-Stubai Complex from the Late Cretaceous to the Quaternary. Our results document a three-stage exhumation history, which we relate to the geological evolution of the Austroalpine nappes during the two orogenic cycles. In particular, our findings suggest a significant time lag between the onset of renewed plate convergence in the Eocene, which resulted in the collision of the European and Adriatic Plates and the development of the current high topography of the European Alps.The Eastern European Alps formed during two orogenies, which took place in the Cretaceous (Eoalpine orogeny) and the Cenozoic [3, 32, 33]. The first phase of mountain building encompassed the closure of the Meliata Ocean between the Late Jurassic and Early Cretaceous and subsequent intracontinental subduction within the Apulian plate [3, 26, 33, 34]. During this southeast-directed subduction, the northern part of the Apulian plate was subducted below the southern Apulian plate and experienced eclogite facies conditions at ca. 100 Ma [35-38]. This phase led to the formation of the Austroalpine (Apulian) nappe stack by west-northwest vergent thrust faulting (Figure 1) and was followed by Late Cretaceous east-southeast-directed extension and exhumation of the nappe stack [3, 31, 39]. Most of the Gosau basins in the Northern Calcareous Alps (Figure 1), which were filled with alluvial, shallow-marine, and deep-water sediments between ~90 and ~60 Ma, appear to have formed in an extensional to transtensional setting as well [40-42], although for some of these basins an origin during crustal shortening was inferred [43]. In the Austroalpine units of Graubünden (eastern Switzerland), the phase of late-orogenic extension was reconstructed by structural analysis and is thought to have lasted from ~80 to ~67 Ma [3]. Overall, the timing of nappe stacking and crustal thinning in the Eastern Alps indicates a westward propagation of both crustal shortening and extension [3, 39].The Cenozoic Alpine orogeny occurred after the southeastward subduction and closure of the Penninic oceans (i.e., Piemont Ocean in the south and Valais Ocean in the north), which led to the subduction of the European passive margin beneath the Adriatic plate in the Eocene [32, 33, 44, 45] and subsequent nappe stacking and crustal thickening [28, 46, 47]. Indentation of the Adriatic plate toward the north led to further shortening and folding of the nappe stack [48], while the formation of N-S trending normal faults and conjugate strike-slip faults resulted in orogen-parallel extension of the Eastern Alps [31, 39]. The combination of shortening, erosion, and normal faulting caused the formation of tectonic windows (e.g., Tauern and Engadine window) that expose European and Penninic units within the Austroalpine nappe stack (Figure 1) [49-52].The Ötztal-Stubai Complex is 70 km long and 50 km wide and belongs to the Austroalpine nappe system of the Eastern Alps (Figure 1) [31, 53]. It is located between the Engadine Window in the west and the Tauern Window in the east. The Ötztal-Stubai Complex is bounded by the Inntal fault in the north, the Brenner normal fault in the east, the Schlinig fault and the Vinschgau shear zone in the south, and the Engadine line in the west (Figure 2(a)) [34, 54]. Metamorphic basement rocks of the Ötztal-Stubai Complex comprise gneisses, schists, granitoids, amphibolites, and minor eclogites [55]. Remnants of the para-autochthonous Permo-Mesozoic sedimentary cover of the Ötztal-Stubai Complex (e.g., Brenner Mesozoic) are only locally preserved (Figure 2(a)). These cover rocks were deposited as transgressive sediments on a paleoweathering horizon in the uppermost Ötztal-Stubai Complex [56]. On top of the Brenner Mesozoic occur the upper Austroalpine Steinach and Blaser nappes, which consist of Paleozoic and Mesozoic sedimentary rocks, respectively (Figure 2) [31]. During WNW-directed thrusting in the Cretaceous, the Ötztal-Stubai nappe was transported above the Campo nappe and was overthrusted by the Drauzug-Gurktal nappe system (which comprises the Tonale nappe, the Meran-Mauls basement, and the Steinach and Blaser nappes; Figure 2). The Schneeberg and Texel units represent parts of the Apulian plate that experienced high-P/low-T metamorphic conditions during the Cretaceous orogeny (Figure 2) [57, 58]. Based on a structural analysis, it was recently suggested that the Ötztal-Stubai Complex and the Texel unit are both part of one coherent nappe [54]. In the Late Cretaceous to Early Cenozoic, extension and normal faulting led to the juxtaposition of the low-grade to unmetamorphosed Steinach and Blaser nappe against the metamorphosed Ötztal-Stubai Complex and its sedimentary cover [31, 59]. The contact between these nappes is strongly deformed and shows a stretching lineation that formed during top-to-ESE-directed extensional shearing [31, 54].The Ötztal-Stubai Complex experienced a polymetamorphic evolution comprising a Caledonian, a Variscan, and a Cretaceous (also referred to as Eoalpine) metamorphic event [3, 60, 61]. The Eoalpine metamorphism resulted from W- to NW-directed thrusting and nappe stacking [3, 53, 61-63]. The Eoalpine metamorphic conditions increase from subgreenschist facies in the northwest to eclogite facies conditions in the southeastern part in the Ötztal-Stubai Complex (Figure 2) [57, 64-68]. This spatial metamorphic gradient is reflected in the distribution of Rb/Sr and K/Ar biotite ages with a trend from Variscan ages in the northwest to Eoalpine ages in the southeast [35, 63, 69, 70]. To the south of Sölden near the southeastern margin of the Ötztal-Stubai Complex, biotite K/Ar ages cluster around 90 Ma [60, 71], whereas Rb/Sr biotite ages show a greater variation and range between 83 and 105 Ma [71-74]. Farther north, in the central part of the complex, biotite K/Ar ages range from ~130 to ~170 Ma [60, 71], while biotite Rb/Sr ages are slightly older and between ~160 and ~190 Ma [71, 72, 75]. These “mixed” ages result from samples that were partly reset in the Cretaceous, suggesting temperatures slightly higher than 300°C [71, 72]. In the very north of the Ötztal-Stubai Complex, Rb/Sr and K/Ar biotite ages are significantly older, with ages of up to 319 and 294 Ma, respectively [71, 73]. The ages >290 Ma reflect the Variscan metamorphic event in the northern part of the Ötztal-Stubai Complex, where the Rb/Sr and K/Ar systems of biotite and muscovite were not reset by the Cretaceous metamorphism. Hence, the metamorphic temperatures during the Cretaceous did not exceed ~300°C. The overall age pattern indicates that the Cretaceous metamorphism mainly affected the central and southern part of the Ötztal-Stubai Complex and that the temperatures in this part of the complex were around 400°C–450°C (Figure 2).Two previous studies employed apatite and zircon fission track dating in the Ötztal-Stubai Complex [30, 31]. The samples of Elias [30] yielded AFT ages between 8.8 ± 0.9 and 21.6 ± 3.8 Ma, whereas the ZFT ages are much older and range from 77.6 ± 3.5 to 98.5 ± 6.1 Ma (Figure 3). The ZFT ages were interpreted by Elias [30] to reflect a phase of cooling after Cretaceous metamorphism during which the samples cooled below the ZFT closure temperature. Owing to the large difference between the ZFT and AFT ages, the cooling history between ~75 and ~25 Ma remains largely unconstrained.The study of Fügenschuh et al. [31] focused on the eastern part of the Ötztal-Stubai Complex, where the sedimentary cover of the Ötztal-Stubai Complex (the Brenner Mesozoic) is still preserved and structurally overlain by the Steinach and Blaser nappes (Figure 2). Here, they took nine samples from the metamorphic basement rocks and five samples from the sedimentary cover rocks and the nappes above. The basement samples yielded AFT ages between 28 ± 6 and 62 ± 10 Ma and ZFT ages ranging from 42 ± 6 to 71 ± 10 Ma (Figure 3). Based on their age data and additional structural information, Fügenschuh et al. [31] suggested a phase of rapid cooling due to normal faulting and crustal extension in the Late Cretaceous. This phase ended about 60 Ma ago, and afterward, the eastern Ötztal-Stubai Complex has cooled slowly until the present day [31].To better constrain the Cretaceous cooling history of the Ötztal-Stubai Complex, we took fourteen gneiss samples for zircon (U–Th)/He (ZHe) thermochronology (Table 1). Twelve samples were collected along two elevation profiles in the Stubaital and Kaunertal; two additional samples are from the southern Ötztal (Figure 4(a)). In the Kaunertal, six samples were taken along a 7.5 km long NNE-SSW profile. Sample elevations range between 1758 and 3132 m. In the Stubaital, the elevations of six samples range from 1736 to 3200 m along a distance of 5 km (Figure 4(b) and (c)). Both profiles are approximately parallel to the isogrades of the Eoalpine metamorphism, and the respective samples experienced temperatures of 400°C–450°C (Figures 2 and 3). The two samples from the Ötztal are from 1380 and 3363 m elevation, respectively. The higher sample is from the top of the Innere Schwarze Schneid (Figure 3). In the following, we summarize the analytical methods; a more detailed description can be found in Wolff et al. [23].For zircon (U–Th)/He dating, we selected inclusion-free and euhedral crystals. Microphotographs were taken using a Keyence digital microscope of the VHX series to determine the crystal shape parameters and to calculate correction factors for the ejection of α-particles [76] using the constants of Hourigan et al. [77]. The single crystals were degassed by heating with an infrared diode laser at the GÖochron laboratories, University of Göttingen. All crystals were checked for complete degassing of helium by sequential reheating and helium measurement. Subsequently, crystals were retrieved from the gas extraction line and were analyzed by the isotope dilution method using an ICP-MS equipped with an APEX microflow nebulizer.In this section, we describe the ZHe ages of the two elevation profiles in the Stubaital and Kaunertal and the two ages from the Ötztal. All single-crystal ZHe ages and their weighted-mean ages are reported in Table 2. In the Stubaital, the ZHe ages range from 55.9 ± 2.5 to 69.2 ± 3.2 Ma, whereas the age range in the Kaunertal is from 49.7 ± 2.4 to 71.1 ± 3.5 Ma (Figure 5). In both elevation profiles (which extend over vertical distances of ~1.4 and ~1.5 km, respectively), the ZHe ages increase with sample elevation. The only exception is sample 20A6 in the Kaunertal (Figure 5(b)) from which only one grain was dated as there were no further zircons of sufficient quality. We therefore consider the age of this sample as less reliable compared with the ZHe ages of the other samples. The two ZHe ages from the Ötztal of 42.6 ± 3.2 and 68.8 ± 2.8 Ma fit into the general pattern of increasing ZHe ages with altitude (online supplementary Figure S1). Taken together, the ages indicate a continuous cooling of the central Ötztal-Stubai Complex during the Late Cretaceous and Paleogene. The relatively large difference in the ages for the lowest and highest samples in both elevation profiles (i.e., ~15 and ~20 Ma, respectively) indicates a rather slow cooling. A least-square regression through the ZHe ages yields slopes of 0.11 ± 0.03 and 0.06 ± 0.01 km/Myr for the Stubaital and Kaunertal profiles, respectively (Figure 5(a) and (b)). Such slopes are often interpreted as exhumation rates during the period of cooling recorded by the data. However, such an interpretation may be invalid because it assumes that the temperature field in the upper crust was in a steady-state (i.e., the isotherms were spatially and temporally invariant during the period of cooling). Interpreting such slopes as exhumation rates also assumes that the rate of cooling was constant through time.We now utilize the ZFT and AFT ages of six samples from Elias [30] to gain further information on the cooling history of the rocks in our study area. The position of the six samples is shown in the map of Figure 3, and their ages are plotted in Figure 5. In the Stubaital transect, three ZFT ages from Elias [30] range from ~87 Ma to ~78 Ma (samples P5/11, P5/13, and S3), while two AFT ages are ~13 Ma and ~8 Ma (samples S1 and P5/13; Figure 5(a)). The rather small age difference between the ZFT ages and our oldest ZHe ages of ~70 Ma suggests relatively fast cooling during the Late Cretaceous. Subsequently, cooling has slowed down, as indicated by the large age difference between the ZHe ages and the AFT ages (Figure 5(a)). An acceleration in the cooling rate is required in the Miocene to account for the rather young AFT ages. For the Kaunertal, a similar cooling history can be proposed because the ZFT and AFT ages of Elias [30] and our ZHe ages are more or less the same as in the Stubaital (Figure 5(b)).In summary, the ZHe ages of the Kaunertal and Stubaital elevation profiles together with the data from Elias [30] indicate that the Ötztal-Stubai Complex did experience a three-phase cooling history. A phase of relatively rapid cooling during the Late Cretaceous was followed by slower cooling in the latest Cretaceous and Paleogene. During the Miocene, cooling must have accelerated again, although the exact timing is not known yet. In the next section, we employ thermokinematic modeling using our new ZHe ages from the Stubaital and Kaunertal elevation profiles as well as the AFT and ZFT ages of three samples from Elias [30], which lie close to our elevation profiles (i.e., samples P4/2, P5/13, and P5/11; Table 3) to constrain the cooling and exhumation histories in a more quantitative way.We applied thermokinematic modeling using the finite-element code PECUBE (version 4.2), which solves the three-dimensional heat-transport equation and predicts the thermal history of crustal blocks under prescribed kinematic and topographic boundary conditions [78, 79]. PECUBE uses the modeled time-temperature histories to calculate apparent ages for a range of thermochronometers. These ages can be compared directly with cooling ages obtained from rock samples. Inverse modeling of the compiled thermochronological data and the use of the nearest-neighborhood algorithm [80, 81] finally allow to find a best-fit model. All model calculations of this study were carried out with the high-performance cluster PALMA II of the University of Münster.For the sampling transects in the Stubaital and Kaunertal, two PECUBE models of slightly different dimensions were built (Table 4). The model constructed for the Stubaital is 16 km long and 15 km wide, whereas the Kaunertal model has a size of 19 × 16 km. These model dimensions ensure that cooling ages for all samples of each elevation profile (Figure 4) can be predicted. Both models are 55 km thick, and the present-day topography was added on top of the models. As the highest peaks in the study area have an elevation of about 3.5 km (Figure 3), the maximum thickness of both models is 58.5 km. Given that the high local relief of the European Alps is the result of considerable glacial and fluvial erosion during the Quaternary [82, 83], we reduced the relief at the beginning of all model runs by two-thirds relative to the present-day relief. In the last 2 Ma of each model run, the initial topography increases linearly to the present-day topography.In all model runs, we use a thermal diffusivity of 1 × 10−6 m2/s and a radiogenic heat production of 3 × 10−6 W/m3 [84], which decreases with depth with an e-folding length of 20 km (Table 4) [52]. The temperature-boundary condition at the base of the models is 800°C and remains constant during all model runs [52]. These thermal parameters and boundary conditions lead to an initial geothermal gradient of ~30°C/km in the upper 10 km of the model that decreases downward.For calculating cooling ages from the thermal histories modeled with PECUBE, we used the diffusion model of Galbraith and Laslett [85] for ZFT and the annealing model of Ketcham [86] for AFT. To compute ZHe ages, we adjusted the diffusion model of Reiners et al. [87] to a slightly lower closure temperature (i.e., 170°C at a cooling rate of 10°C/Myr) to account for the rather low radiation damage of our zircon samples (1.28 × 1016 to 2.26 × 1017 alpha decay events/g; Table 2) [88, 89]. The misfit between our measured ZHe ages (and AFT and ZFT ages from Elias [30]) and the ages predicted by our model is calculated as:where N is the number of data points, αi,data is the observed data, αi,model is the predicted value, and σi,data is the uncertainty of the ages [79]. To find the best-fitting model, we used a two-step neighborhood algorithm inversion and sampled the parameter space iteratively to minimize the misfit between observed and predicted ages [80]. For both the Kaunertal and Stubaital models, we performed a total of 12,880 model runs (i.e., 161 iterations with 80 models for each iteration). To ensure a sufficiently explorative sampling of the parameter space, we chose a high resampling ratio of 0.9 (cf. 79). To derive quantitative constraints for the uncertainties of the five free model parameters, we used marginal probability density functions [81-90].Our new ZHe ages and the AFT and ZFT ages of Elias [30] indicate a three-phase cooling history for the Ötztal-Stubai Complex, as described in section 5. We now use an inverse-modeling approach to reproduce these cooling ages and to reconstruct the three-phase exhumation history in a quantitative way. Exhumation and material transport in all models occur in the vertical direction. In each model run, the exhumation rate remains constant during each of the three different exhumation phases. With a large number of 12,880 model runs, we explored the following parameter space. The models start at 110 Ma, and the first exhumation phase is allowed to end between 80 and 65 Ma. The exhumation rate during this first phase varies between 0.1 and 0.7 km/Ma. During the second phase, the exhumation rate is lower and was allowed to vary between 0 and 0.05 km/Myr. The third exhumation phase starts between 20 and 15 Ma with permissible exhumation rates ranging from 0.15 to 0.35 km/Myr. Please note that these five free parameters were varied simultaneously.The results of the inverse modeling are illustrated in three plots for each of the two elevation profiles (Figure 6). The horizontal and vertical axes of these plots show the parameter space that was explored. The two best-fit models—indicated by a white star—suggest a similar exhumation history for the Stubaital and the Kaunertal transects. For the first exhumation phase, the best-fit models yield exhumation rates of 0.27 ± 0.03 and 0.23 ± 0.03 km/Myr, respectively. The end of this phase (and the beginning of phase two) occurs at 67.9 ± 3.7 Ma (Stubaital) and 71.7 ± 3.0 Ma (Kaunertal), respectively (Figure 6(a) and(b)). The second exhumation phase is characterized by low exhumation rates of ~0.004 km/Myr (Stubaital) and ~0.007 km/Myr (Kaunertal), whereas the exhumation rates during the third phase are 0.23 ± 0.14 and 0.23 ± 0.02 km/Myr, respectively (Figure 6(a)and(b)). The onset of the third phase is nearly identical for both profiles (i.e., 16.0 ± 1.2 and 16.2 ± 1.2 Ma). The ZHe, AFT, and ZFT ages predicted by the two best-fit models are in good agreement with the cooling ages observed in the Stubaital and Kaunertal elevation profiles (Figure 7).Our new ZHe ages from the two elevation profiles in the central Ötztal-Stubai Complex range from ~70 to ~50 Ma (i.e., from 69 ± 3 to 56 ± 3 Ma in the Stubaital and from 71 ± 4 to 50 ± 2 Ma in the Kaunertal; Figures 4 and 5). Together with the additional ZHe ages from the Ötztal, the combined data set defines a narrow band in age-elevation space (online supplementary Figure S1), which indicates that the central part of the Ötztal-Stubai Complex behaved as a coherent crustal block since the Late Cretaceous. This interpretation is corroborated by our two best-fit thermokinematic models, which reveal a similar exhumation history for the two elevation profiles (Figure 6) and reproduce nearly all of the ZHe ages very well (Figure 7).Our thermokinematic models suggest that the cooling and exhumation histories of the Ötztal-Stubai Complex are well described by three distinct phases with markedly different exhumation rates (Figure 6). In the first phase, exhumation during the Late Cretaceous proceeded at a rate of ~250 m/Myr (i.e., 270 ± 30 and 230 ± 30 m/Myr for the Stubaital and Kaunertal models, respectively). This exhumation phase lasted until ~70 Ma (i.e., 68 ± 4 Ma in the Stubaital and 72 ± 3 Ma in the Kaunertal models) and was followed by a long period with slow exhumation during the Paleogene and Early Miocene. During this phase, our samples cooled slowly through the partial retention zone for helium in zircon, which explains the age difference of ~15–20 Ma between the highest and lowest samples in both profiles. Exhumation rates during this second phase were 4 ± 7 m/Myr (Stubaital) and 7 ± 6 m/Myr (Kaunertal; Figure 6). Despite the significant uncertainties of these values, our modeling clearly indicates that the exhumation rates were low. Higher exhumation rates during the second phase would lead to significantly younger ZHe cooling ages with a smaller age difference between the highest and lowest samples. Therefore, it is only possible to reproduce the observed cooling ages with our thermokinematic model when applying exhumation rates <10 m/Myr during the second exhumation phase. Such low rates may be somewhat surprising, but we note that similar erosion rates of 2–15 m/Myr have been measured in different present-day mountain settings using cosmogenic nuclides [91-93]. It is also worth noting that the exhumation rate during the second phase is significantly lower than the slopes of the regression lines in the two age-elevation plots (Figure 5). This observation highlights that it is not advisable to interpret such slopes as rates of exhumation (cf. 20, 52) because it is difficult, if not impossible, to show that the underlying assumptions are met (see section 5). According to our models, the third exhumation phase began in the Miocene at 16 ± 1 Ma (Figure 6). The exhumation rate during this phase is ~230 m/Myr and reproduces the AFT ages of samples P5/13 and P4/2 (which are located within the boundaries of the two models) very well (Figure 7). Finally, we note that in our models, the two changes in exhumation rate occur instantaneously, whereas in nature, a more gradual change, probably over 1–2 million years, appears to be more likely.By multiplying the duration of the three exhumation phases with the respective exhumation rates, we can quantify the total amount of exhumation from our two best-fit models. The models suggest that ~15 and ~13 km of rocks have been removed in the Stubaital and Kaunertal, respectively. Given the great similarity of the exhumation histories and the amounts of exhumation of the models for the Stubaital and Kaunertal, we use the single exhumation history depicted in Figure 8 as the basis for the following discussion.In this section, the three-phase exhumation history of the Ötztal-Stubai Complex derived from our thermokinematic modeling (Figure 8) is discussed in the context of the tectonic and geological evolution of the Eastern Alps. Following the consumption of the Meliata Ocean in an SE to E-dipping subduction zone, the stacking of the Austroalpine basement nappes and their sedimentary cover caused crustal shortening and thickening, which began in the Early Cretaceous (at ca. 135 Ma) and propagated from east to west through time [e.g., 39, 60, 94]. Parts of the Austroalpine nappes experienced an Eoalpine high-P/low-T metamorphism due to intracontinental subduction [e.g., 36, 37, 57, 65]. Using Lu–Hf dating of garnet, this eclogite facies metamorphism has recently been precisely dated at ca. 102–93 Ma at different localities to the east and south of the Tauern window [38], but its age remains poorly constrained in the southern Ötztal-Stubai Complex (Texel unit) due to the presence of garnet with Variscan cores and Eoalpine rims [95], which results in mixed ages [38].At the southern margin of the Ötztal-Stubai Complex, the ductile Vinschgau shear zone preserves evidence for top-to-the-W-directed thrust faulting and tectonic transport of the Ötztal-Stubai nappe above the Campo basement nappe (Figure 2) [54, 70]. The activity of this major shear zone has been correlated with the Trupchun deformation phase in eastern Switzerland and adjacent Italy, which lasted from ~100 Ma to ~80 Ma [3, 96]. Based on K/Ar dating of dynamically recrystallized mica and Rb/Sr isotopic studies on thin mylonite slices, an age of 85–80 Ma has been suggested for ductile shearing in the Vinschgau shear zone, which was active at decreasing temperatures [63]. In the central and northern part of the Ötztal-Stubai Complex, no thrust faults related to Eoalpine nappe stacking and crustal thickening are present. ZFT ages in our study area range from 92.0 ± 9.0 to 77.6 ± 3.5 Ma (Figure 3) [30] and show that cooling and exhumation of the central part of the Ötztal-Stubai Complex were already underway at the time when the Ötztal-Stubai nappe was thrusted toward the west. The results of our inverse thermokinematic modeling indicate that exhumation proceeded at a rate of ~250 m/Myr during this stage (Figure 8). We argue that the exhumation was most likely accomplished by the progressive erosion of the Eoalpine mountain belt although normal faulting may have also contributed to the exhumation as discussed in the next paragraph.The Schlinig fault, which defines the southwestern boundary of the Ötztal-Stubai Complex (Figure 2), was originally interpreted as a brittle, W-vergent thrust fault and the westward continuation of the Vinschgau shear zone [70 and references therein]. However, it was later shown that the Schlinig fault is actually an E-dipping, low-angle normal fault with a top-to-the-ESE shear sense that offsets the basal thrust fault of the Ötztal-Stubai nappe [62]. Another normal fault, the Arlui fault, forms the western margin of an occurrence of Mesozoic cover rocks in the hanging wall of the Schlinig fault (Figure 2) [62]. These normal faults were active in the Late Cretaceous between approximately 80 and 67 Ma (Ducan-Ela phase of extensional deformation; Figure 8) [3]. In the eastern part of the Ötztal-Stubai Complex, the tectonic contact between the Brenner Mesozoic in the footwall and the Steinach and Blaser nappes in the hanging wall is also a Late Cretaceous top-to-the-ESE normal fault (Figure 2) [31, 54]. ZFT and AFT ages in the eastern Ötztal-Stubai Complex combined with one-dimensional thermal modeling suggest a rapid cooling and exhumation of the fault footwall (700, 200 m/Myr) from greenschist-facies conditions to a temperature of ~100°C, which was reached at ~60 Ma (Figure 8) [31]. Since then, further exhumation occurred at a low rate of ~40 m/Myr [31]. The occurrence of Late Cretaceous normal faults east and southwest of our study area may be taken to indicate that normal faulting also contributed to the exhumation of the central Ötztal-Stubai Complex. So far, normal faults have not been mapped in this region, but minor normal faults may be difficult to detect because suitable marker horizons in the crystalline gneisses are lacking, and the Mesozoic sedimentary cover has been completely removed. In any case, the coincidence between the end of the first exhumation phase in our model at 70 ± 3 Ma and the end of the extensional Ducan-Ela phase at ~67 Ma [3] is striking (Figure 8).Most of the Late Cretaceous Gosau basins in the Northern Calcareous Alps (Figure 1) also document normal faulting and crustal extension [39, 40, 42, 97]. The Lower Gosau Subgroup (Coniac-Santon) consists of terrestrial alluvial to shallow marine deposits, whereas the Upper Gosau Subgroup (Campan-Maastricht) is characterized by deep-water clastics and hemipelagic slope sediments [40]. The rapid subsidence of the Gosau basins from ~85 Ma onward and the increase in water depth have been interpreted to indicate subduction erosion during oblique subduction of the South-Penninic Ocean beneath the Austroalpine nappe stack [40]. In the lower Campanian, the heavy mineral spectrum of the Gosau sediments changed markedly from chrome-spinel-dominated to mineral associations rich in garnet, staurolite, and chloritoid [98]. The presence of these metamorphic minerals—together with sediment transport directions toward the NW to NE—suggests that the Austroalpine nappe stack was an important source area for the Gosau sediments [40, 99]. Hence, we suggest that the sediments that were produced by erosional denudation of the Ötztal-Stubai Complex were transported northward and deposited in the South-Penninic Ocean.The pronounced decrease in exhumation rate at 70 ± 3 Ma—from a value of ~250 m/Myr to <10 m/Myr (Figure 8)—indicates that denudation of the central Ötztal-Stubai Complex by erosion (and possibly normal faulting) has slowed down by more than an order of magnitude. We suggest that this reduction reflects a significant decrease in the paleoelevation of the Eoalpine mountain belt due to prolonged erosion, late-orogenic normal faulting, and probably subduction erosion—the latter was already suggested for the Northern Calcareous Alps by Wagreich and Faupl [40] and Faupl and Wagreich [99]. The combined effect of these processes has presumably caused a substantial thinning of the continental crust of the Austroalpine mountain belt and thus decreased its elevation.The observation that the rate of exhumation remained low (<10 m/Myr) from the latest Cretaceous to the mid-Miocene (Figure 8) suggests that the Ötztal-Stubai Complex remained at a rather low elevation during the southeastward subduction of the Piemont Ocean underneath the Austroalpine nappe stack and during the early stage of the Alpine orogeny, which began with the subduction of the European passive margin in the Eocene at ca. 40–45 Ma [3, 33, 44, 100]. A low elevation prevailing during the early stage of continental subduction could be explained by the fact that the distal European margin had been highly extended during the opening of the Piemont Ocean [45]. Therefore, the subduction of the thinned European crust initially did not result in a recognizable uplift signal. The north-south shortening associated with this Early Tertiary collisional deformation has been referred to as Blaisun phase by Froitzheim et al. [3] and lasted until ~35 Ma (Figure 8).The inference that the Ötztal-Stubai Complex in the upper plate of the subduction zone had a low elevation during the Eocene is supported by the formation of an Eocene to Oligocene low-relief surface (the Dachstein paleosurface), which is locally preserved on elevated karst plateaus in the central and eastern parts of the Northern Calcareous Alps (Figure 8) [101]. The peneplanation of this paleosurface must have occurred at low elevation because it is sealed by Oligocene to earliest Miocene conglomerates and sandstones (Augenstein formation) [49, 101]. A low topography during Paleocene and Eocene time was also inferred for the Austroalpine nappes east of the Tauern Window based on detrital thermochronology, the formation of paleosols, and the deposition of lateritic and shallow marine sediments [102, 103].Our results indicate that the exhumation rate of the Ötztal-Stubai Complex increased significantly in the mid-Miocene (at ~16 Ma). Since then, about 4 km of rocks have been removed at an average rate of ~230 m/Myr (Figure 8). Hence, our modeling is consistent with the suggestion of a significant increase in surface elevation of the Eastern Alps in the Early Miocene, which has been attributed to the indentation of the Adriatic plate [100]. Apart from crustal thickening, surface uplift, and accelerated erosion, the indentation caused further north-south shortening of the mountain belt [47, 100, 104, 105] and the onset of orogen-parallel extension by a combination of strike-slip and normal faulting [20, 31, 39, 48, 52, 106-108]. One of these normal faults—the low-angle Brenner normal fault—bounds the Ötztal-Stubai Complex in the east (Figure 2). This west-dipping fault was active between ~19 and ~9 Ma and caused 35 ± 10 km of E-W extension [20, 52]. Given that the Ötztal-Stubai Complex is located in the hanging wall of the Brenner fault, its exhumation during normal faulting may seem counterintuitive. We note, however, that normal faulting in this setting does not cause subsidence and basin formation in the hanging wall block because both, the hanging wall and footwall, are continuously exhumed, albeit at different rates. As the Ötztal-Stubai-Complex is situated on the hanging wall of the Brenner fault and was not affected by internal deformation during the Neogene [107], its exhumation since ~16 Ma must have been accomplished by erosion and not by normal faulting. To the west of the Ötztal-Stubai Complex, the Engadine line constitutes a steeply dipping sinistral strike-slip fault with a normal-fault component (Figures 1 and 2) [109]. The fault was active during the Late Oligocene and Early Miocene and has accommodated N-S shortening as well as E-W extension during the indentation by the Adriatic plate and the related lateral escape of material toward the east [109].A similar three-stage exhumation history as presented in this study for the Ötztal-Stubai Complex was recently derived for the Gurktal Alps, which are located east of the Tauern Window (Figure 1) [21]. In the Gurktal Alps, normal faulting and erosion in the Late Cretaceous caused exhumation at a rate of 600 ± 60 m/Myr between ~100 and ~83 Ma. Afterward, the exhumation rate decreased to only 20 ± 10 m/Myr from ~83 to ~34 Ma before it increased again to ~160 km/Ma at ~34 Ma in the course of the Europe-Adria continental collision [21]. The long phase of slow exhumation between ca. 83 and 34 Ma is comparable to the second exhumation phase of our model (Figure 8) and indicates that other areas of the Eastern Alps were also barely affected by the Eocene phase of deformation that is related to the subduction of the European continental margin. Our results are thus broadly in agreement with those of Wölfler et al. [21] who proposed that the exhumation history of the different Austroalpine nappes is largely controlled by their structural position within the nappe pile and their distance to the Adriatic indenter.Topographic reconstructions suggest that during the Quaternary, the valley-scale relief of the European Alps increased by a factor of up to two [110]. The incision of large valleys proceeded at rates as high as 1–1.5 km/Myr [82, 83, 111]. Therefore, it seems likely that spatially averaged erosion rates over valleys and ridges have also increased during the glacial-interglacial cycles of the Quaternary. In the central and northern Ötztal-Stubai Complex, the presence of deep U-shaped valleys (Figure 3) also indicates a significant increase in local relief due to glacial erosion. Although this Quaternary relief increase is crudely captured by our thermokinematic models (see section 6.1), it remains unresolved if rates of exhumation and erosion have increased in the Quaternary. This is because the exhumation rate during each period of the modeled three-phase exhumation history is constant, and the last phase integrates over the past 16 Ma (Figure 8).In this study, we have reconstructed the cooling and exhumation histories of the Ötztal-Stubai Complex in the Eastern European Alps. Our new ZHe ages along two elevation profiles in the Kaunertal and Stubaital range from ~70 to ~50 Ma and increase with elevation. Inverse thermokinematic modeling of these ZHe ages together with few previously published AFT and ZFT ages indicate rapid exhumation (~250 m/Myr) during the Late Cretaceous due to progressive erosion of the Eoalpine mountain belt and late-orogenic normal faulting. This phase was followed by a long period of slow erosion (<10 m/Myr) between ~70 Ma and ~16 Ma, which indicates a rather low elevation of the study area after the Eoalpine orogeny. In the Miocene, exhumation accelerated to ~230 m/Myr due to surface uplift and enhanced erosion of the Ötztal-Stubai Complex in front of the Adriatic indenter.With respect to deriving exhumation histories from thermochronological data, our study highlights that cooling ages from closely spaced samples along elevation profiles in combination with thermokinematic modeling allow deciphering cooling and exhumation histories of crustal blocks in continental collision zones in detail. Compared with the conventional approach of using spatially distributed samples and thermal modeling of individual samples, this methodology has the advantage that thermokinematic inverse modeling allows deriving time-depth information from suites of samples and accounting for temporal changes of the temperature field due to heat advection and thermal relaxation.The code PECUBE by J. Braun used in this work is available at https://github.com/jeanbraun. All data used for this study can be found in the manuscript text, figures, and tables.The authors declare that there is no conflict of interest regarding the publication of this paper.This study has been supported by funds from the University Münster to Ralf Hetzel.