The relation between peak metamorphic temperatures and subsequent cooling during continent–continent collision (western Central Alps, Switzerland)

The maximum temperature (Tmax) and subsequent exhumation reflect the relations between advective and conductive heat transport, which in turn depend on the tectonic evolution. To unravel these relations in an orogen, precise Tmax data need to be combined with relative time information for the displacements of adjacent units. We present new Tmax data based on Raman spectroscopy of carbonaceous material (RSCM) and zircon fission track (FT) data, which are combined with previous data and then discussed jointly. We follow this approach in the Central Alps at the western edge of the Lepontine dome. Our analysis indicates two main tectono-metamorphic domains in this area: domain A comprises the Lower Helvetic units involving the Aar Massif; domain B is situated south of the Helvetic main thrust, in the footwall of the Simplon line. In domain A, thrusted Helvetic units were overprinted mainly by reverse faulting in the Aar Massif. The thermal evolution is related to the inversion of the former Doldenhorn basin. Tectonic transport during inversion brought into contact units with substantially different Tmax. Temperature gradients were then reduced by conductive heat transfer, but thermal overprinting during cooling involved subsequent vertical movements as well. Zircon FT data yield apparent ages between 12 and 18 Ma in the external part, but 8–9 Ma in the internal part of the Aar Massif. The youngest ages are taken as the cooling at a given temperature, whereas the other data are discussed as being only partially resetted along a temperature path in the partial annealing zone of the zircon FT. When combined with age data for Tmax and apatite FT data from the literature, the youngest group exhibits exhumation rates between 0.5 and 1.2 km/Ma in the time range between 20 Ma and today. In all of domain B, Tmax was significantly higher than in domain A. In domain B the estimated rates of exhumation are 0.8–1.0 km/Ma for the post-20 Ma time interval. Despite of different temperature evolution, the exhumation rates are similar in both domains. The study shows the necessity to combine detailed tectonic data to interpret the T–t evolution of such an area.


Introduction
The metamorphic field gradient in an orogenic belt reflects its time-integrated thermal history, which is tightly connected to the tectonic evolution. To understand such interrelations, metamorphic data are required for each kinematic unit. In order to interpret the metamorphic field gradient dynamically, precise P-T constraints should be combined with temporal information, notably age data correlating to well-defined stages of the tectonic evolution. Petrochronological data  pertaining to the thermal peak (T max ) are needed, as well as data for the temperature-time evolution during cooling and exhumation. For the former, thermally robust mineral chronometers are preferable, whereas various thermo-chronometers (e.g., Wagner and van de Haute 1992;Braun et al. 2006;Malusa and Fitzgerald 2019) have often been used for cooling and exhumation stages. At any stage of an orogenic evolution, tectonic movements between adjacent units in a given time interval can severely change the thermal structure (e.g., Shi and Wang 1987;Schmalholz and Duretz 2015;Jaquet et al. 2017). During post-tectonic cooling, isotherms disturbed by tectonics will relax due to conductive heat transport. Understanding such interactions requires detailed knowledge of the geometric relationships at different times.
Here, we combine tectonic, thermometric, and geochronological data from very low-grade to amphibolite facies conditions to gain insight into the thermal structure and its evolution, especially during cooling of upper crustal levels. Such evolutions have more commonly been investigated in medium-to high-grade metamorphic terrains (e.g., Kohn 2014), but relations in very low-and low-grade areas have hardly been tested, chiefly because suitable metamorphic and petrochronological data are rarely available. We study the western end of the Lepontine dome in the Central Alps, which represents one of the well-investigated examples of a Barrovian metamorphic terrain. Several studies are available that provide details on the metamorphic field gradient (e.g., Niggli and Niggli 1965;Frey et al. 1980Frey et al. , 1999Todd and Engi 1997;Engi 2011;Berger et al. 2011;Bousquet et al. 2008;Bousquet 2012;Nibourel et al. 2018). Overall, the field gradient shows spatially continuous zoning, but locally some mineral isograds (and isotherms) are offset along shear zones (e.g., Todd and Engi 1997). Such metamorphic discontinuities are most prominent near the western border of the Lepontine dome, which is the focus of this contribution. From north to south, the area studied comprises sedimentary units (Helvetic nappes) and two main basement complexes, i.e. the Aar Massif and the nappes in the Simplon area. Wedged within the latter are Valaisan metasediments, the only unit that experienced Eocene HP-LT metamorphism prior to the Oligocene-Miocene Barrovian overprint . Valaisan units in similar position at the NE-margin of the Lepontine dome show that this thermal overprint did not completely erase the Eocene imprint (Wiederkehr et al. 2009, but at the NW-margin this relation is not established (Fig. 1). In the area we studied, the overall metamorphic field gradient shows an increase from very low-grade to amphibolite facies conditions. The tectonic evolution in the area is known in some detail (e.g., Milnes 1973Milnes , 1974Steck 1984;Herwegh and Pfiffner 2005;Krayenbuhl and Steck 2009), but the relation to the temperature-time evolution remains less clear.
Our study adds results from Raman spectroscopy of carbonaceous material and zircon fission track data to the available metamorphic and geochronological data. The combined data set is then used to quantify the field gradient reflecting the thermal maximum and the conditions of subsequent cooling. The thermal evolution in the different tectonic units involved is discussed in relation to the known tectonic phases of the orogenic evolution. We show that deformation related to the exhumation of the Lepontine dome and the uplift in frontal parts of the Central Alps created the regional metamorphic imprint now visible.

Tectonic and metamorphic framework
The Barrovian metamorphic dome in the Central Alps displays a continuous field gradient from very low-grade to partial melting conditions (Frey et al. 1980(Frey et al. , 1999Todd and Engi 1997;Bousquet et al. 2008;Bousquet 2012). The transition from greenschist to amphibolite facies conditions occurs in the Northern Steep Belt (Milnes 1974) that separates the Lepontine nappe stack from the Aar Massif adjacent to the north (Fig. 1). The Aar Massif itself has been further subdivided into its internal and external parts, and the Gastern sub-massif (e.g., Steck et al. 2001;Krayenbuhl and Steck 2009;Berger et al. 2017). During Alpine orogeny, major sediment volumes were sheared off from the crystalline basement, producing the Helvetic nappes. The sedimentary sequences are described in several contributions (Herwegh and Pfiffner 2005 and references therein). The tectonic evolution has been subdivided into two main stages (Burkhard 1988;Herwegh and Pfiffner 2005;Pfiffner 2015; Fig. 2): (1) thrusting of the main Helvetic nappes (Prabé phase); (2) inversion of an underlying basin to produce the Doldenhorn Nappe (Kiental phase; Herwegh and Pfiffner 2005). Deformation included thrusting of the Doldenhorn Nappe above the Gastern sub-massif and basement units of the external Aar Massif. Typical fold and thrust geometries developed that include basement units in internal parts of this thrust sheet. This fold and thrust belt was then overprinted by the uplift of the Aar Massif (Grindelwald phase; Burkhard 1988). Doming and uplift of the basement is connected to the internal deformation of the Aar Massif, first by reverse faulting (Handegg phase; Wehrens et al. 2017) then by NW-directed thrusting above a basal thrust system (Pfaffenchopf phase; Berger et al. 2017;Herwegh et al. 2020). Coeval with Pfaffenchopf thrusting, dextral strike-slip faulting partly affected the southern border of the Aar Massif (Oberaar phase, Wehrens et al. 2017). The Handegg-, Pfaffenchopf-and Oberaar phases thus represent kinematic sub-events of the Grindelwald phase (Fig. 2).
East of Brig, several units of different paleogeographic provenience are involved: continental basement and sediments from the European margin and the Briançonnais, as well as Valaisan units. These units, separated by major orogenic thrusts (i.e. Pennine front, Helvetic main thrust), underwent basement thrusting and postnappe folding that involved both metasedimentary and basement units (e.g., Milnes 1974;Steck 1984;Sartori et al. 2017). Thrusting in this area occurred during an early stage, with coeval or slightly later isoclinal folding (D1 and D2 of Sartori et al. 2017). These deformations are responsible for nappe stacking in the Simplon area. This includes also Valaisan units, which represent the previously subducted accretionary prism (with HP-LT imprint). Further south, the Grand St. Bernard nappecomplex and equivalent gneiss sheets were interleaved with their sedimentary covers. This nappe stack subsequently became part of the lid of the future Helvetic domain (Fig. 2). The Gotthard Nappe reflects the lowest nappe of the European margin, which is thrusted on top of the Aar Massif. After the nappe geometry was established ( Fig. 2), the thickened crust was then folded. This major post-nappe folding (Berisal folds) is related to the uplift of the Aar Massif Ricchi et al. 2019;Herwegh et al. 2017). During and after development of these folds, the Rhône Simplon fault (= RSF) became active (e.g., Steck 1984;Mancktelow 1990Mancktelow , 1992Campani et al. 2010Campani et al. , 2014. The RSF has a pure low angle normal fault proportion, reported at the surface as Simplon line. It continuously changes into a lateral ramp known as RSF. In the following, the Simplon fault sensu strictu and the lateral ramp will be here named as RSF ( Figs. 1 and 3). At the western rim of the Aar Massif, owing to dextral slip movements, the RSF splits up into several minor branches (e.g., Campani et al. 2010;Berger et al. 2013). Movements along the RSF essentially occurred from 12 to 5 Ma, but strike-slip motion remained active up to recent times (e.g., Champagnac et al. 2003;Diehl et al. 2018).

Zircon fission track dating
Some zircon separates were available for dating from the samples of Reinecker et al. (2008). An additional sample set was separated at the University of Bern. This involved crushing by the SelFrag system (Giese et al. 2010). After zircon separation, minerals were mounted in PFA-Teflon and afterwards grinded and polished to expose internal  Niggli and Niggli (1965) and Bousquet et al. (2008). In addition, the isotherms from Engi (2011) are shown in °C. Note the end of all these data along the RSF (= Rhône Simplon fault). The area of Fig. 3 is indicated. The thick lines outline the Aar Massif and the Tavetsch and Gotthard nappes. The thin lines are the tectonic boundaries in the Lepontine surfaces. After attachment of an external mica detector, thermal neutron irradiation was carried out at the FRM-II reactor (Germany). Fission track analysis was made with an optical microscope (Zeiss Axioscope 2) under 1000× magnification using a dry objective. Ages were calculated using a zeta calibration factor (zeta = 101.8 ± 0.6 a/cm 2 , S. Erne-Schmid) determined on dosimeter glass IRMM-541 and Fish Canyon Tuff age standards. Calculation, visualization and statistics were performed using Trackkey 4.2 g (Dunkl 2002). All zircon fission track ages are displayed as central ages from 20 individual grains, errors shown are ± 1σ. The critical question, how this central zircon FT age can be used to constrain the thermal history of samples, depends on the temperature limits adopted for the partial annealing zone (= PAZ), notably its upper limit (= total stability zone) and lower limit (= total annealing zone). The absolute temperatures of both limits of the PAZ remain controversial. Here we considered limiting values of either 220° and 270 °C (Brandon et al. 1998 (Rahn et al. 2004). The behaviour of FT inside the PAZ depends on several factors, including the time interval spent inside the PAZ (Wagner and van de Haute 1992, see also chapter 5.2) and the accumulated radiation damage (e.g., Marsellos and Garver 2010). For instance, a shorter time spent inside the PAZ (faster cooling) results in reduced annealing and higher PAZ temperatures and FT ages. FT counting in zircons might be biased towards grains with 'countable' track densities (e.g., Rahn et al. 2019). Old and/or grains with high U content might have track densities too high to be counted accurately. Most of our samples are young and only those samples with considerable older ages (LBS-10 and LBS-11) do show low average U contents of counted grains and a general trend with lower U content associated with older ages and vice versa. This has been observed previously and interpreted to reflect radiation damage controlled differences in FT annealing (e.g., Marsellos and Garver 2010), with a negative relation between FT age and U content (or accumulated radiation damage).

Raman spectroscopy of carbonaceous material (RSCM)
Progressive heating induces ordering of carbonaceous material, changing the Raman signal. The partial ordering state of the crystalline structure can be quantified using characteristic Raman bands at certain wavenumbers (Wang et al. 1989). Beyssac et al. (2002a, b) showed that the peak intensity ratio and the peak area ratio, both calculated from bands in the first order region, are useful to quantify such changes. The peak area ratios are calibrated against temperature (Beyssac et al. 2002a, b). We performed micro-Raman spectroscopy at the Raman Laboratory of the Institute of Geological Sciences in Bern. The measurements were made using a Jobin-Yvon LabRAM HR800 instrument. A Nd-YAG continuous wave laser (20 mW beam spot of approximately 1 µm diameter, wavelength: 532.12 nm) was combined with an Olympus BX41 microscope at  Table 1 are shown. Isotherms are inferred. The 400 °C isotherm is based on the occurrence of Alpine biotite. See also isotherms and mineral occurrences in Fig. 1 100× magnification in confocal mode. The spectra were taken in the range between wavenumbers 1070 to 1750 cm −1 . Measurement times were 60 s. Peak fitting was done with the software "PEAKfit v4.06". The fitting itself used the Voigt area, the algorithm combined Gaussian and Lorentzian profiles. In each sample, at least ten and as many as 40 separate spots were analyzed (see Additional file 1: Appendix S1). Data sets for each sample show some scatter and occasional outliers, which probably reflect structural heterogeneity in the carbonaceous material. After removal of outliers (> 3σ), 1σ precision of the mean typically is 5-15 °C; the absolute accuracy of RSCM data is estimated at ± 50 °C. We use these measurements and calculate the weighted mean and the standard deviation as the error (see Additional file 1: Appendix S1 for more information).
Temperature data south of the Aar Massif are available from the Gotthard-and Valaisan-metasediments (Table 1; Hafner 2016). Temperatures are ~ 470-500 °C at the contact to the Aar Massif, they increase to 550 °C in the Monte-Leone Nappe and adjacent units, and they reach 575 °C south of the Berisal fold ( Fig. 3; Table 1). Toward the west, maximum temperatures were lower, reaching ~ 400-450 °C in the area of Visp (Fig. 3). In contrast to other thermochronometers the temperature data based on the RSCM method represent T max . In the Simplon area these data are consistent with metamorphic multi-equilibrium thermometry (e.g., Todd and Engi 1997).

Zircon fission track data
Available zircon FT data are reported in Tables 2 and  3 (Figs. 3 and 4). Our new data extend earlier literature data (Table 2; Michalski and Soom 1990) and are consistent with these. In the Lötschberg section (Fig. 4), from north to south, a general trend is observed from older to younger zircon FT ages. We subdivide these data into three groups (see Fig. 4, Table 2): (A1) apparent mixed ages (74-110 Ma); (A2) only partially resetted ages (12-18 Ma); and (A3) cooling ages (8-9 Ma). Data in group A1 show a clear relationship between age and age scatter, which indicates slow cooling and prolonged stay in the partial annealing zone.
Data for group A2 samples show some spread in ages, with an average of 14.5 Ma (Table 2; Fig. 4). Group A3 is limited to the southern rim of the Aar Massif, recording ages of ~ 8 Ma (Table 2; Fig. 4). This age group has been related to late stage movements along the RSF (Campani et al. 2010) and/or the movements of the Rote Chue Gampel fault ( Fig. 4; Sartori et al. 2017;Krayenbuhl and Steck 2009;Dolivo 1982). Our new thermochronological samples are from outcrops slightly north of the Rote Chue Gampel fault, and they appear to be related to strain distributed along this major fault (Sartori et al. 2017). In the footwall of the RSF, zircon FT ages are in the range of 10-14 Ma ( Table 2, Campani et al. 2010;Soom 1990).

Timing of T max
The timing of T max in the Lötschberg section is not well constrained because few chronometers pertain to such low temperatures, and data tend to be difficult to interpret. K/Ar and Ar/Ar data for metasediments at the base of the Doldenhorn Nappe yield mixed ages, indicating detrital input from a Variscan hinterland mixed with sheet silicates that developed during Alpine metamorphism (Kirschner et al. 2003;Frank and Stettler 1979;Fig. 5). These data show a clear relationship between the fraction of 2M polytype of mica/illite and the ages obtained (Fig. 5). This correlation gives a first order indication about when new sheet silicates had formed in these samples. Extrapolation of the data to 100% illite-2M indicates resetting between 14 and 21 Ma. The older age group is consistent with the youngest age in Kirschner et al. (2003). The base of the Doldenhorn Nappe reached T max before and during the Doldenhorn thrusting event (Kiental phase). The age of 14-21 Ma is consistent with the timing of T max further east in the Grimsel area (Bt Arage; Rolland et al. 2009).
Valaisan units from the NE-margin of the Lepontine belt show a subduction-related HP-LT imprint of Eocene age (41.2 ± 1.2 Ma; Ar-Ar in situ white mica age;  (Wiederkehr et al. 2009). Early Alpine HP-LT conditions are also known from the Valaisan units to the W and SW of the Simplon area; while not dated, these are presumed to be Eocene as well (Villa et al. 2014). This indicates different P-T paths for the Valaisan units compared to their surroundings (see Berger et al. 2011 for discussion). The potential nearly isothermal decompression and/or additional heating during Barrovian overprint allow a different T max in the Valaisan units as in the sediments without an HP-LT event. For the northern Lepontine belt in general, few petrochonological data date prograde mineral growth to amphibolite facies conditions. In the Simplon area (Steinental), Vance and O'Nions (1992) found prograde garnet growth using Sm/Nd methods between 32 and 25 Ma, but these data may not date the thermal peak. In the northern part of the Lepontine belt, T max is best constrained at 18-19 Ma (Janots et al. 2009), but only in samples some 50-70 km NE of Simplon pass. Th/Pb data in samples from the Robiei area (~ 17 km ENE of Simplon) show allanite growth at 20 Ma at T max (Boston et al. 2017). Further south (N of Domodossola) the same study found monazite growth at 22 Ma, also near T max . All of these late Oligocene to early Miocene ages reflect conditions that postdate nappe stacking, but are close to T max .

Thermo-tectonic domains
As mentioned in the introduction, understanding how a metamorphic field gradient formed, requires insight from the distribution of T max data and their age. The study area, from the Kander valley to the Simplon area, encompasses a wide range of metamorphic temperatures and tectonic evolution. These differences are not readily visible in the metamorphic field gradient documented by various studies, which seems to be continuous (Figs. 1 and 2; Bousquet 2012). In order to discuss the thermal evolution, we subdivide the area into following major tectonic domains (Fig. 6): (A) The Lower Helvetic nappes and the Aar Massif; (B) the footwall of the RSF south of the Gotthard Nappe; and (C) the hanging wall of the RSF, which will not be analysed in this contribution. The Upper Helvetic nappes are separated from the Lower Helvetic nappes by the Helvetic main thrust (Figs. 3 and 6). The units below this thrust include the Aar Massif and the Doldenhorn Nappe, i.e. our domain A. It is characterized by an eroded lid, owing to thrusting above the Lower Helvetic nappes and the Aar Massif before domain A reached T max . The exact timing for T max is a matter of debate, but relative age constraints show that T max in this domain occurred during or shortly after nappe stacking (e.g., Burkhard 1988;Herwegh and Pfiffner 2005). Domain B differs in this respect, as nappe stacking precedes T max by several millions of years Boston et al. 2017). A clear temperature hiatus is evident along the RSF (Bousquet 2012;Frey et al. 1999, Fig . 3).
In domain B, the metamorphic and structural imprint shows marked spatial gradients, indicating late-Alpine tectonic effects (Fig. 8, Campani et al. 2014). These are primarily due to the development of the Berisal fold with its axial plunge, the 3D projections of which are visible Comments: Italic numbers are the numbers of spectra analysed. Easting, Northing Swiss coordinates RSCM Raman spectroscopy of carbonaceous material, Cc-dol calcite-dolomite thermometry, ∂C carbon isotopic thermometry. n.g. not given. For methods see Additional file 1: Appendix S1 in the structural profiles (e.g., Schmidt and Preiswerk 1905;Milnes 1973;Steck 1984Steck , 2008 and in 3D structural models (Campani et al. 2014). A reorientation of the metamorphic field gradient has been proposed (Chatterjee 1961;Streckeisen et al. 1974) based on the combination of mineral assemblages observed in the Simplon tunnel with those documented from surface outcrops (up to 1 km above the tunnel-level).

Interpretation of the zircon FT ages
The zircon FT ages are grouped into: (A1) apparent ages between 74 and 117 Ma in the Gastern sub-massif; (A2) ages between 12 and 18 Ma in a central section; and (A3) ages between 8 and 9 Ma in the south (Fig. 4) (2001) shows a negatively-skewed length distribution, indicating also an incomplete or partial resetting of the zircon FT inside this sample, located west of the study area inside the Rawil depression (Leuk area; Rahn 2001). Incomplete or partial resetting of FT is also indicated by a large spread in zircon FT ages in group A2. In contrast, the samples of group A3 reached temperatures high enough for complete resetting of FT and cooled faster and therefore have tentative longer track length. The group A2 ages indicate an average of 14.5 Ma, but the observations summarized above suggest that this may be not a geological significant age. Sample KAW65 (not included in group A2) is also located in the area of group A2, but shows an age of only 9 Ma (Michalski and Soom 1990). This could be related to local complete resetting and/ or slightly different annealing kinetics (depending on U content, fluids, etc.). Therefore, we mention in the following group A2 as an average age of 14.5 Ma, but interpret this as an age with incompletely/partially reset FT. In contrast, the constant age of 8-9 Ma in group A3 is interpreted as a cooling age.

Domain A: Thermal and tectonic interplay in the Aar Massif
The above defined age groups in the Lötschberg section (Fig. 4) are compared with temperature data from nearby locations. The temperature data are limited to the metasediments, which are grouped into samples between the front of the Gastern sub-massif to the Lötschen valley (= group X) and samples near the southern rim of the Aar Massif (= group Y). Temperature data in group X are not from the same tectonic units as the zircon FT data of group A1. The mixed zircon FT ages from group A1 reflect lower T max than group A2. Using the results of the FT data and the different T-t evolution between group A1 and A2 samples implies that T max in this area must have been reached before the present day geometry was established, i.e. before the Kiental deformation (compare sample location marked by black stars with those shown by black circles in Fig. 7b). In a next stage, during Kiental thrusting, samples of group X (temperature) were brought into direct contact with group A1 samples. This is related of the tectonics of the inversion of the Doldenhorn basin, which includes decoupling between the basement and the sediments. Coevally, the Jungfraukeil developed at this deformation phase, as did the related mylonites (Herwegh and Pfiffner 2005). This phase is responsible for producing a new generation of sheet silicates, which yield ages of 20-22 Ma (Fig. 5, Kirschner et al. 2003). At this stage, the fold and thrust geometry of the Doldenhorn Nappe developed, and deformation occurred also in basement units indicating a change from thin-skinned to thick-skinned tectonics (Herwegh and Pfiffner 2005;Krayenbuhl and Steck 2009). This deformation bent the isotherms, an effect also found in thermomechanical models simulating the evolution of fold and thrust belts (e.g., Shi and Wang 1987;Schmalholz and Duretz 2015;Jaquet et al. 2017). The geometry and elapsed time assumed in a thermo-mechanical model of Jaquet et al. (2017) corresponds well with the situation of the Doldenhorn Nappe (Fig. 7, Jaquet et al. 2017 their Fig. 4). The above-mentioned transport between group A1 and group X samples developed during the thrusting of the Doldenhorn Nappe ( Fig. 7 compare a and  b). Reverse faulting movements along steep SE dipping The tectono-metamorphic domains of the study area. Note the repetition of the Pennine frontal thrust due to folding shear zones and foliation planes during the subsequent Handegg deformation phase allowed for partial conductive equilibration of the isotherms. This is indicated by similar T max in the Kander valley and in the Jungfraukeil, despite the teconically doming between these two areas ( Fig. 7c; Herwegh and Pfiffner 2005). The doming and related numerous zones of Handegg deformation allow the occurrence of the samples of group A1 and A2 at   Fig. 7 Tectonic evolution of Domain "A" inspired by Herwegh and Pfiffner (2005). Stars represent zircon FT samples. Circles represent samples used for thermometry in each group, respectively, a-d show different time steps of the tectonic evolution. a The present day close by locations of zircon FT samples of group A1 and T max samples of Group X have large distance in the situation before inversion of the Doldenhorn basin. e Redrawn from temperature modelling of Jaquet et al. (2017, their Fig. 4) different time intervals in the PAZ of the zircon FT (see also Sect. 5.2). The group A2 samples are more resetted as samples of group A1. During the combination of reverse faulting and the relaxation of the overturned isotherms by conduction, these samples passed through the zircon PAZ, but still did not allow complete resetting of the FT (see Sect. 5.2). The group A3 samples are related to temperatures of group Y and show faster differential uplift (Fig. 7d). The apatite FT ages of the same samples show also a difference between group A2 and A3, indicating differential movements between the internal and external Aar Massif (Reinecker et al. 2008).

When did the Valaisan units reach T max ?
Within the Simplon nappe stack (domain B), tectonic slices of Valaisan metasediments occur interleaved and post-nappe folded with other units (e.g., Bousquet et al. 2008). In corresponding units near the NE margin of the Lepontine dome, isotherms associated with the Barrovian overprint crosscut refolded nappe contacts (Fig. 1, Wiederkehr et al. 2008). In an area further east, several units-notably the Valaisan and the Adulareflect bimodal P-T paths, with some rocks retaining their HP-imprint. Hence these units attained T max earlier, in the Eocene (40-42 Ma, Wiederkehr et al. 2009, see also discussion in Villa et al. 2014) compared to the Barrovian overprint (16-19 Ma, Wiederkehr et al. 2009;Allaz et al. 2011;Boston et al. 2017). To the west of the Simplon area, RSCM data for three Valaisan samples from the north flank of the Rhône valley (Sion-Courmayeur Zone) indicate RSCM-temperatures of 361 ± 9 to 385 ± 9 °C (Table 1). In the Simplon area (between Brig and Monte Leone), sparse HP-relics occur in Valaisan rocks (e.g., chloritoid + pseudomorphs after carpholite; Bousquet et al. 2008), but no intact HP-assemblages have been reported. Contacts between Valaisan slivers and adjacent metasedimentary and gneissic units are sheared and jointly folded. This implies that the (post?-) Eocene decompression of the Valaisan slices preceded the formation of the Simplon nappe stack. In the eastern Lepontine, Wiederkehr et al. (2008Wiederkehr et al. ( , 2009Wiederkehr et al. ( , 2011 attributed decompression and nappe stacking to the same orogenic stage including different P-T paths depending on the amount of heating Berger et al. 2011;Roselle et al. 2002). Therefore, it is not clear at what stage the Valaisan units in the Simplon area reached T max . In an attempt to clarify the situation, RSCM data were scrutinized for each tectonic unit by Hafner (2016; using weighted kriging analysis). This study applies the weighted kriging to (a) the separate datasets from each unit, and (b) jointly for all of the units in the nappe stack (NE of the Simplon-Rhône line). The analysis showed that the resulting isotherm patterns for (a) and (b) are not significantly different. As the spatial T max pattern for the samples from Valaisan units is the same (within error) as for the adjacent units, we surmise that they are likely to be of the same age and thus that T max in domain B was attained during or after formation of the Simplon nappe stack.

The metamorphic field gradient
In external parts of the Lepontine belt, the metamorphic field gradient of the Barrovian metamorphism is oriented subparallel to the structural domains, such as the Aar Massif and the Gotthard Nappe (Fig. 1). This is well visible from the mapped mineral isograds and isotherms (e.g., Bousquet 2012;Todd and Engi 1997;Niggli and Niggli 1965;Fig . 1). In contrast, near the triple junction of the RSF, Helvetic thrust and the Pennine front in the Rhône valley, this simple relationship is lost (Fig. 8, Streckeisen et al. 1974;Milnes 1975;Chatterjee 1961). This perturbation may be related to: (1) tectonic transport after T max ; (2) a 3D sectional effect (compare Figs. 3 and 8, Campani et al. 2014); (3) variable timing of T max in different units.
It has been proposed early, that the metamorphic field gradient is steep in the area south of Brig (Streckeisen et al. 1974;Chatterjee 1961;see also Fig. 8). The projection of all available data (this study) into a profile allows such an interpretation (Fig. 8). At a larger spatial scale, the thermal field gradient is steep in the front of the Aar Massif (Mair et al. 2018;Nibourel 2019;Fig. 8). In southward direction, it shows a gradual increase to lower amphibolite facies, followed by a second marked increase in the region of the Berisal fold (Fig. 8). In any case, isotherms are bent and crosscut tectonic boundaries (Fig. 8). In order to explain such a field gradient, three possibilities are envisaged as idealized end member scenarios ( Fig. 9): (1) metamorphic data may reflect (static) thermal relaxation after thrusting and folding (Fig. 9e); (2) metamorphic data may reflect conditions after thrusting, but before folding (Fig. 9b), or (3) metamorphic data may reflect conditions that postdates thrusting and folding (Fig. 9c). In the area between the Aar Massif and the Berisal fold, the field gradient visible does not correspond to any such end member. Instead, a combination of thrusting and folding was involved (Figs. 7 and 9d). The inferred isotherms are a combination of thrusting, folding and conductive heat exchange at conditions near T max (Fig. 9d). This metamorphic situation is later followed by exhumation and cooling (Sect. 5.6).
Considering the temporal relations, the older nappe stack south of the Aar Massif (domain B) reached T max (at ~ 20 Ma), i.e. well after nappe stacking (e.g., Wiederkehr et al. 2009). However, T max was reached, while nappe formation occurred further north (Doldenhorn Nappe;domain A;Figs. 2,4,8). In domain B the T max structure was reached after decompression and nappe stacking (at ~ 20 Ma, Sect. 5.1). This T max structure was then modified by folding (Berisal fold) and subsequently underwent a limited overprint by conductive heat transfer (Fig. 8). The Berisal folding is connected in time and kinematics to the uplift of the Aar Massif (Handegg deformation-phase, Fig. 2, see also Campani et al. 2010Campani et al. , 2014.

Cooling/exhumation rates
As discussed above for the timing of T max , the cooling (and related exhumation) will give insights into the post-T max tectonics. The apparent cooling rates are derived from T max data and the assumed temperature value of the closure for the apatite and zircon FT (see also Sect. 5.2). Average values are estimated locally to obtain local cooling rates, which are then transferred into exhumation rates on the basis of an assumed average geothermal gradient of 25°/km. These calculations yield exhumation rates between 0.5 and 1.2 km/Ma ( Fig. 10; Table 4). We can further subdivide the area in the external and internal Aar Massif and the domain B including the area of the Monte Leone Nappe and the Berisal area (Figs. 7 and 10). This subdivision shows the general increase of cooling rate from the Aar Massif to the Monte Leone Nappe for the youngest interval (Fig. 10). The estimated exhumation rates in the Aar Massif are in agreement with available results of thermal modelling from Reinecker et al. (2008). These calculations yield exhumation rates between 0.5 and 1.2 km/Ma (Fig. 10, Table 4). We can further subdivide the area in the external and internal Aar Massif and the domain B including the area of the Monte Leone Nappe and the Berisal area (Figs. 7 and 10). The calculated exhumation rates in the Aar Massif are in agreement with available results from thermal modelling by Reinecker et al. (2008). In general, cooling initially following T max was slow, which may indicate dominantly conductive heat flow without substantial exhumation. In a second phase, cooling attended uplift and exhumation of the orogenic domains in their final tectonic geometry. The time interval of rapid cooling (and exhumation) changed over time, and the highest present day exhumation and cooling rates are located in the southern Aar Massif (e.g., Herwegh et al. 2020).

Conclusion
This study discusses: (1) how apparent cooling rates are influenced by tectonics in low-grade metamorphic units; and (2) how major tectonic boundaries were active at different times leading to a separation into thermo-tectonic domains: RSCM data were projected into the plane of the cross section. Color coding for temperature is shown above the profile. Small inset map (top right) shows metamorphic reaction isograds and maximum isotherms within the Aar/Gotthard massifs and northern Lepontine dome and the traces of the two cross section parts. For reference only, a geothermal gradient of 25 °C/km is shown to the left of the section. Cross section A-A′ inspired by Ziegler and Isler (2013) and Herwegh and Pfiffner (2005), cross section B-B′ modified after Campani et al. (2014). The data point at the 600 °C isotherm in the south is based on sample Si9410 of Todd and Engi (1997). The slight shift between the contour line and the single data point is based on the contouring. ZSC Zone of Sion Courmayeur (Valaisan), FAP fold axial plane • In domain A, the sediments in the Lower Helvetic units (Aar Massif and Doldenhorn Nappe) record the inversion of a sedimentary basin with a thrust and fold geometry and subsequent thermal equilibration by conduction. These processes occurred in a time interval of 5-10 million years, whereas the developed geometry is maintained during subsequent exhumation in a block-like manner and over a longer time span.
• In domain B a deeper crustal level is exhumed, where the nappe stack developed over several million years related to deformation of D1 and D2 in the sense of Sartori et al. (2017) (Fig. 2), and T max was reached after nappe stacking. The thermal structure of domain B was folded at km-scale at conditions near T max , which did not result in a passively folded T max structure, since the metamorphic field gradient may have been slightly modified after folding, but prior to regional cooling. • While nappe stacking occurred in domain A, T max was reached in domain B several million years after nappe stacking in this area. • The isotherms were locally bent due to late exhumation of the Aar Massif (including Berisal folding). • Exhumation rates estimated from FT data range between 0.7 and 1.5 km/Ma.
Data using the RSCM method have proven to be suitable to refine and extend data on T max in a classic Alpine terrain that has seen a complex tectonic and metamorphic evolution. To understand the metamorphic imprint and field gradient properly, it is critical to combine thermal data with low-temperature age constraints, which are here based mostly on zircon and apatite FT data. A complete discussion of the interplay between deformation and thermal effects would require a 3D view of the evolution, taking into account the influence of strike-slip faults in particular. However, the data presented here confirm the sensitive interdependence of the tectonic and metamorphic evolution, which is well established for high-grade terrains (e.g., Nepal Himalaya, Kohn 2014), but it is here confirmed also at low to medium grade.  Table 4)