Modeling surface GPS velocities in the Southern and Eastern Alps by finite dislocations at crustal depths

Tectonophysics 590 (2013) 136–150

Contents lists available at SciVerse ScienceDirect

Tectonophysics journal homepage: www.elsevier.com/locate/tecto

Modeling surface GPS velocities in the Southern and Eastern Alps by finite dislocations at crustal depths A. Caporali a,⁎, F. Neubauer b, L. Ostini a, G. Stangl c, D. Zuliani d a

Department of Geosciences, University of Padova, Via Gradenigo 6, I-35131 Padova, Italy Department of Geography and Geology, University of Salzburg, Hellbrunnerstr, 34, A-5020 Salzburg, Austria Austrian Academy of Sciences and BEV, Schmiedlstr, 6, A-8042 Graz, Austria d Istituto Nazionale di Oceanografia e Geofisica Sperimentale, Centro Ricerche Sismologiche, Via Treviso 55, I-31100Udine, Italy b c

a r t i c l e

i n f o

Article history: Received 3 August 2012 Received in revised form 15 January 2013 Accepted 22 January 2013 Available online 31 January 2013 Keywords: Southern Eastern Alps Tauern Window Adria indenter Seismicity Crustal deformation Shear heating

a b s t r a c t The indentation of the Adria plate into the Southern and Eastern Alps is an ongoing collisional process accompanied by seismicity, surface and rock uplift and lateral escape. We present a 3D quantitative description of the process by combining GPS and structural data with an elastic dislocation model. Horizontal velocities of 70 Austrian and Italian permanent GPS stations in the Eastern and Southern Alps serve as boundary condition on the free surface of an elastic half space containing six rectangular faults, each with a uniform slip rate. The geometry of the rectangular faults and the slip rate vector are constrained by least squares, taking into account the structural setting of the area and the geographic distribution of the velocity data. We find that the surface velocities of the order of some mm/yr require reverse (North side of the Tauern Window), transpressional (Giudicarie, North Alpine Wrench Corridor, Pustertal, Dinarides) and normal (Brenner fault) slips ranging from 10 to 30 mm/yr at crustal depths. The regional stress pattern computed from fault plane solutions agrees with the principal directions of our rectangular fault planes. The model, although constrained by horizontal velocities only, predicts a pattern of vertical motion, which qualitatively agrees with known phenomena such as the surface uplift in the Tauern Window area, of the order of up to few mm/yr. If the heat on the shearing fault planes is removed mostly by upwards diffusion, the absence of large heat anomalies on the Earth's surface suggests, for nominal geotherms, shear stresses and concentration of subcrustal radiogenic elements, that the time of initiation of the slip dates to Pliocene, hence more recent than Late Oligocene–Miocene time of collision of the Adria indenter. © 2013 Elsevier B.V. All rights reserved.

1. Introduction In the Eastern and Southern Alps a complex fault geometry accommodates the northward indentation of the Adria plate, the uplift of the Tauern Window, and a lateral extrusion towards the Pannonian Basin. The northeastern edge of the Adria is associated with the seismically active Friuli (Anderson and Jackson, 1987; Bressan et al., 1998, 2007; Schmid et al., 2004, 2008). The compressive Mw = 6.5 earthquake of May 6, 1976 is the largest recorded event in Friuli. The area of the Southern and Eastern Alps is characterized by significant crustal thickness variations (Brückl et al., 2007, 2010). Seismic profiles show that the Eurasian and Adriatic plates interact with a thinner Pannonian unit as a structurally separated entity (Brückl et al., 2007, 2010). Late Oligocene–Middle Miocene indentation tectonics is considered as the primary agent driving substantial lateral material transfer, or “lateral extrusion” (Neubauer et al., 2000; Ratschbacher et al., 1991a; Willingshofer and Cloetingh, 2003; Wölfler et al., 2011). The indentation tectonics and resulting lateral extrusion is driven by a free ⁎ Corresponding author. Tel.: +39 49 8279122. E-mail address: [email protected] (A. Caporali). 0040-1951/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tecto.2013.01.016

boundary in the east due to subduction of a remnant land-locked basin within the present-day Carpathians (e.g., Wortel and Spakman, 2000). From the structural point of view, increasing evidence demonstrates inversion of the entire Alpine–Carpathian–Pannonian system at ca. Miocene/Pliocene boundary (ca. at 5 Myr before present, e.g., Peresson and Decker, 1997) and the sudden onset of surface uplift (e.g., Genser et al., 2007; Hergarten et al., 2010; Wagner et al., 2010). Extension and related normal and transtensional faults as well as subsidence in sectors of the Eastern Alps including the Pannonian Basin were replaced by E–W shortening structures and related surface uplift. In analog experiments Adria has been modeled as a rigid indenter relative to the indented Eastern Alps, in the sense that the deformable area is narrow relative to the size of the indenter (Ratschbacher et al., 1991a). Robl and Stüwe (2005) point out that the assumption of rigidity can to some extent be relaxed, and favor a model in which the velocities of GPS stations are described with a Newtonian fluid. Analog experiments (Ratschbacher et al., 1991b; Regenauer-Lieb and Petit, 1997; Rosenberg et al., 2007) have shown that the deformation resulting from the indentation of the rigid Adria into a plastic Eurasia depends on the ratio of the width of the indenter and the horizontal extent of the deformable foreland in the direction of the indentation. These

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

analog experiments have shed light on the way the Alpine collision cuts the European lithosphere and is therefore linked to intraplate grabens, e.g., the Upper Rhine valley. Using the analytic model of finite dislocation in an elastic half space of Okada (1985) and a dense set of GPS velocities, we present in this paper a first map of slip at depth. Based on structural surface data we identify a number of rectangular faults, which approximate the largest geological structures and which have the potential to accommodate the slip required to fit the GPS data. We show that the GPS velocities, of the order of a few mm/yr, are consistent with a dynamic process where indentation and lateral extrusion involve slips at crustal depths ranging between 10 and 30 mm/yr. The available velocity data are consistent with slip taking place on at least six rectangular faults located in the upper crust: we associate them primarily with the Giudicarie Fault,

a

137

the North Alpine Wrench Corridor, the Brenner fault, the SEMP fault along the northern part of the Tauern Window, the Pustertal–Gailtal fault and the Dinarides, e.g., the Idrija and the eastern segments of the Fella–Save faults. Whereas most of these zones represent well defined fault zones, the North Alpine Wrench Corridor corresponds to a zone of distributed seismicity close to the northern margin of Eastern Alps (Lenhardt et al., 2007; Reinecker and Lenhardt, 1999) and consequently a zone of distributed deformation. Using the root mean square (r.m.s.) of the (observed-minus-modeled) velocities as an indicator of the goodness of the fit, we constrain the geometrical parameters of the rectangular faults and slip rates. There are nine parameters for each fault: three dimensional coordinates of an origin, length and width of the fault, two-dimensional slip vector, strike and dip angles. The least squares adjustment is done on a neighborhood of a priori values of these parameters,

12˚

14˚

C

NAW

48˚

48˚

Salzburg

Alpine front

P

SEM

Innsbruck 030721

TW

Tauern window

La

Bf KB

Eastern Alps

Eastern Alps

010717

MF

PG

G

o

FS

G

i

Bolzano

760506

Southern Alps

980412 040712

Southern Alps

760915

Id

F

760617

Friuli

46˚

D

770916

041124

020602

Trieste

DI

es id ar

in

Mo

46˚

980831

Fr

SV

Mw=5

Venice

Adria

Mw=6

Padova Mw=7

12˚

14˚

Fig. 1. a: Large scale fault structure in the Eastern–Southern Alps and recent seismicity. To = Tonale fault, PG = Pustertal–Gailtal, Gi = Giudicarie fault, TW = Tauern Window, NAWC = North Alpine Wrench Corridor, La = Lavant Fault, MF = Möll Valley–Hochstuhl fault, Di = Dinarides, Mo = Montello, Fr = Friuli, SEMP = Salzachtal–Ennstal–Mariazell– Puchberg fault, Go = Görtschitztal fault, IdF = Idrija fault, KF= Katschberg Fault. Vergence and faults style are from the structural model of Italy (CNR, 1990). CMT mechanisms of events from 1976 to 1999 (black compression) are from the Harvard CMT Catalogue. Later CMT solutions (blue compression) are from the Regional Catalogue of Pondrelli et al. (2011). The events are labeled by their date in the format yymmdd. Historic events are taken from the CPTI04 Catalogue (blue solid dots: Gruppo di lavoro CPTI04 (2004) and from the ZAMG Catalogue (http://www.zamg.ac.at/forschung/geophysik/erdbebenforschung/: light blue solid dots). b: Index map for a.

138

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

Fig. 1 (continued).

which come from independent information such as structural geology, average direction of P/T axes of fault plane solutions, and regional strain rate field from GPS velocities. Shear heating is likely to increase the temperature on the fault planes. The resulting heat flow adds to that generated by radiogenic sources in the upper continental crust. The available data (Clark, 1961; Della Vedova et al., 2001; Viganò et al., 2008, 2011) indicate that the total heat flow observed on the surface does not exceed 60 to 80 mW/m 2, although in middle Jurassic values as high as 85 to

105 mW/m2 have been reported by Carminati et al. (2010) using organic matter maturity data from outcropping sediments. Following Turcotte and Schubert (2002) we model the surface heat flow and temperature increase on the fault planes as due to a sudden increase of heat flow caused by shear heating on the fault plane. It is reasonable to assume that the total temperature (for a nominal thermal gradient plus shear heating) on the fault plane is in the range 600–800 °C (Vosteen et al., 2006, for the TRANSALP profile), and that the heat flow on the Earth's surface, when added to the radiogenic

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

in the Plio-Pleistocene, hence more recent than the late Oligocene to Miocene collision of the Adria indenter with the stable European foreland. According to the analog models of Ratschbacher et al. (1991a), this collision is responsible for the fold and fault structure. Hence it can be concluded that the collision was ‘head on’ in the first 15–18 Myr, and has been accommodated by slip on inclined fault planes only in the past 5–7 Myr, since a longer slipping phase would imply an exceedingly large amount of frictionally generated heat to reach the Earth's surface. Specific assumptions on the local geotherm are clearly needed to make these concepts more quantitative.

Table 1 The two solutions generating the velocities used in this study. The analysis of the raw GPS data yielding weekly normal equations and the normal equation stacking yielding the velocities from the time series were done with the Bernese 5.0 software (Dach et al., 2007). A total of 70 stations in the Lat/Long interval (44.5–49.5; 10–15) were used in the fit. Network Time span (GPS week)

# of processed stations

Austrian

119

5

ITRF2005

229

28

ITRF2005

Italian

1106–1565 (2001.03.18– 2010.01.09) 995–1615 (1999.01.31– 2010.12.25)

ITRF stations used for frame alignment

139

Eurasian pole for velocity reduction

2. Geological setting The Eastern Alps represent the Late Eocene–Oligocene collision zone between the stable European plate in northern lower plate position and the overriding Adriatic plate. During Late Oligocene–Miocene (e.g., Handy et al., 2010 for a recent review), due to indentation of Adria and its frontal Southalpine unit, N–S shortening gradually changed to eastwards lateral extrusion accommodated by a conjugate set of mainly

heat (of the order of 50 mW/m 2), does not exceed 60–80 mW/m2. The model then implies that the time of initiation of heat production by shear in the half space must be relatively recent, for shear stresses in the range 100–300 MPa. We infer that such epoch should be somewhere

12˚

14˚ PFRK LINZ RIED

ASCH

WELS

C

NAW

48˚

48˚

GMND ROSE

FRLG SBGZ

BAD2

Salzburg WIND

Alpine front

MP

BGDN

SE

RTMN

GRM2

Innsbruck

MUEN

KTZ2 HKBL

HFL2 PATK

TW

Tauern window

TAMS

La

KOLM

Bf

PRBZ

KB

STBZ

Eastern Alps

Eastern Alps

MF

BRBZ

PG

SILL

TREI

G

o

MABZ

FS AFAL

BZRG

LAN2 VILH ZOUF

Bolzano

SARN

KLA2

ACOM TARV

Southern Alps RADO

PEJO

G

i

FDOS

Southern Alps

MPRA BARC

Id

JOAN

F

MOCA

46˚

Friuli

SCHI

TRIE

TREV BOSC

BEVA SDNA

Trieste

DI

es id ar

PAZO

PARR

CITT

46˚

Fr MDEA GORI PALM

in

Mo

ASIA

CODR

D

ROVE

SV

VICE CAVA PADO TEOLPadova

12˚

Adria 3 mm/yr

Venice

14˚

Fig. 2. The geographical distribution of the GPS velocities used in this study, referred to the Eurasian plate, after removal of a mean residual velocity.

140

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

strike-slip faults, the dextral Periadriatic fault and the sinistral Mur-Mürz and Salzachtal–Ennstal–Mariazell–Puchberg (SEMP) faults (e. g., Ratschbacher et al., 1991b). Lateral extrusion interfered with contemporaneous east-directed slab rollback in the Pannonian–Carpathian realm (e.g., Peresson and Decker, 1997). Miocene–Pliocene exhumation and surface uplift of the Tauern Window area, which exposes a piece of detached European continental basement, was accommodated by the sub-Tauern thrust ramp indicating continuing shortening (Cassinis, 2006; TRANSALP Working Group, 2002). Little is known about the present-day processes shaping the Eastern and Southern Alps. For example, it is unclear if the Late Miocene–Recent horizontal and vertical motions in the Alpine Dinaric region are controlled by deep lithospheric, post-glacial isostatic adjustment or surface/climatic processes (Cederbom et al., 2004; Rosenberg and Berger, 2009; Willett et al., 2006). Another open question is how do NW-striking dextral fault systems in the Alpine–Dinaric region (example: the Lavanttal and Möll Valley–Hochstuhl fault systems to the East, and the Schio–Vicenza fault to the West) kinematically link with E-striking thrust faults in Friuli, the dextral Periadriatic fault (example: Pustertal–Gailtal fault segments of the Periadriatic fault), and the left-lateral Salzachtal–Ennstal–Mariazell–Puchberg (SEMP) fault. It is also unclear how that kinematic transition is topographically expressed (Fig. 1a). In the northern part of the eastwards extruding block, the SEMP fault has been proven active (Plan et al., 2010) and also active in the Late Pleistocene (Keil and Neubauer, 2011a). This fault consists of several branches as Keil and Neubauer (2011b) recently demonstrated. In any case both E–W extension and E–W compression have been recorded for Pliocene–Quaternary deformation in the Eastern Alps (e.g., Peresson and Decker, 1997). There is seismologic, geologic and geodetic indication of the existence of a triple junction east of the Tauern Window, joining the European, Adria and a thin Pannonian microplate (Brückl et al., 2007, 2010). According to the palinspastic reconstruction by Linzer et al. (2002), the average velocity of extrusion east of the Tauern Window was 5 to 6 mm/yr during the most active phase in the Oligocene and Miocene. However geodetic data show that recent extrusion occurs at much lower velocities of only about 1 mm/yr (Caporali et al., 2009; Grenerczy and Kenyeres, 2006; Weber et al., 2006). Based on analog experiments, Rosenberg et al. (2007) report that the lateral extrusion velocity can reach at most 20% of the indenter velocity. How much slip, if any, is today involved along the Pustertal–Gailtal (Periadriatic) and the SEMP faults is yet unclear. According to the lateral extrusion model (Castellarin et al., 2006; TRANSALP Working Group, 2002), thrusting along the sub-Tauern ramp has been a major component of the exhumation of the Tauern Window area. Between Adria and the Southern Eastern Alps, convergence of 3 mm/yr is expected to result in an uplift rate smaller than 1 mm/yr. According to Barletta et al. (2006), such value could be significantly increased by the post-glacial rebound or the effect of recent glacier melting. By contrast, there is little or no knowledge on the amount of present-day strike slip on the lateral fault systems, such as the Schio–Vicenza and Giudicarie faults in the W (Massironi et al., 2006) and the northern Dinarides in the E (Carulli et al., 1990). Here the present-day seismicity is lower than in the Friuli region, which contains the tip of the indenter. In the neighborhood of the Mölltal fault at least two events of M = 6.1 have occurred, in 1201 and 1690, according to the historical Catalogue at the Austrian Zentralanstalt für Meteorologie und Geodynamik (http://www.zamg.ac.at/forschung/ geophysik/erdbebenforschung). On the northern edge of the Giudicarie fault, a sinistral strike-slip earthquake of Mw = 4.9 occurred in 2001 near the town of Merano (Caporali et al., 2005). According to recent geodetic and paleomagnetic data, the Adria microplate can be considered a separate structure from the Nubia plate (Battaglia et al., 2004; Marton et al., 2011). The departures from rigidity of the crust at or near plate boundaries cause specific patterns in the velocities of GNSS stations in these areas and provide a boundary

condition for the deformation at depth. Faults and slip rates have been mapped and are described in detail in the Database of Individual Seismogenic Sources (Basili et al., 2008; Benedetti et al., 2000; Galadini et al., 2005). The fault profiles of Galadini et al. (2005) were used by Bechtold et al. (2009), in conjunction with reliable GPS velocities, to map areas of excess strain and identify associated active faults. Detailed profiles in the N–S and E–W directions have been published on the basis of dense GPS velocity solutions (Bechtold et al., 2009; Caporali et al., 2009; D'Agostino et al., 2005, 2008), but we do not know about the present-day causative slip at depth. It is then natural to ask if such patterns of superficial horizontal velocities could be due to slip on a limited number of rectangular faults. Their geometry (position, orientation, size) and amount of slip should correlate with the tectonic structure of the area, which is independently known from seismology and structural geology. 3. Data, method of analysis and accuracy 3.1. Data We used velocities data resulting from the multi-year cumulative solutions done at the Austrian Academy of Sciences in Graz and at the University of Padova (Table 1). Both are active as EUREF Analysis Centers and use the processing standards recommended by the International GNSS (Global Navigation Satellite Systems) Service (http://igscb.jpl.nasa.gov/) and the European Permanent Network of EUREF (http://www.epncb.oma.be/_organisation/guidelines/ guidelines_analysis_centres.pdf). The common processing strategies and a common frame for position and velocities are such that the velocities from the different solutions form one homogenous set. Small discrepancies may exist due to the different time spans, for example, but these are typically smaller than 0.5 mm/yr and have a negligible impact on the final interpretation. The estimated velocities are shown on Fig. 2. All the velocities were computed in ITRF2005 (Altamimi et al., 2007) by means of the Bernese software 5.0 (Dach et al., 2007) and were reduced to a Eurasian reference frame with the same Euler rotation vector. The mean of these European fixed velocities was further subtracted, so that the velocities used to fit the model describe the internal deformation of a block at rest. Formal error ellipses have been rescaled to account for colored noise in the coordinate time series (Caporali, 2003). The GPS data fit reasonably well with observations on structures of Quaternary to recent sedimentary formations. The NNW direction of the Veneto–Friuli earthquake belt is well monitored by Upper Pliocene to Holocene conglomerates (Caputo et al., 2010). The principal NNW motion direction extends

U

N u

α

E

h d2 W

d3

d d1 L

δ

Fig. 3. The geometry of the Okada model of dislocation in an elastic half space, relating the displacement u̵ measured at the surface by GPS geodesy to the slip d at depth, constant on the rectangular patch. The N, E, and U directions define the local reference system at the surface (North, East, Up).

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

towards the NW corner of the Southalpine unit and across the Tauern Window, although rates are decreasing. There are, in a broad sense, four distinct domains of velocities (Fig. 2). Domain 1: In the western part of the figure (Giudicarie fault area) the velocity vectors tend to drift NW, in very good agreement with the direction of the most compressional axis of the stress tensor resulting from the inversion of seismological events (see Fig. 10 in Bressan et al., 1998). Domain 2: South of the Pustertal–Gailtal fault there is a dominant northwards component, indicative of the ongoing indentation on the Adria plate and strike slip along the Dinarides. Domain 3: On the northern part, N of NAWC (North Alpine Wrench Corridor), the velocities are smaller in size and heading south. This resembles the back-slip which is opposite to the direction of plate convergence (Savage, 1983). A relatively small back-slip can be interpreted as an indication of a partially locked interface.

a

141

Domain 4: On the central-east part of the area, between NAWC and the Pustertal fault, there is a marked component in the east direction, as already noted earlier (Caporali et al., 2009; Grenerczy and Kenyeres, 2006; Weber et al., 2006), suggesting that an eastward motion is still active at present (Brückl et al., 2010). Besides these four domains, some additional features can be recognized: in the eastern part of the Tauern Window (TW), east of the Möll Valley–Hochstuhl fault, NE motion has been found. Together, these data imply about E–W transtension in that area. Extensive late-stage E–W extensional structures including conjugate Mohr shear fractures, although poorly dated, were demonstrated by Wang and Neubauer (1998). Reiter et al. (2004) also found evidence for ca. E–W extensional neotectonic activity along the Brenner fault. The area of the Brenner fault, on the western side of the Tauern Window, is likely to be subject to E–W extension (Fügenschuh et al., 1997; Fügenschuh et al., 2011; Rosenberg and Garcia, 2011).

12˚

14˚ PFRK

LINZ RIED

ASCH

WELS

C

NAW

48˚

2

48˚

GMND

ROSE

FRLG SBGZ

BAD2

Salzburg WIND

Alpine front

P

SEM

BGDN

RTMN

GRM2 MUEN

Innsbruck

4KTZ2 HKBL

HFL2 PATK

La

TauernPRBZ window

Bf

KOLM

KB

Eastern Alps

TW TAMS

5

STBZ

Eastern Alps

BRBZ

3

SILL

MF

PG

TREI

G

o

MABZ

Bolzano

BZRG

FS ZOUF

AFAL

LAN2 VILH

SARN

Southern Alps RADO

1

PEJO

KLA2

ACOM TARV

FDOS Southern Alps

G

i

MPRA BARC

Id

JOAN

F

MOCA

46˚

FriuliCODR PAZO

PARR

6

TRIE

TREV CITT

BEVA

Trieste

SDNA

DI

es

SCHI

BOSC

46˚

GORI

id ar in

Mo

ASIA

Fr MDEA PALM

D

ROVE

SV

VICE CAVA PADO Padova TEOL

Venice

Adria

3 mm/yr

12˚

14˚

Fig. 4. a. Geometry of the six rectangular slip planes (gray rectangles) projected onto the topographic surface. The gray line parallel to one of the sides of each rectangle represents the intersection of the fault plane with the topographic surface. The ‘beach balls’ give a pictorial view of the data in Table 2. The number in black above each beach ball refers to the indexing in Table 2. The blue arrows represent the measured velocities with 1σ error ellipse, and the white arrows the predicted model velocities based on Table 2. b. Gridded model to highlight the surface flow associated with the slip at depth.

142

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

b

12˚

14˚

C

NAW

48˚

48˚

P

SEM

TW

La

Bf

G

o

MF

G

i

PG

46˚

46˚

Fr

Mo DI SV

3 mm/yr

12˚

14˚ Fig. 4 (continued).

3.2. Method of analysis and accuracy To estimate which displacements at depth are associated to the observed surface velocities, we use the Okada (1985) model of displacement of an elastic half space, in terms of uniform slip on a rectangular fault. The extension of the considered area and the spatial resolution of

the available velocity data suggest that several rectangular faults need to be introduced. The model velocities on the surface are computed as the vector sum of the displacements associated to each rectangular fault. A total of nine independent parameters need to be specified for each rectangular fault (Fig. 3): the position (latitude/longitude/depth) of the center of the rectangle; the strike α and dip δ angles; the length L

Table 2 The nine parameters of each rectangular faults used in the analysis. The first three columns (Lat, Long, depth) identify the center of the rectangular fault. The last column gives the product of the previous three columns (slip area x slip rate) times the shear modulus μ = 30 GPa. Uncertainties (1σ), in the sense of square root of the corresponding element in the variance covariance matrix scaled by the r.m.s. of the post fit residuals, are: 0.05° for longitude and latitude, 2 km for depth, 3° for strike and dip, 0.005 m/yr for slip, 5 km and 2 km for length and respectively width of the fault. Fault id.

Name

Long. (deg)

Lat. (deg)

Depth (km)

Strike (deg)

Dip (deg)

Right lat. (m/yr)

Reverse (m/yr)

Length (km)

Width (km)

Moment rate 1018 J/yr

1 2 3 4 5 6

Giudicarie NAWC Pustertal TW north Brenner fault Dinarides

11.04 12.18 12.91 12.22 11.49 14.13

46.31 47.85 46.68 47.34 47.02 45.62

30 10 20 8 2 5

211 84 282 261 188 312

80 60 89 89 45 45

−0.01 −0.01 0.01 0.00 0.00 0.01

0.03 0.02 0.01 0.02 −0.02 0.01

58 297 123 61 35 122

10 12 30 20 4 14

0.56 2.02 1.41 0.55 0.08 0.73

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

143

a

b

Fig. 5. a. Three dimensional view of the model, seen from West. The rectangular faults are white rectangles, and a thicker white line on the surface represents the intersection of the prolongation of the fault with the Earth's surface. A=Austria, I=Italy, SLO=Slovenia. The gray surface on the bottom is a smoothed approximation of the Moho, to emphasize that the rectangular faults are within the crust. (b) Relation of vertical deformation (in mm/yr) to GTOPO30 topography. Also shown are the Individual Seismic Sources (black rectangles) and the Composite Seismic Sources (light blue stripes) from the DISS 3.1.1 Catalogue of Basili et al. (2008).

and width W of the rectangle and the two dimensional slip vector (d1, d2,0) of the hanging wall relative to the foot wall. We assume d3 = 0. The ratio of the P and S velocities of seismic waves was fixed to Vp/Vs = 1.89, a generally accepted value.

For our numerical modeling work we used the software Coulomb 3.20 developed at the U.S. Geological Survey (Lin and Stein, 2004; Toda et al., 2005), complemented with a module for numerical computation of partial derivatives, setup of the normal equations and

144

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

Fig. 6. Geometry of the slip u on a rectangular fault with a dip δ and production of a heat flow q by shear heating under the action of a shear stress τ associated to a deviatoric stress (large horizontal arrows, for a thrust fault). The horizontal line indicates the Earth's surface.

parameter adjustment by weighted least squares. Formal uncertainties in the estimated parameters were eventually computed from the diagonal elements of the variance covariance matrix for individual fault elements. To setup the fault structure which is needed to fit the data we proceeded in a two-step approach. In a first step we considered the three most significant domains (Giudicarie, NAWC, Pustertal) and fit them to the GPS data. These three domains are considered as more significant in the sense that they alone account for most of the GPS signal. Nevertheless some systematic residuals between measured and computed velocities suggested the introduction of three additional structures: the Brenner normal fault, a north-verging thrust on the northern termination of the Tauern Window and the Dinarides. The result is shown in Fig. 4a and demonstrates that these six structures are sufficient to account for most of the observed velocities. A number of GPS sites remain unaccounted by the model and are likely to respond to structures too local to be described with large scale rectangular faults. An interesting example is possibly given by VILH and LAN2, which could be on different sides of the Mölltal fault. It should be also remarked that the Tauern area is so densely faulted that it is not unlikely that several different faults determine the motion of each of these more dubious stations (example: PRBZ). The r.m.s. of the velocities before the fit (i.e. no model) of 1.3 mm/yr dropped to 0.8 mm/yr after the 6-element fit, roughly a 40% improvement. We present in the Discussion section an analysis of the effect of adding more faults to the model. Fig. 4b shows the surface velocities computed with the model and interpolated to a regular grid, to highlight the predicted regional scale surface kinematics. Table 2 gives the best fitting parameters for each of the six rectangular faults. Formal uncertainties in the parameters result from the diagonal components of their variance/covariance matrix scaled by the r.m.s. of the post-fit residuals. It is certainly a first approximation to consider that the slip is constant across each rectangular fault and ends abruptly at its edge. In principle it would be possible to divide each rectangular fault into sub-patches and assign a slip vector to each sub-patch. This would imply an increase in the number of unknowns to levels incompatible

with the amount of observational data available to constrain such model parameters. The model parameters are to some extent correlated with each other. For example decreasing the slip rate and increasing the slip area (or vice versa) can compensate each other, in the sense that the velocity pattern on the surface may not change. Likewise the dip angle and the absolute value of the slip vector are non-negligibly correlated. The knowledge that the upper brittle crust extends up to 10 km depth in most of the study area (Viganò et al., 2011), as demonstrated by heat/temperature data and hypocentral depths, was used to place all slip planes except those of larger dip angle (Giudicarie, Pustertal– Gailtal). This depth of the brittle crust is consistent with strength calculations along the TRANSALP section for weak rheologies. For strong crustal rheologies the brittle–ductile transition might be somewhat deeper, up to 12–14 km (Willingshofer and Cloetingh, 2003). The latter faults need to be at greater depth to fit the large scale velocity pattern. Consequently, the upper/lower parts of these faults lie in the brittle and respectively ductile part of the crust. For the dip angles and vergence of the faults, geological information (CNR, 1990) was used as a priori. The last column in Table 2 contains as a derived information the ‘potency rate’, that is the product of the slip area times the slip rate. We scale it by the shear modulus μ = 30 GPa to obtain a moment rate. This quantity and its lateral variation are likely to correlate with the local seismicity (regional ‘a’ and ‘b’ parameters of the Gutenberg–Richter law), with the maximum expected magnitude and with the geodetic strain rate, as suggested by recent studies (Caporali et al., 2011). Fig. 5 gives a 3D view of the fault planes. Likewise Fig. 5b shows the relationship to the topography of the projected fault planes and of the Individual Seismic Sources compiled by Basili et al. (2008). These are available as part of the DISS 3 of the Istituto Nazionale di Geofisica e Vulcanologia (http://diss.rm.ingv.it/diss/index.html). The model predicts uplift and subsidence in different areas at rates of up to few mm/yr which are in principle well measurable by precision leveling. The area of most intense uplift is the northern part of the

Table 3 For each structural unit, the heat flow on the slip surface due to shear is computed, for an assumed shear stress of 100 MPa. In the following columns this heat flow is mapped with Eq. (3) to the Earth's surface at four epochs counted since the start of the shear heating. In the last two columns the temperature increase at the slip surface as a consequence of the frictional heating is estimated at two epochs, since the beginning of the slipping phase, using Eq. (4). We take k = 3 W/(m°K) and κ = 1 mm2/s. The slip rate, dip angle and mean depth yb follow from Table 2. Fault id.

1 2 3 4 5 6

Name

Giudicarie NAWC Pustertal TW north Brenner fault Dinarides

Heat flow at depth (mW/m2)

100 62 41 48 63 45

Heat flow at the Earth's surface T0 My after slip initiation (mW/m2)

Temperature increase on slip plane T0 My after slip initiation (°C)

T0 = 5

T0 = 10

T0 = 15

T0 = 20

T0 = 5

T0 = 10

T0 = 15

T0 = 20

9 32 11 31 57 31

23 40 17 36 59 35

32 44 21 38 60 37

39 46 23 40 60 38

474 294 192 228 300 212

670 416 271 322 424 300

821 510 332 395 519 367

947 589 384 456 599 424

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

145

2000

150

100 MPa 200 MPa 300 MPa

100 MPa 200 MPa 300 MPa

Temperature increase on the fault plane (K)

Heat flow at the Earth surface (mW/m2)

1800

100

50

δ = 80°

1600

1400

1200

1000

800

600

δ = 45° 400

Pliocene-Quaternary →

Pliocene-Quaternary → ← Oligocene

Miocene

← Oligocene

Miocene

200

0

0

5

10

15

20

25

30

Time of initiation of frictional heating (Myr)

0

0

5

10

15

20

25

30

Time of initiation of frictional heating (Myr)

Fig. 7. (Left) Heat flux on the Earth's surface due to shear heating on a rectangular fault at depth. The continuous curves correspond to a dip of 80°, and the dotted curves to a dip of 45°. Different colors are assigned to different shear stresses. (Right) Temperature increase on the fault surface, due to shear heating, for various values of the shear stress.

Tauern Window (ca 6 mm/yr). However, as it will be discussed later, the measured height changes are the sum of the elastic and several other causes, and their interpretation is far from obvious.

4. Shear heating and heat flow The problem of advective and conductive transport of heat in a region of thrust faulting has been discussed by Molnar and England (1990). They demonstrated that a simplified analytical approach in two dimensions was sufficiently accurate, provided that the involved velocities were larger than few mm/yr, so that lateral conduction of heat becomes negligible. The role of lateral heat transport by fluids and its potential for masking frictional heating in slip zones has been addressed by Fulton et al. (2010). They used a two dimensional coupled fluid flow and heat transport model to show that advection after a large earthquake is in general significant only for large permeabilities. In such case, shear heating associated to the slip along the infracrustal subduction planes is expected to increase the temperature on each plane. Temperatures could become sufficiently high to induce weakening of the rocks and hence facilitate the slip, perhaps lowering the seismicity at moderate depths. Such heat, once mapped to the topographic surface, will add to the radiogenic continental heat flow, provided that lateral transport e.g. by ground water circulation, is negligible. This additional heat flow depends on the shear stress, dip angle and thermal conductivity of the rocks. To generate a thermal anomaly on the Earth's surface it is necessary that sufficient time has elapsed since the beginning of the shear heating at depth, so

that the heat has time to reach the surface of the Earth and be measured as a heat anomaly. In our study area the thermal anomaly data of Della Vedova et al. (2001) and Viganò et al. (2011) in the southernmost part are typically in the range 50–60 mW/m2. Values as high as 80 mW/m2 are reported in the Tauern Window area (Clark, 1961; Sachsenhofer, 2001). In a hypothetical scenario where the collision of the Adria with the European foreland took place in Late Oligocene–Miocene (some 25–20 Myr ago) and the slipping phase was more recent, this ‘no anomaly’ situation can help in setting an upper limit to the time of initiation of the slip along the fault planes. We can consider the high temperature on the plane and the upper limit on the heat flow at the Earth's surface as boundary conditions in a problem of heating a semi-infinite half space by a constant heat flux on the slip surface (Turcotte and Schubert, 2002). In such case the time-dependent temperature profile T(z,t) and the heat flux q(z,t) at a distance z from the slip surface and at a time t since the beginning of the slip phase (i.e. the onset of the heat flow) are given by the equations: 2 3 rffiffiffiffiffi z2   2q0 6 κt − z z 7 T ðz; t Þ−T 0 ¼ e 4κt − erfc pffiffiffiffiffi 5 4 π 2 k 2 κt   z qðz; t Þ ¼ q0 erf c pffiffiffiffiffi : 2 κt

ð1Þ

T0 being the reference temperature, that is the temperature on the slip plane prior to initiation of the slip phase. Hence T0 is controlled by the

146

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

local geothermal gradient at the depth of the plane. The complementary error function is defined as: η

2 −η′2 erfcðηÞ ¼ 1− pffiffiffi ∫ e dη′: π0

a

ð2Þ

The rock properties are defined by the coefficient of thermal conductivity k and the coefficient of diffusion κ. The heat flow associated to the friction is q0 = u ∗ τ, where u is the slip rate and τ is the shear stress. If δ is the dip angle, the vertical distance from the slip plane to the Earth's surface is y = z ∗ sinδ (Fig. 6).

c

14˚ TW

14˚ TW

TAMS

TAMS

KOLM

KOLM

TREI

TREI

MF

MF

PG

PG

LAN2 VILH ZOUF

LAN2 VILH

KLA2 ZOUF

ACOM

KLA2

ACOM

TARV

TARV

RADO

RADO

MPRA

MPRA

Fr

BARC

Fr

BARC

JOAN

46˚

46˚ CODR

JOAN

46˚

46˚ CODR

GORI

GORI

MDEA

MDEA

PALM

PALM

PAZO

PAZO

TRIE

TRIE

BEVA

BEVA

DI

SDNA

CAVA

CAVA

3 mm/yr

3 mm/yr

14˚

b

DI

SDNA

14˚

14˚ TW TAMS

KOLM

TREI

MF

PG

LAN2 VILH ZOUF

KLA2

ACOM TARV

RADO

MPRA

Fr

BARC

JOAN

46˚

46˚ CODR

GORI MDEA PALM

PAZO

TRIE BEVA

DI

SDNA

CAVA

3 mm/yr

14˚ Fig. 8. Effect of adding specific faults to the original six fault model in Fig. 4a and Table 2. Blue/white arrows represent the observed/computed velocities of the GPS stations. (a) Zoom of Fig. 4a in the indentation area, serving as reference. (b) Effect of adding a dextral strike-slip shear zone along the Mölltal fault (MF) of 2 cm/yr, dipping 80° and extending from 0 to 10 km depth. (c) Effect of adding a reverse fault in Friuli, dipping North at 45° and with an updip slip of 2 cm/yr.

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

The boundary condition on the heat flux at a mean vertical distance yb from the slipping area at depth and the Earth's surface at present time T0 is:    y  yb b pffiffiffiffiffiffiffi ¼ q ; t0 : τ  u  erfc 2 sinδ κt 0 sinδ

ð3Þ

The boundary condition on the temperature increase at the surface of the slip plane z = 0, relative to the nominal temperature T0 is: rffiffiffiffiffiffiffi τ  u κt 0 ¼ T ð0; t 0 Þ−T 0 : ð4Þ k π For plotting purposes we assume k = 3 W/(m K), κ = 1 mm 2/s, δ = 45 and 80°, τ = [100,200,300] MPa, u = 0.02 m/yr and a mean depth yb = 10 km, which are typical for our problem, according to Table 2. More specific values are given in Table 3. Fig. 7 shows the heat flow on the Earth's surface and the temperature on the fault plane for two dip angles and three choices of the shear stress. The figure suggests that the frictionally induced heat flow at the Earth's surface is smaller than, say, 20 mW/m 2 if the frictional heating initiated more recently than during the Late Oligocene–Miocene time of deformation in the Eastern and Southern Alps, that is roughly in Pliocene– Pleistocene times. In such case the shear heating on the fault plane was associated with a moderate shear stress of some 100 MPa and lasted enough for a temperature increase of 200 to 400 °C on the slip plane. Table 3 gives for each fault more specific values than Fig. 7, and shows how the temperature at depth and heat flow on the Earth's surface would increase after 5, 10, 15, and 20 My from the initiation of the slip phase. The values in Table 3 assume a shear stress of 100 MPa. According to the rheological profiles (Fig. 6 of Viganò et al. (2011)), the lower crust can be considered ductile for these values of the tectonic stress. The temperature increase induced by the shear stress on the fault plane adds to the T0 of a normal continental geotherm, assumed in the range 10–30 K/km, bringing the temperature on the deeper portion of the fault (in all the six cases b 35 km, see Table 2) to a total of 600–800 °C. Initiation of shear heating at earlier times would, according to the model, result in heat flux which, when added to the radiogenic component, would give a total heat flux larger than observed. Smaller values of the thermal conductivity k are possible. In such case Eq. (4) shows that the initiation of shear heating would be even at more recent times. As an example we assume that the data of Viganò et al. (2011) valid for the Po Plain are applicable to the Giudicarie (element #1). Their reported thermal flow at the Earth's surface of 46± 9 mW/m 2 would result of a radiogenic contribution of 37 mW/m 2 and a shear heating contribution of 9 mW/m 2, if the slip phase initiated 5 My ago (Table 3). An earlier initiation of the slip phase would imply a smaller radiogenic heat flow and a larger shear heating component. The estimates of heat flow at the Moho would, in such case, have to be revised to account for the contribution of shear heating. According to Table 2, the Giudicarie shear zone would lie at a depth between 26 and 34 km, and the corresponding temperatures between 600 and 780 °C respectively, for a gradient of 23± 5 °C/km. Taking the temperature at the center of the fault to be 690 °C, we conclude from Table 3 that 216 °C are of radiogenic origin and the remaining 474 °C are due to shear heating. An earlier initiation of the slip phase would, according to data of Table 3, imply temperatures considerably higher than those predicted by the measured thermal gradient. Hence the hypothesis that the slip phase must have initiated roughly a few Myrs ago. Similar reasoning can be done for the other areas, if the measurements of the geotherm and heat flow are available. 2

5. Discussion The indentation of Adria into the eastern Southern Alps, and related phenomena such as uplift and lateral extrusion, can be modeled in

147

terms of dislocation in an elastic half space. Slip on a finite number of rectangular faults maps into displacement on the free surface of the half space, which is compared with the horizontal velocities of GPS stations. The location, size and orientation of the rectangular faults are close to those of regional tectonic lineaments, which exhibit a variable degree of seismicity. The deformation pattern requires the Adria indenter to deform in the contact zone with large scale structures. Because of the slip on selected planes, the stress associated with the Nubia push is partly slowly dissipated e.g. by heat, partly accommodated by seismic events at or near some of the planes, partly stored in the rocks. The flow is controlled by the transpressional character of the Periadriatic (Pustertal–Gailtal) fault, Giudicarie fault, Dinarides and North Alpine Wrench Corridor. The normal character of the Brenner fault seems to fit well the velocities of STBZ (Vipiteno/Sterzing) and PATK (Patscherkofel). Our model accounts for a lateral escape uniquely as a consequence of convergence and indentation. We have not attempted to include gravitational collapse and spreading as a driver for extrusion (Ratschbacher et al., 1991a,b). More recent studies (Robl and Stüwe, 2005) suggest that gravity is unlikely to play a major role. Vertical GPS velocities were not used to constrain the model, as they are insufficiently well determined in several cases. Relative to previous approaches based mostly on 2D deformation, the 3D model presented here predicts vertical motion which is in qualitative agreement with independent data (Fig. 5). In the Tauern Window the uplift is measured to be of the order of 1 mm/yr (Höggerl, 2001). Our model predicts values as high as 4–6 mm/yr, with an uncertainty difficult to calculate, but which is likely to be of the order of few mm/yr. Hence the discrepancy is probably inconspicuous. We emphasize that the vertical motions predicted by an elastic model need to be complemented with estimates of erosion rates, post-glacial uplift and departures from elasticity, before they can be compared with data such as leveling, absolute gravity or GPS/DInSAR heighting. It is a remarkable feature of the model (Fig. 5) the prediction of an increased surface uplift e.g. in Julian Alps, where such is demonstrated by other geodetic leveling data (Rižnar et al., 2007). The thermal effects associated with the subduction are analyzed with a simple analytical model, which is based on an instantaneous heat flux in a half space. We speculate that if, on the one hand, the total heat at the surface is at nominal levels, as suggested by data, and, on the other hand, the temperature on the fault plane is high enough to facilitate the slip, then the slip phase must be relatively recent. If this is the case, then, in the absence of lateral dissipation of heat e.g. by ground water transport, the frictional heat has had insufficient time to reach the Earth's surface and generate a thermal anomaly. For example for dip angles of approximately 45° and shear stresses of ca. 100 MPa, a slip of ca. 2 cm/yr at a mean depth of some 10 km, a shear induced heat flow initiated in the Pleistocene reaches the Earth's surface still at a low enough level that it adds negligibly (Southern Alps) or marginally (Tauern) to the radiogenic heat flux. This inference maintains its validity for a smaller conductivity k and/or shear stress τ than assumed in the plots (Fig. 7). The high temperatures on the fault surface on the one hand and, on the other hand, the fact that earthquakes on the continental crust seem to be confined at depths where the temperatures are less than 350 °C (Chen and Molnar, 1983) can help in explaining why seismicity is higher in the Friuli Montello area than in the Tauern Window area to the North: the fault zone in Friuli has a lower slip and hence shear heating than the Tauern area. Hence the temperature of the Friuli fault must be lower than in the Tauern faults, implying that a brittle rheology in Friuli is more likely than in the Tauern. This hypothesis would be consistent with the different heat fluxes in the two areas. The relatively high seismicity on the southern end of the Giudicarie (fault #1) could be related to a high strain regime on a relatively cold edge of the rectangular fault (Fig. 1). For the Pliocene, there is indirect evidence for an increase of the intra-crustal temperature from both subsidence analysis and calculations

148

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

of the effective elastic thickness of flexural basins. Such data were recorded from the North Alpine Molasse basin (north of the North Alpine fault, Fig. 2; Genser et al., 2007), the intra-orogenic Klagenfurt basin north of the eastern part of the Periadriatic fault (Nemes et al., 1997) and the westernmost Styrian basin (Sachsenhofer et al., 1997). Sachsenhofer et al. (2001) report high Miocene to Recent heat flow for the Alpine– Pannonian–Dinaric junction. Over large portions of Eastern Alps, the present-day heat flow is in the order of 55–75 mW/m2 (Sachsenhofer, 2001). Interestingly, the heat flow is still elevated to 80–90 mW/m2 in major portions of the Tauern Window (Schubert, 2003; Zötl and Goldbrunner, 1993). For the Tauern Window area, a Late Miocene–Pliocene thermal dome was postulated (Genser et al., 1996; Luth and Willingshofer, 2008). In general, thermal springs along some major faults like SEMP fault along northern margins in Eastern Alps could represent an indication of heat dissipation by advection along such faults. Many major Late Oligocene–Miocene faults of Eastern Alps include fault rocks formed within a succession of initial ductile, then semiductile and finally brittle deformation (e.g., Handy et al., 2005; Rosenberg and Berger, 2009; Wang and Neubauer, 1998) during decreasing temperature conditions during exhumation. In cases where these faults reflect higher temperatures as adjacent units as along the Periadriatic fault, these faults could be interpreted, consequently, as zones of subvertical heat dissipation (see also Handy et al., 2005) similar as postulated for the Pliocene–Recent tectonic setting in our study. These weak zones of viscous rheology are between zones of brittle rheology. To explain the Oligocene–Miocene indentation tectonics, Robl and Stüwe (2005) used a Newtonian rheology and found that a three times higher viscosity for the Adriatic indenter than for Eastern Alps is sufficient to explain the observed structure. Our model favors a change of the structural setting of the Eastern Alps not before ca. 5 Myr (onset of Pliocene). This is a matter of debate. For example, Pischinger et al. (2008) argue for an earlier onset of E–W compression within eastern sectors of Eastern Alps). Our proposal is in convincing agreement with structural data, which indicate inversion of the entire system since the onset of Pliocene. The evidence includes: (1) sharp onset of the subsidence in the Venetian platform (Friuli area) and northernmost Adria since ca. 3 Myr (Mancin et al., 2009); (2) in contrast to subsidence in Friuli, sudden increase of surface uplift in Eastern Alps (Genser et al., 2007; Hergarten et al., 2010; Sachsenhofer et al., 1997; Wagner et al., 2010), (3) a revival of ca. N–S indentation tectonics (Peresson and Decker, 1997; Bada et al., 2001), and (4) a change to an overall compressive regime with no significant lateral extrusion (Bada et al., 2001). This tectonic pattern is different from the Oligocene– Miocene collision pattern and is characterized by stronger coupling between the orogen and its forelands. E.g., even the northern foreland of Alps is affected by surface uplift (Genser et al., 2007). The northwestern Dinarides are a zone of Pliocene to Recent dextral motion, along the Idrija fault respectively dextral wrenching along the interface between Adriatic Sea and Dinaric Mountains (e.g., Picha, 2002). Our analysis has focused on six faults; these are the principal responsible faults of the observed pattern of surface velocities. In the following we provide arguments suggesting that additional faults would bring no benefit, in the sense that the r.m.s. of the fit would be constant or larger than that corresponding to the chosen six faults. We selected two additional faults in the indentation area: one along the North verging Friuli faults, stretching roughly E–W, and the other along the Mölltal fault, a dextral strike slip fault striking roughly SE– NW. We summarize our results in Fig. 8: Fig. 8a is a zoom of Fig. 4a in the selected area, and serves as reference, that is no further fault modeled beyond the six faults in Table 2; Fig. 8b shows the effect of introducing a dextral strike-slip shear zone roughly along the Mölltal fault, dipping NE at 80°, with 2 cm/yr right lateral slip and a depth ranging between 0 and 10 km. Stations such as LAN2, VILH, ACOM, TARV, RADO and KLA2 are clearly affected by the existence of this structure, and should exhibit a velocity which is less in agreement with the observed velocities than in the reference case described by Fig. 8a.

Likewise for Fig. 8c, where a Friuli thrust dipping at 45° with a 2 cm/yr up-dip slip and with depth between 5 and 15 km, centered on the mean hypocentral depth of 8 km is postulated. To support this hypothesis, the two stations MPRA and JOAN must have a North velocity higher than observed, suggesting that the hypothesis is unlikely. 6. Conclusion The model presented here describes in an area of active indentation the relationship in 3D between fault structure, seismology, heat flow and surface kinematics. Unlike previous studies based on rigid indenters with Mohr–Coulomb rheology for the upper crust and viscous rheology for the lower crust (Ratschbacher et al., 1991a), we concentrate on a purely elastic rheology in three dimensions for both the indenter and the indented block, with slip confined in selected rectangular faults. The rectangular faults which can be constrained by the available data resemble known structures. The model is able to discriminate, compatibly with the available GPS data, which structure is undergoing slip, so that further research can be stimulated on the relation between seismicity, slip at depth and surface velocity field. The structure with the largest slip at depth is the NW part of the indenter (Giudicarie fault). The Pustertal–Gailtal fault and the North Alpine Wrench Corridor play a major role in accommodating the indentation. Two other faults, the Brenner fault and the Salzachtal–Ennstal– Mariazell–Puchberg fault at the northern edge of the Tauern dome enable smaller scale structures to be accounted for. Other candidates (e.g. Southern side of the Tauern dome) could not be tested due to insufficient coverage of GPS data. The Schio–Vicenza fault is reasonably well covered and does not seem to imply a velocity gradient. The characterization of the active faults in Table 2 reflects both active tectonics and the geographic gaps in the data set. There are certainly additional deformation areas, but we do not yet have sufficient data to constrain their model. It is only a matter of time to accumulate sufficiently long time series from existing and planned stations to fill the gaps in the velocity map. New preliminary data presented recently by the Slovenian group (Medved et al., 2011) indicate that the velocity pattern in the northern Dinarides is very well in keeping with the pattern predicted by our model (Fig. 4a and b). DInSAR data are also expected to play a role, particularly for the vertical motion. A final important remark is on the concept of convergence rate. For a rigid indenter the convergence rate near the margin is independent of depth. For a deformable margin, this is not the case, and larger velocities at depth are expected. This gradient of velocity with depth is likely to have an impact on geodynamic models of the indentation and on the time scale of the collisional process. We have used heat flow data in conjunction with an analytical model of shear heating of a half space to set up constraints on the time of initiation of the slip phase. The fact that no thermal anomaly is reported in correspondence to the rectangular faults can be interpreted as caused by the shear induced heat flow having insufficient time to reach the Earth's surface. However, indirect structural evidence indicates increasing temperatures at depth. We conclude that a slip phase initiated in the Pliocene would be consistent with the data and would be more recent than the time of initiation of the collision of the Adria microplate with Eurasia (Oligocene/Miocene). This is well supported by structural evidence, which indicates a second phase of indentation since ca. the onset of Pliocene (ca. 5 Myr ago). Changes in heat flux could be correlated with different seismicity by virtue of the colder temperature in the slipping planes associated to seismic provinces and a higher temperature on slipping planes of less seismic provinces. Acknowledgments We acknowledge detailed reviews by two anonymous journal reviewers who helped to clarify statements and presentations. We thank G. Ranalli and A. Viganò for helpful discussions. The research of AC and LO was in part supported by the University of Padova

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

under the Grant ‘Space techniques to monitor the physics of seismicity in NorthEast Italy’. Work of FN has been partially supported by the Austrian Science Fund (FWF) grant no. P22,110.

References Altamimi, Z., Collilieux, X., Legrand, J., Garayt, B., Boucher, C., 2007. ITRF2005: a new release of the International Terrestrial Reference Frame based on time series of station positions and Earth Orientation Parameters. Journal of Geophysical Research 112, B09401. http://dx.doi.org/10.1029/2007JB004949. Anderson, H., Jackson, J., 1987. Active tectonics of the Adriatic region. Geophysical Journal of the Royal Astronomical Society 91, 937–983. Bada, G., Horvath, F., Cloetingh, S., Coblentz, D.D., Toth, T., 2001. Role of topography-induced gravitational stresses in basin inversion: the case study of Pannonian Basin. Tectonics 20, 343–363. Barletta, V.R., Ferrari, C., Diolaiuti, G., Carnielli, T., Sabadini, R., Smiraglia, C., 2006. Glacier shrinkage and modeled uplift of the Alps. Geophysical Research Letters 33, L14307. http://dx.doi.org/10.1029/2006GL026490. Basili, R., Valensise, G., Vannoli, P., Burrato, P., Fracassi, U., Mariano, S., Tiberti, M.M., Boschi, E., 2008. The Database of Individual Seismogenic Sources (DISS), version 3: summarizing 20 years of research on Italy's earthquake geology. Tectonophysics 453, 20–43. http://dx.doi.org/10.1016/j.tecto.2007.04.014. Battaglia, M., Murray, M.H., Serpelloni, E., Burgmann, R., 2004. The Adriatic region: an independent microplate within the Africa Eurasia collision zone. Geophysical Research Letters 1, L09605. http://dx.doi.org/10.1029/2004GL019723. Bechtold, M., Battaglia, M., Tanner, D.C., Zuliani, D., 2009. Constraints on the active tectonics of the Friuli/NW Slovenia area from CGPS measurements and threedimensional kinematic modeling. Journal of Geophysical Research 114, B03408. http://dx.doi.org/10.1029/2008JB005638. Benedetti, L., Tapponnier, P., King, G.C.P., Meyer, B., Manighetti, I., 2000. Growth folding and active thrusting in the Montello region, Veneto, northern Italy. Journal of Geophysical Research 105 (B1), 739–766. Bressan, G., Snidarcig, A., Venturini, C., 1998. Present state of tectonic stress of the Friuli Area (Eastern Southern Alps). Tectonophysics 292, 211–227. Bressan, G., Kravanja, S., Franceschina, G., 2007. Source parameters and stress release of seismic sequences occurred in the Friuli–Venezia Giulia region (NorthEastern Italy) and in Western Slovenia. Physics of the Earth and Planetary Interiors 160, 192–214. Brückl, E., Bleibinhaus, F., Gosar, A., Grad, M., Guterch, A., Hrubcovà, P., Randy Keller, G., Majdanski, M., Sumanovac, M., Tiira, T., Yliniemi, J., Hegedus, E., Thybo, H., 2007. Crustal structure due to collisional and escape tectonics in the Eastern Alps region based on profiles Alp01 and Alp02 from the ALP 2002 seismic experiment. Journal of Geophysical Research 112, B06308. http://dx.doi.org/10.1029/2006JB004687. Brückl, E., Behm, M., Decker, K., Grad, M., Guterch, A., Keller, G.R., Thybo, H., 2010. Crustal structure and active tectonics in the Eastern Alps. Tectonics 29, TC2011. http://dx.doi.org/10.1029/2009TC002491. Caporali, A., 2003. Average strain rate in the Italian crust inferred from a permanent GPS network — I. Statistical analysis of time series of permanent GPS stations. Geophysical Journal International 155, 241–253. Caporali, A., Braitenberg, C., Massironi, M., 2005. Geodetic and hydrological aspects of the Merano earthquake of 17 July 2001. Journal of Geodynamics 39, 317–336. http://dx.doi.org/10.1016/j.jog.2005.01.001. Caporali, A., Aichhorn, C., Barlik, M., Becker, M., Fejes, I., Gerhatova, L., Ghitau, D., Grenerczy, G., Hefty, J., Krauss, S., Medak, D., Milev, G., Mojzes, M., Mulic, M., Nardo, A., Pesec, P., Rus, T., Simek, J., Sledzinski, J., Solaric, M., Stangl, G., Stopar, B., Vespe, F., Virag, G., 2009. Surface kinematics in the Alpine–Carpathian–Dinaric and Balkan region inferred from a new multi-network GPS combination solution. Tectonophysics 474, 295–321. http:// dx.doi.org/10.1016/j.tecto.2009.04.035. Caporali, A., Barba, S., Carafa, M.M., Devoti, R., Pietrantonio, G., Riguzzi, F., 2011. Static stress drop as determined from geodetic strain rates and statistical seismicity. Journal of Geophysical Research 116, B02410. http://dx.doi.org/10.1029/2010JB007671. Caputo, R., Poli, M.E., Zanferrari, A., 2010. Neogene–Quaternary tectonic stratigraphy of the eastern Southern Alps, NE Italy. Journal of Structural Geology 32, 1009–1027. Carminati, E., Cavazza, D., Scrocca, D., Fantoni, R., Doglioni, C., 2010. Thermal and tectonic evolution of the Southern Alps (northern Italy) rifting: coupled organic matter maturity analysis and thermokinematic modeling. AAPG Bulletin 94 (3), 369–397. Carulli, G.B., Nicolich, R., Rebez, A., Slejko, D., 1990. Seismotectonics of the Northwest External Dinarides. Tectonophysics 179, 11–25. Cassinis, R., 2006. Reviewing pre-TRANSALP DSS models. Tectonophysics 414, 79–86. Castellarin, A., Nicolich, R., Fantoni, R., Cantelli, L., Sella, M., Selli, L., 2006. Structure of the lithosphere beneath the Eastern Alps (southern sector of the TRANSALP transect). Tectonophysics 414, 259–282. http://dx.doi.org/10.1016/j.tecto.2005.10.013. Cederbom, C.E., Sinclair, H.D., Schlunegger, F., Rahn, M.K., 2004. Climate-induced rebound and exhumation of the European Alps. Geology 32, 709–712. http://dx.doi.org/ 10.1130/G20491.1. Chen, W.P., Molnar, P., 1983. Focal depths of intracontinental and intraplate earthquakes and their implications for the thermal and mechanical properties of the lithosphere. Journal of Geophysical Research 88, 4183–4214. Clark, S., 1961. Heat flow in the Austrian Alps. Geophysical Journal of the Royal Astronomical Society 6, 54–63. CNR, 1990. Structural model of Italy. In: Barberi, F. (Ed.), SELCA Editions, Firenze 1990. D'Agostino, N., Cheloni, D., Mantenuto, S., Selvaggi, G., Michelini, A., Zuliani, D., 2005. Strain accumulation in the Southern Alps (NE Italy) and deformation at the

149

NorthEastern boundary of Adria observed by CGPS. Geophysical Research Letters 32, L19306. http://dx.doi.org/10.1029/2005GL024266. D'Agostino, N., Avallone, A., Cheloni, D., D'Anastasio, E., Mantenuto, S., Selvaggi, G., 2008. Active tectonics Adriatic region from GPS and earthquake slip vectors. Journal of Geophysical Research 113, B12413. http://dx.doi.org/10.1029/2008JB005860. Dach, R., Hugenthobler, U., Friedez, P., Meindl, M., 2007. Bernese GPS Software Version 5.0. Astronomical Institute, University of Bern (612 pp.). Della Vedova, B., Bellani, S., Pellis, G., Squarci, P., 2001. Deep temperatures and surface heat flow distribution. In: Vai, G.B., Martini, I.P. (Eds.), Anatomy of an Orogen, the Apennines and Adjacent Mediterranean Basins. Kluwer Ac. Publ., Great Britain, pp. 65–76. Fügenschuh, B., Seward, D., Mantckelow, N., 1997. Exhumation in a convergent orogen: the western Tauern Window. Terra Nova 9, 213–217. Fügenschuh, B., Mancktelow, N.S., Schmid, S.S., 2011. Comment on Rosenberg and Garcia: estimating displacement along the Brenner Fault and orogen-parallel extension in the Eastern Alps. Int J Earth Sci (Geol. Rundsch.) (2011) 100:1129–1145. International Journal of Earth Sciences 101, 1451–1455. http:// dx.doi.org/10.1007/s00531-011-0725-4. Fulton, P.M., Harris, R.N., Saffer, D.M., Brodsky, E.E., 2010. Does hydrologic circulation mask frictional heat on faults after large earthquakes? Journal of Geophysical Research 115, B09402. http://dx.doi.org/10.1029/JB007103. Galadini, F., Poli, M.E., Zanferrari, A., 2005. Seismogenic sources potentially responsible for earthquakes with M > 6 in the eastern Southern Alps (Thiene–Udine sector, NE Italy). Geophysical Journal International 161 (3), 739–762. http://dx.doi.org/ 10.1111/j.1365-246X.2005.02571.x. Genser, J., van Wees, J.D., Cloetingh, S., Neubauer, F., 1996. Eastern Alpine tectonometamorphic evolution: constraints from two-dimensional P–T–t modelling. Tectonics 15, 584–604. Genser, J., Cloetingh, S., Neubauer, F., 2007. Late orogenic rebound and oblique Alpine convergence: new constraints from subsidence analysis of the Austrian Molasse basin. Global and Planetary Change 58, 214–223. Grenerczy, G., Kenyeres, A., 2006. Crustal deformation between Adria and the European platform from space geodesy. In: Pinter, N., et al. (Ed.), The Adria microplate: GPS geodesy, tectonics and hazards. : NATO Sci. Ser. IV, Vol. 61. Springer, Dordrecht, Netherlands, pp. 321–334. http://dx.doi.org/10.1007/1-4020-4235-3_22. Gruppo di lavoro CPTI04, 2004. Catalogo paramentrico dei Terremoti Italiani, versione 2004 (CPTI04), INGV. In: Gasperini, P., Camassi, R., Mirto, C., Stucchi, M. (Eds.), National Research Council of Italy, Bologna ((http://emidius.mi.ingv.it/CPTI04/). Handy, M.R., Babist, J., Wagner, R., Rosenberg, C., Konrad, M., 2005. Decoupling and its relation to strain partitioning in continental lithosphere: insight from the Periadriatic fault system (European Alps). In: Gapais, D., Brun, J.P., Cobbold, P.R. (Eds.), Deformation mechanisms, rheology and tectonics: from minerals to the lithosphere: Geol. Soc. [London] Spec. Publ., 243, pp. 249–276. http://dx.doi.org/10.1144/GSL.SP.2005.243.01.17. Handy, M.R., Schmid, S.M., Bousquet, R., Kissling, E., Bernoulli, D., 2010. Reconciling platetectonic reconstructions of Alpine Tethys with the geological–geophysical record of spreading and subduction in the Alps. Earth-Science Reviews 102, 121–158. Hergarten, S., Wagner, T., Stüwe, K., 2010. Age and prematurity of the Alps derived from topography. Earth and Planetary Science Letters 297, 453–460. Höggerl, N., 2001. Bestimmung von rezenten Höhenänderungen durch wiederholte geodätische Messungen. In: Hammerl, C., et al. (Ed.), Die Zentralanstalt für Meteorologie und Geodynamik 1851–2001. Leykam, Graz, Austria, pp. 630–644. Keil, M., Neubauer, F., 2011a. Neotectonics, drainage pattern and geomorphology of the orogen-parallel Upper Enns Valley, Eastern Alps. Geologica Carpathica 62, 279–295. Keil, M., Neubauer, F., 2011b. The Miocene Enns Valley basin (Austria) and the North Enns Valley fault. Austrian Journal of Earth Sciences 104, 49–65. Lenhardt, W.A., Freudenthaler, C., Lippitsch, R., Fiegweil, E., 2007. Focal-depth distribution in the Austrian Eastern Alps based on macroseismic data. Austrian Journal of Earth Sciences 100, 66–79. Lin, J., Stein, R.S., 2004. Stress triggering in thrust and subduction earthquakes, and stress interaction between the southern San Andreas and nearby thrust and strike-slip faults. Journal of Geophysical Research 109, B02303. http://dx.doi.org/ 10.1029/2003JB002607. Linzer, H.-G., Decker, K., Peresson, H., Dell'Mour, R., Frisch, W., 2002. Balancing lateral orogenic float of the Eastern Alps. Tectonophysics 354 (3–4), 211–237. http:// dx.doi.org/10.1016/S0040-1951(02)00337-2. Luth, S.W., Willingshofer, E., 2008. Mapping of the post-collisional cooling history of the Eastern Alps. Swiss Journal of Geosciences 101, 207–223. Mancin, N., Di Giulio, A., Cobianchi, M., 2009. Tectonic vs. climate forcing in the Cenozoic sedimentary evolution of a foreland basin (Eastern Southalpine system, Italy). Basin Research 21, 799–823. http://dx.doi.org/10.1111/j.1365-2117.2009.00402.x. Marton, E., Zampieri, D., Kazmer, M., Dunkl, I., Frisch, W., 2011. New Paleocene–Eocene paleomagnetic results from the foreland of the Southern Alps confirm decoupling of stable Adria from the African plate. Tectonophysics 504, 89–99. http:// dx.doi.org/10.1016/j.tecto.2011.03.006. Massironi, M., Zampieri, D., Caporali, A., 2006. Miocene to present major fault linkages through the Adriatic indenter and the Austroalpine–Penninic collisional wedge (Alps of NE Italy). In: Moratti, G., Chalouan, A. (Eds.), Tectonics of the Western Mediterranean and North Africa: Geological Society, [London] Special Publications, 262, pp. 245–258. Medved, K., Stopar, B., Bajec, K., 2011. National report of Slovenia. Paper Presented at the 2011 EUREF Symposium, Moldavia (Available on line at http://www.euref. eu/euref_symposia_meetings.html). Molnar, P., England, P., 1990. Temperatures, heat flux, and frictional stress near major thrust faults. Journal of Geophysical Research 95B4, 4833–4856. Nemes, F., Neubauer, F., Cloetingh, S., Genser, J., 1997. The Klagenfurt Basin in the Eastern Alps: a decoupled intra-orogenic flexural basin? Tectonophysics 282, 189–204. Neubauer, F., Fritz, H., Genser, J., Kurz, W., Nemes, F., Wallbrecher, E., Wang, X., Willingshofer, E., 2000. Structural evolution within an extruding wedge: model and application to the

150

A. Caporali et al. / Tectonophysics 590 (2013) 136–150

Alpine–Pannonian system. In: Urai, J., Lehner, F., van Zee, W. (Eds.), Aspects of Tectonic Faulting (Festschrift in Honour of Georg Mandl). Springer, pp. 141–153. Okada, Y., 1985. Surface deformation due to shear and tensile faults in a half-space. Bulletin of the Seismological Society of America 75, 1135–1154. Peresson, H., Decker, K., 1997. Far-field effects of Late Miocene subduction in the Eastern Carpathians: E–W compression and inversion of structures in the Alpine– Carpathian–Pannonian region. Tectonics 16, 38–56. Picha, F.J., 2002. Late orogenic strike-slip faulting and escape tectonics in frontal Dinarides–Hellenides, Croatia, Yugoslavia, Albania and Greece. The American Association of Petroleum Geologists Bulletin 86, 1659–1671. Pischinger, G., Kurz, W., Übleis, M., Egger, M., Fritz, H., Brosch, F.J., Stingl, K., 2008. Fault slip analysis in the Koralm Massif (Eastern Alps) and consequences for the final uplift of “cold spots” in Miocene times. Swiss Journal of Geosciences 101, 235–254. Plan, L., Grasemann, B., Spötl, C., Decker, K., Boch, R., Kramers, J., 2010. Neotectonic extrusion of the Eastern Alps: constraints from U/Th dating of tectonically damaged speleothems. Geology 38, 483–486. http://dx.doi.org/10.1130/G30854.1. Pondrelli, S., Salimbeni, S., Morelli, A., Ekström, G., Postpischl, L., Vannucci, G., Boschi, E., 2011. European–Mediterranean regional centroid moment tensor catalog: solutions for 2005–2008. Physics of the Earth and Planetary Interiors 185, 74–81. http://dx.doi.org/10.1016/j.pepi.2011.01.007. Ratschbacher, L., Merle, O., Davy, P., Cobbold, P., 1991a. Lateral extrusion in the Eastern Alps, part 1: boundary conditions and experiments scaled for gravity. Tectonics 10, 2. http://dx.doi.org/10.1029/90TC02622. Ratschbacher, L., Frisch, W., Linzer, H.-G., Merle, O., 1991b. Lateral extrusion in the Eastern Alps, part 2. Structural analysis. Tectonics 10, 257–271. Regenauer-Lieb, K., Petit, J.P., 1997. Cutting of the European continental lithosphere; plasticity theory applied to the present Alpine collision. Journal of Geophysical Research 102, 7731–7746. Reinecker, J., Lenhardt, W.A., 1999. Present-day stress field and deformation in eastern Austria. International Journal of Earth Sciences 88, 532–550. Reiter, F., Lenhardt, W.A., Brandner, R., 2004. Indications for activity of the Brenner Normal Fault zone (Tyrol, Austria) from seismological and GPS data. Austrian Journal of Earth Sciences 97, 16–24. Rižnar, I., Koler, B., Bavec, M., 2007. Recent activity of the regional geologic structures in western Slovenia. Geologija 50, 111–120. Robl, J., Stüwe, K., 2005. Continental collision with finite indenter strength: 2 European Eastern Alps. Tectonics 24, TC4014. http://dx.doi.org/10.1029/2004TC001741. Rosenberg, C.L., Berger, A., 2009. On the causes and modes of exhumation and lateral growth of the Alps. Tectonics 28, TC6001. http://dx.doi.org/10.1029/2008TC002442. Rosenberg, C.L., Garcia, S., 2011. Estimating displacement along the Brenner Fault and orogen-parallel extension in the Eastern Alps. International Journal of Earth Sciences 100, 1129–1145. http://dx.doi.org/10.1007/s00531-011-0645-3. Rosenberg, C.L., Brun, J.-P., Cagnard, F., Gapais, D., 2007. Oblique indentation in the Eastern Alps: insights from laboratory experiments. Tectonics 26, TC2003. http:// dx.doi.org/10.1029/2006TC001960. Sachsenhofer, R.F., 2001. Syn- and post-collisional heat flow in the Cenozoic Eastern Alps. International Journal of Earth Sciences 90, 579–592. Sachsenhofer, R.F., Lankreijer, A., Cloetingh, S., Ebner, F., 1997. Subsidence analysis and quantitative basin modelling in the Styrian Basin (Pannonian Basin System, Austria). Tectonophysics 272, 175–196. Sachsenhofer, R.F., Jelen, B., Hasenhüttl, C., Dunkl, I., Rainer, T., 2001. Thermal history of Tertiary basins in Slovenia (Alpine–Dinaride–Pannonian junction). Tectonophysics 334, 77–99.

Savage, J.C., 1983. A dislocation model of strain accumulation and release at a subduction zone. Journal of Geophysical Research 88B6, 4984–4996. Schmid, S.M., Fügenschuh, B., Kissling, E., Schuster, R., 2004. Tectonic map and overall architecture of the Alpine orogen. Eclogae Geologicae Helvetiae 97, 93–117. Schmid, S.M., Bernoulli, D., Fuegenscuh, B., Matenco, L., Schefer, S., Schuster, R., Tischler, M., Ustaszewski, K., 2008. The Alpine–Carpathian–Dinaric orogenic system: correlation and evolution of tectonic units. Swiss Journal of Geosciences 101, 139–183. Schubert, G., 2003. Hydrogeological map of Austria 1:500.000. Geologische Bundesanstalt, Vienna. T.R.A.N.S.A.L.P. Working Group, 2002. First deep seismic reflection images of the Eastern Alps reveal giant crustal wedges and transcrustal ramps. Geophysical Research Letters 29 (10), 1452. http://dx.doi.org/10.1029/2002GL014911. Toda, S., Stein, R.S., Richards-Dinger, K., Bozkurt, S., 2005. Forecasting the evolution of seismicity in southern California: animations built on earthquake stress transfer. Journal of Geophysical Research B05S16. http://dx.doi.org/10.1029/2004JB003415. Turcotte, D.L., Schubert, G., 2002. Geodynamics, second edition. Cambridge University Press, Cambridge, p. 472. Viganò, A., Bressan, G., Ranalli, G., Martin, S., 2008. Focal mechanism inversion in the Giudicarie Lessini seismotectonic region (Southern Alps, Italy): insights on tectonic stress and strain. Tectonophysics 460, 106–115. Viganò, A., Della Vedova, B., Ranalli, G., Martin, S., Scafidi, D., 2011. Geothermal and rheological regime in the Po Plain sector of Adria (Northern Italy). Italian Journal of Geosciences 131 (2), 228–240. http://dx.doi.org/10.3301/IJG.2012.09. Vosteen, H.D., Rath, V., Clauser, C., Lammerer, B., 2006. A review of the thermal regime of the Eastern Alps with respect tothe effects of paleoclimate and exhumation. Tectonophysics 414, 157–167. http://dx.doi.org/10.1016/j.tecto.2005.10.022. Wagner, T., Fabel, D., Fiebig, M., Häuselmann, P., Sahy, D., Xu, S., Stüwe, K., 2010. Young uplift in the non-glaciated parts of the Eastern Alps. Earth and Planetary Science Letters 295, 159–169. Wang, X., Neubauer, F., 1998. Orogen-parallel strike-slip faults bordering metamorphic core complexes: the Salzach–Enns fault zone in the Eastern Alps. Journal of Structural Geology 20, 799–818. Weber, J., Vrabec, M., Stopar, B., Pavlovcic‐Preseren, P., Dixon, T., 2006. The PIVO‐2003 experiment: a GPS study of Istria Peninsula and Adria Microplate motion, and active tectonics in Slovenia. In: Pinter, N., Grenerczy, G., Weber, J., Stein, S., Medak, D. (Eds.), The Adria microplate: GPS geodesy, tectonics and hazards. : NATO Sci. Ser. IV, Vol. 61. Springer, New York, pp. 321–334. Willett, S.D., Schlunegger, F., Picotti, V., 2006. Messinian climate change and erosional destruction of the central European Alps. Geology 34, 613–616. http://dx.doi.org/ 10.1130/G22280.1. Willingshofer, E., Cloetingh, S., 2003. Present‐day lithospheric strength of the Eastern Alps and its relationship to neotectonics. Tectonics 22, 1075. http://dx.doi.org/ 10.1029/2002TC001463. Wölfler, A., Kurz, W., Fritz, H., Stüwe, K., 2011. Lateral extrusion in the Eastern Alps revisited: refining the model by thermochronological, sedimentary, and seismic data. Tectonics 30, TC4006. http://dx.doi.org/10.1029/2010TC002782. Wortel, M.J.R., Spakman, W., 2000. Subduction and slab detachment in the Mediterranean– Carpathian region. Science 290, 1910–1917. Zötl, J., Goldbrunner, J.E., 1993. Die Mineral- und Heilwässer Österreichs. Springer, Wien.