orogenic structure of the eastern alps, europe, from transalp deep seismic reflection profiling

www.elsevier.com/locate/tecto

Tectonophysics 388

Orogenic structure of the Eastern Alps, Europe, from TRANSALP

deep seismic reflection profiling

Ewald Lqschena,*, Bernd Lammererb, Helmut Gebrandea, Karl Millahnc,

Rinaldo Nicolichd, TRANSALP Working Group1

aDepartment fur Geo-, und Umweltwissenschaften, Universitat Munchen, Sektion Geophysik, Theresienstrasse 41, D-80333 Munchen, GermanybSektion Geologie, Theresienstrasse 41, D-80333 Munchen, Germany

cMontanuniversitat Leoben, Institut fur Geophysik, Franz-Josef-Strasse 18, A-8700 Leoben, AustriadUniversita di Trieste, Dipartimento di Ingegneria Civile, Via Valerio 10, I-34127 Trieste, Italy

Received 14 October 2003; received in revised form 31 January 2004; accepted 13 June 2004

Abstract

The TRANSALP Group, comprising of partner institutions from Italy, Austria and Germany, acquired data on a 340 km long

deep seismic reflection line crossing the Eastern Alps between Munich and Venice. Although the field work was split into four

campaigns, between fall 1998 and summer 2001, the project gathered for the first time a continuous profile across the Alps

using consistent field acquisition and data processing parameters. These sections span the orogen itself, at its broadest width, as

well as the two adjacent basins. Vibroseis and explosion data, complementary in their depth penetration and resolution

characteristics, were obtained along with wide-angle and teleseismic data. The profile shows a bi-vergent asymmetric structure

of the crust beneath the Alpine axis which reaches a maximum thickness of 55 km, and 80–100 km long transcrustal ramps, the

southward dipping dSub-Tauern-RampT and the northward-dipping dSub-Dolomites-RampT. Strongly reflective patterns of theseramps can be traced towards the north to the Inn Valley and towards the south to the Valsugana thrust belt, both of which show

enhanced seismicity in the brittle upper crust. The seismic sections do not reveal any direct evidence for the presence of the

Periadriatic Fault system, the presumed equivalent to the Insubric Line in the Western Alps. According to our new evolutionary

model, the Sub-Tauern-Ramp is linked at depth with remnants of the subducted Penninic Ocean. The dcrocodileT-type model

describes an upper/lower crustal decoupling and wedging of both the European and the Adriatic–African continents.

D 2004 Elsevier B.V. All rights reserved.

Keywords: Alpine orogeny; Seismic reflection profiling; Crustal structure; Eastern Alps

0040-1951/$ - s

doi:10.1016/j.tec

* Correspon

E-mail addr1 Universiti

(2004) 85–102

ee front matter D 2004 Elsevier B.V. All rights reserved.

to.2004.07.024

ding author. Tel./fax: +49 89 2180 4201/4205.

ess: [email protected] (E. Lqschen).es of Munich, Leoben, Salzburg, Bologna, Trieste, Milan, Rome, Zurich, GFZ Potsdam, ENI-AGIP Milan.

E. Luschen et al. / Tectonophysics 388 (2004) 85–10286

1. Introduction

When seismic research in the Alps started in the

mid-1970s, using the refraction technique, the first

information on crustal thickness and its average

velocity structure was provided showing some asym-

metry in a north–south direction and a crustal root at

about 50–55 km depth (Miller et al., 1977; Scarascia

and Cassinis, 1997). The northern and southern

Molasse basins, including the corresponding thrust

belts at the orogenic front, were the targets for seismic

and drilling exploration for hydrocarbons in the 1970s

and 1980s (Wessely and Liebl, 1996). However, the

internal structure of the orogen remained widely

unknown. This changed dramatically, when seismic

reflection profiling, as used in oil and gas exploration,

came into use in the Western Alps (Roure et al.,

1990). Considerable progress in our understanding of

the Alps evolved from national research programmes

during the 1980s and early 1990s conducted by the

Swiss NFP20, the Italian CROP and the French

ECORS programmes (Roure et al., 1990) in the

Western Alps. These were partly integrated into the

pan-European initiative of the N–S European Geo-

traverse (Blundell et al., 1992). The results gave rise

to the idea of dindenterT-tectonics to describe the

complex interactions of the European and the Adri-

atic–African continental plates during their collision,

which started approximately 50 Ma ago, after the

closure and subduction of the Penninic ocean beneath

the Adriatic plate (Pfiffner et al., 1997; Pfiffner, 1992;

Blundell et al., 1992). This stimulated models for the

Eastern Alps, although, as seen in geological and

tectonic maps (Fig. 1), their surface structures are

significantly different from those of the Western Alps.

The lack of seismic data in the Eastern Alps, as

compared to the Western Alps, was the motivation to

create the TRANSALP project, which conducted its

field operations between 1998 and 2001 by partner

institutions from Germany, Austria and Italy

(TRANSALP Working Group, 2001, 2002). Similar

Fig. 1. Location map showing the TRANSALP transect in the Eastern Alp

national programmes NFP20 (Switzerland), ECORS (France) and CROP (I

simplified from European Geotraverse EGT (Blundell et al., 1992).

3=Austroalpine units including Northern Calcareous Alps, 4=Souther

(paleogene). The topographic map on the right shows the TRANSALP ma

sectors acquired between 1998 and 2001 are marked by different colours.

undershooting deployments.

research programmes crossing whole mountain ranges

in Europe were conducted in the Alpine-age Pyrenees

(ECORS Pyrenees Team, 1988) and in the Variscan-

age Ural mountains (Berzin et al., 1996), which might

serve for comparison. One of the main advantages of

the new profile is its length of about 340 km and the

use of consistent acquisition and processing parame-

ters, enabling application of modern imaging techni-

ques. This is the first seismic line to traverse the

complete orogen at its broadest width, where max-

imum tectonic compression in the central part (Tauern

Window) is reported (Lammerer and Weger, 1998)

and part of the adjacent foreland basins. Earlier

programs in the Western Alps produced only short

and laterally displaced sections. The design of the

TRANSALP experiment was highly complex. Vibro-

seis near-vertical seismic profiling formed the core of

the field data acquisition, complemented by explosive

near-vertical seismic profiling, cross-line recording

for three-dimensional control, active-source tomogra-

phy or wide-angle recording of Vibroseis and explo-

sive sources by a stationary array for velocity control

and passive tomography by another 9- to 11-month

stationary array for seismological/lithospheric studies.

2. Geological setting

The Eastern Alps are composed of a thin-skinned

orogenic wedge, mostly of Adriatic origin to the

north, a thick-skinned wedge to the south and an

uplifted part of European basement and cover together

with oceanic rocks in the center, the Tauern window.

Its uplift re-deformed the Alpine edifice in Neogene

time and led to lateral eastward extrusion of blocks

between conjugate strike-slip faults in the nappe stack

(Ratschbacher et al., 1991; Frisch et al., 1998) and to

ductile stretching in the Tauern Window (Lammerer

and Weger, 1998).

The TRANSALP transect crosses (from north to

south) the flexural foreland basin of the Molasse at its

s. Previously completed seismic profiling in the Western Alps by the

taly) are shown for comparison on the left-hand side. Tectonic map is

1=Helvetic units, 2=Penninic units including Tauern Window,

n Alps including Dolomite mountains, 5=Periadriatic intrusions

in line, together with 7 cross-lines (Q1–Q7 from north to south). The

Gaps at the Inn Valley and at the Tauern mountains are closed using

E. Luschen et al. / Tectonophysics 388 (2004) 85–102 87


widest and a narrow folded part, which was overthrust

by its own Mesozoic substrata units (Helvetic

nappes), and rootless oceanic sediments of the

Rhenodanubian Flysch nappes. To the south, the

Austroalpine nappes are represented in the northern

part by the Northern Calcareous Alps (NCA), which

were sheared off from their basement and folded

already in Cretaceous time. Two subunits are exposed,

the thin Allg7u nappe along the northern margin and

the thick Lechtal nappe, which comprises most of the

section north of the Inn valley (e.g. Linzer et al., 1995;

Mandl, 2000; Auer and Eisbacher, 2003). The weakly

metamorphic Paleozoic basement to the NCA

stretches between the Inn valley and the Tauern

mountains, consisting of phyllites, quartzphyllites

and minor volcanic and intrusive rocks. The Tauern

window consists of imbricated and tightly folded

Hercynian granitic sills and Paleozoic and Precam-

brian paragneisses and amphibolites of European

origin in an overall east–west elongated dome

structure with a steep southern limb (Lammerer and

Weger, 1998; references therein). Steeply dipping

foliation also occurs in the Austroalpine basement

gneisses of the Adriatic plate further south until the

prominent right-lateral Periadriatic fault, the presumed

equivalent to the Insubric line in the Western Alps.

From here to the south, the Dolomite mountains are

crossed, mostly within or close to its quartzphyllite

basement, which crops out in its southernmost

position at the Agordo–Valsugana thrust. Further to

the south, several minor thrusts give rise to the

Tertiary Belluno basin and flexures, before the

Mesozoic rocks plunge under the clastic sediments

of the Venetian plain (Castellarin and Cantelli, 2000).

3. Subprojects and seismic data acquisition

The Vibroseis survey was designed to mainly

achieve a high resolution and depth penetration for

the upper and middle crust (compare table in

TRANSALP Working Group, 2002). A vibratorpoint

spacing of 100 m with four heavy vibrators, a sweep

signal of 10–48 Hz and 28 s length, a geophone group

spacing 50 m, and a spread length of 18 km in split

spread configuration with 360 recording channels

resulted in nominal 90-fold common midpoint cover-

age. Data were eightfold diversity-stacked and uncor-

related as well as field-correlated stored on magnetic

tape. Field-correlation resulted in 20 s long records.

Except for the northernmost approximately 60 km

of the transect within the Bavarian Molasse, the

Vibroseis survey was accompanied by explosive

seismic recording using shotpoints of 90 kg charge

in three 30 m deep boreholes and 5 km nominal

spacing. The explosive seismic survey was designed

to provide low-fold, but high-energy signals from the

deeper parts of the crust. Shots were fired when the

Vibroseis rolling spread arrived at both the north and

south off-end configuration, including the spare

spread with up to 1145 channels. In this manner, both

Vibroseis and explosive data production were running

simultaneously using the same recording unit. A daily

progress of up to 5 km in Vibroseis production

included recording of about two to four explosive

shots. In areas where high-impedance rocks are

exposed at the surface, such as the NCA, Vibroseis

penetration was relatively low. The explosives pro-

vided a valuable and economic seismic source for

imaging at greater depth. Previously this procedure

had been successful in the Western Alps (Pfiffner et

al., 1997) and in the Ural Mountains (Berzin et al.,

1996). It turned out to be particularly successful in the

Eastern Alps in the presence of noisy recording

conditions along highly populated and noise-contami-

nated north–south running valleys.

Seven receiver cross-lines (Q1–Q7; Fig. 6), each

approximately 20 km long, recorded off-end explo-

sive shotpoints and also passively recorded the

Vibroseis and explosive sources of the main line

between adjacent cross-line/main-line tiepoints in

order to provide three-dimensional control. All these

measurements were performed by contractor compa-

nies. The cross-line recording spreads were directly

connected to the main-line spread at the tiepoints or

were operated by additional recording systems in

slave mode connected to the main line via radio link.

While the main-line operations were rolling south-

ward, two cross-lines at any time were recording

passively. The cross-lines were mainly designed for

low-fold 3-D prestack depth migration along the

entire transect. The subsurface coverage of the

cross-lines was continuous, at least at greater depth,

and their receiver lines were much less noisy than

parts of the main line. Thus, alternative north–south

sections could be constructed. Additionally, comple-


mentary information in terms of seismic anisotropy

could be obtained. Passive cross-line recording

proved particularly useful and furthermore the addi-

tional costs were relatively low.

The whole survey was divided into three cam-

paigns due to financial constraints. The northernmost

120 km, between Freising (Bavaria) and the Inn

Valley (Austria), including cross-lines Q1 and Q2,

were completed in autumn 1998. The southernmost 50

km between Belluno and Treviso (Italy), including

cross-line Q7, were completed in winter 1998/1999

(using slightly different source parameters of 10–62

Hz sweep frequencies). These two sectors used an

asymmetric split spread configuration in order to

account for the prevailing structural dip. The central

sector of 170 km length between the Inn Valley and

Belluno used a symmetric split spread configuration,

including the cross-lines Q2 (repeated), Q3, Q4, Q5

and Q6, and was completed in autumn 1999. For the

tie-points at the Inn Valley and near Belluno over-

lapping source and receiver configurations were

deployed in order to assure a continuous survey.

The field crew handled more than 1100 receiver

channels simultaneously, sometimes up to 1500,

distributed on the main line, on cross-lines and

temporarily on both sides of the main Alpine crest.

The transect utilized predominantly north–south strik-

ing valleys, and only one major pass was crossed in

winter conditions, the Falzarego Pass in the Dolomite

mountains in Italy with heights of about 2200 m. The

receiver line was often deployed in highly populated

and noise-contaminated valleys, for example, in the

Ziller Valley in Austria and in the Valle di Tures in

Italy, thus providing difficult recording conditions.

The Vibroseis technique proved to be less sensitive to

noisy conditions than the explosive technique, mainly

because of vertical stacking using the diversity stack

and correlation algorithm. For permit reasons, both

datasets had to be recorded at daytimes, partly at late

hours. The explosive technique was superior in

regions with high-impedance rocks at the surface as

mentioned above. Permit difficulties required many of

the 5-km-spaced, 90-kg-shotpoints to be displaced by

several kilometers and partially subdivided into

several shots with smaller charges. Despite subdivid-

ing into several sectors, the whole survey was

continuous due to overlapping deployments, apart

from a 2-km-receiver gap in the Inn Valley and a 6-

km-gap at the main Alpine crest. These gaps were

filled by undershooting deployments. At the Alpine

crest this was achieved by a second recording spread

on the opposite side of the gap, operated in slave

mode by radio link, while the vibrators were rolling

from north to south. Synchronising all the operations

described above required extensive radio link through

several relay stations.

Owing to noise problems in the Valle di Tures

between S. Giovanni and Brunico and some bad shots

in this area, this sector did not illuminate the deeper

crust very well with the explosive technique. This was

the motivation for additional explosive measurements

in this sector in July/August 2001. A stationary 20-

km-long receiver spread was used at the eastern flank

of the valley in an almost completely noise-free

environment to record four shotpoints. Thus, these

final measurements were highly successful in record-

ing deep crustal reflections.

The seismic reflection profiling, consisting of

Vibroseis and explosive measurements and cross-line

recording, was also complemented by passive record-

ing using a stationary, three-component network along

the traverse. This array recorded all seismic sources in

wide-angle configuration for velocity control. Another

stationary three-component network recorded local

and teleseismic earthquake events continuously over 9

months in 1998/1999 and 2 months in 1999 for

lithospheric tomography (TRANSALP Working

Group, 2001).

4. Data processing and seismic sections

The Vibroseis data were processed using a conven-

tional common-midpoint (CMP) processing scheme,

complemented by non-conventional schemes, such as

dip-moveout (DMO) processing and pre-stack depth

migration with subsequent stacking (TRANSALP

Working Group, 2001, 2002). The conventional

CMP technique proved to be very robust despite the

presence of several strongly dipping and even cross-

dip reflection patterns. Thus, many different versions

of Vibroseis stacked and migrated sections in terms of

processing schemes, parameters and plotting scales

(1:50,000 to 1:200,000) have been produced, all

having their specific advantages for interpretation.

By iterative improvement of the processing parame-


ters crooked geometry and strongly varying recording

conditions could be accounted for. Amplitude scaling

was achieved by geometrical spreading correction,

optional automatic gain control (400 ms AGC) and

trace equalisation. Static corrections included eleva-

tion statics, velocity statics, based on tomographic

inversion of the first breaks, and subsequent residual

statics. The velocity model required for depth and

time migration was obtained from various sources. In

the layered Molasse basins stacking velocities were

used. In the central part results of tomographic

inversion of the first arrivals of Vibroseis and

explosive data recorded by the near-vertical spreads

and by the wide-aperture stationary network (obser-

vations up to 80 km offset) yielded a velocity model

for the upper 15 km. A macro-velocity model from

older, but still compatible deep seismic refraction

measurements (Miller et al., 1977) completed the

model for greater depths.

For the explosive seismic data, a separate process-

ing path was chosen. Because of the large shotpoint

interval of 5 km, and the low one- to twofold

coverage, some bad shots were omitted and the

highest quality traces were selected to form a single-

fold section, which was then normal-moveout cor-

rected and (poststack) time- and depth-migrated using

velocity models from the Vibroseis survey. After the

additional experiment in 2001, these data were

integrated into the complete explosive dataset.

Compiled stacked and migrated Vibroseis and

explosives data are compared in their complete

length (Fig. 2). The Vibroseis sections are about

300 km long after the CMP-binning of the crooked-

line geometry of the original length of 340 km. The

Vibroseis sections start in the north about 90 km

north of the Alpine front, whereas the explosive

sections start directly at the Alpine front. As

expected, the Vibroseis sections are superior in

resolution in the upper and middle crust, but show

deficits in the lower crust due to a lack of signal

energy (e.g. Steer et al., 1996). The explosive

sections fill these gaps (except a short sector

beneath the Ziller Valley, due to bad shots between

km 110 and 130), as they have been designed to

Fig. 2. Compilation of final sections of conventional CMP-processing at

surface mapping; Vibroseis stack and depth-migrated section; dynamite s

scale without any vertical exaggeration.

have greater imaging capabilities mainly in the

lower crust. Well-known difficulties in the migra-

tion of deep-crustal events require a synoptic view

on stack (zero-offset time), time and depth-migrated

sections. Incoherent reflection events, particularly at

greater depths, tend to create migration-dsmilesTbecause of their truncated character. Nevertheless,

both migration techniques were reasonably success-

ful, as demonstrated by the predominant criss-cross

reflection pattern at km 175–200 in the Vibroseis

stack section. After depth migration and similarly

after time migration with both methods, the

elements focus well and migrate to their plausible

positions and produce a bi-vergent pattern at km

150–220. However, care has to be taken on lower

crustal reflective spots and sectors (e.g. at km 130–

140 and about 16 s in the Vibroseis section). The

effects of smiles are inherent of the migration

principles, generally for deeper crustal targets in 2-

D seismic surveys. This is a drawback, particularly

for the low-fold explosive survey, which exhibits a

much less coherent image at all depths. Ray-

theoretical and Fresnel-zone principles as well as

a retrospective examination of unprocessed field

records help to distinguish real and spurious

signals.

The northernmost and southernmost parts of the

300 km long Vibroseis section display the stratified

Molasse basins, with the Tertiary base in the Bavarian

Molasse the most prominent reflection. Several

former hydrocarbon exploration targets can be clearly

recognised at antithetic normal faults. These faults

originated in an extensional upper crustal regime due

to the downbending of the whole crust.

At km 70 (Fig. 2), the transition from the unfolded

Molasse sediments (maximum 6 km depth) to the

folded Molasse sediments into the Northern Calca-

reous Alps (maximum 10–11 km depth) can be seen.

The NCA were formed from Triassic–Jurassic shelf

sediments of the former Adriatic–African continental

margin that have been upthrusted onto the European

basement. The sudden displacement by 4–5 km of the

Tertiary base and several onlap structures beneath the

Alpine front were not detected by previous industrial

full length. From top to bottom: simplified geological section from

tack and time-migrated section. All sections are shown at the same

Fig. 3. Vibroseis stack section along line-km 175–200, showing

upper part of the giant bi-vergent pattern south of the Periadriatic

Lineament.

Fig. 4. Vibroseis depth-migrated section at km 150–222 corre-

sponding to the stack section in Fig. 3. Note that the south-dipping

events have been migrated about 20 km to the north. The position of

the Periadriatic Lineament is marked according to geologic–tectonic

maps. Possible dip directions marked by the arrows are discussed in

the text.


drilling and seismic exploration in this area. This

could be an indication of a pre-Alpine, Mesozoic

graben structure, filled by Mesozoic sediments, or

normal faulting due to the additional load of the NCA

during their thrusting onto the former European

crustal basement. As expected, the folded Molasse

and the Flysch zones are characterised by a chaotic

signature, whereas the structure of the NCA seems to

be well defined with their south-dipping overturned

folds and nappes. At about 9–10 km depth the

Northern Calcareous Alps and their substrata are

bounded by a basal, almost horizontal, reflection

pattern above a relatively transparent upper crystalline

crust. This pattern may be caused by the Mesozoic

European sediments possibly including tectonic slip

surfaces, which were active during upthrusting of the

NCA.

Proceeding further south, at km 110, a prominent

south-dipping reflection pattern outcropping at the

southern flank of the Inn Valley is visible. This is in

contrast to pre-existing models where north-dipping

structures were expected (Roeder, 1989). The seismic

data can be interpreted as a thrust fault system along

which the NCA were overthrusted by their former

basement, the Greywacke Zone. Hereafter we inter-

pret this structure as the outcrop of one possible

branch of the dSub-Tauern-RampT, which can be

traced by similar reflective patterns, almost conti-

nuously towards greater depth (see further discussion

below). Beneath the Inn Valley the subhorizontal

basal reflection pattern of the NCA is again vertically

displaced by about 5 km. As a result, this pattern can

be traced subhorizontally from the Alpine front to the

Tauern Window and includes two almost vertical

displacements.

South of km 170, part of a giant bi-vergent pattern,

which is distributed throughout the crust at the Alpine

axis, is shown in Figs. 3 and 4. A criss-cross pattern in

the zero-offset stack section (Fig. 3) migrates laterally

over nearly 20 km to the opposite dip direction of the

individual elements. This sector contains one of the

most critical questions, particularly regarding the

subsurface structures of the Tauern Window and the

Periadriatic Lineament. A very pronounced pattern of

subparallel south-dipping reflections is visible as part

of the bi-vergent pattern close to the centre of the

Alps. Projected to the surface, these reflections may

be linked with structures of the Tauern Window. The

Periadriatic Lineament cannot be clearly identified.

The northward-dipping arrow marks its position

according to previous models (e.g. Castellarin and

Cantelli, 2000). One might conclude that this linea-

ment terminates all of the pronounced reflections at

their northern end but is itself almost non-reflective.

In that case, this fault would end at depth at the dSub-Tauern-RampT, which is continuous, as shown below,


located at depths somewhat greater than those

displayed in the sections of Figs. 3 and 4. In the

search for any direct reflection of a possible north-

dipping structure (e.g. by using pre-stack migration

and inspecting original pre-stack shot gathers, or by

using the passive cross-line data), no indications of

any reflective character were found, although the

acquisition and processing parameters with their

relatively large illumination aperture would allow

for it. In a similar situation in the Bohemian Massif,

Harjes et al. (1997) have shown that a steeply dipping,

several hundred meter thick fault system with low

velocities in a crystalline environment could be well

imaged by the seismic survey, even at greater depth.

If, on the other hand, one assumes that the Periadriatic

Fault system is seismically reflective, there is no other

choice than to trace it according to the southward-

dipping arrow in Fig. 4. In this case, it is obvious that

corresponding reflections are not particularly prom-

inent in comparison with the surrounding fabric. This

implies that the Periadriatic Lineament is not a first-

order structure on a N–S section and its role may have

been over-estimated in previous mountain-building

models.

One of the elements of major importance in the

TRANSALP transect is the dSub-Tauern-RampT (Fig.

Fig. 5. Vibroseis stack section between the Inn Valley (left) and the Dolom

km-long part of this section was the target for an additional explosive exp

5). The ramp between the Inn Valley and the

Dolomites is outlined over 90 km by a dotted line

on the section. Just north of the Periadriatic Linea-

ment, a 20-km-long section has been repeatedly

measured by an additional explosive experiment in

2001, as mentioned above. The resulting stack section

of this experiment demonstrates that at midcrustal

levels the south-dipping pattern, part of the dSub-Tauern-RampT, is a very dominant feature. This

pattern is characterised by an almost 2 s TWT, or

about a 6-km-thick layered sequence, which can be

traced from the surface at the Inn Valley to a depth of

more than 30 km beneath the bi-vergent structure.

Several individual elements are discernible to even

greater depths. Above this pattern, within the 20-km-

long complementary explosive sector (Fig. 5), there is

a relatively transparent zone, also visible on the left

side of Fig. 4. This confirms the existence of this

zone, since there were doubts about it in the Vibroseis

section because of the noisy recording conditions in

the Valle di Tures, which could have obscured the

reflections, and the abrupt north termination of the

criss-cross reflection pattern further south. The com-

plementary section also shows that lower crustal

reflections of the European crust actually terminate

and disappear in this orogenic root zone.

ites (right) emphasising the dSub-Tauern-RampT (dotted line). A 20-

eriment shot in 2001, shown overlain as a stack section.


The cross-line recordings were subjected to a

variety of different processing experiments. Owing

to space requirements, not all the experiments

according to all available subsurface coverage shown

in Fig. 6 can be presented. Fig. 6 describes a

conventional processing approach, applied to the

cross-line Q4 (south of Tauern Window). North–south

running binning lines with a bin width of 5 km have

been used to select the traces and to construct CMP

stack sections, according to processing steps adopted

from the main line. These sections, if mounted

together, provide an alternative stack section, com-

Fig. 6. Left: Schematic example of the layout of one particular cross-line

subsurface coverage (same layout applies for all other cross-lines). Right: C

line, using recordings of cross-line Q4. The subsurface coverage is approxi

Fig. 5. Note the predominance of the midcrustal reflection pattern dipping

plementary to the main-line stack section of an almost

identical subsurface coverage. The dominant reflec-

tion pattern on the north–south section obtained from

cross-line Q4 corresponds again to the south-dipping

dSub-Tauern-RampT, which is even more pronounced

than on the main line and can actually be regarded as

the most dominant feature in the Alpine crust. The

section of Q4 is in an almost identical location to the

complementary explosive section gathered in 2001

(Fig. 5). It is even more evident in the section of Q4

that the zone above the ramp is actually void of

significant reflections. Experiments with varying

(Q5), with all sources recorded on this line and the corresponding

onventional stack section after CMP-processing along N–S binning

mately identical with the additional explosive experiment of 2001 in

to the south, interpreted as part of the dSub-Tauern-RampT.


azimuths of the binning line and corresponding

stacking have shown that the dominant dip direction

is south and north, respectively. With these examples,

we can show that passive cross-line recordings are a

very useful and economic way to provide further

constraints for two-dimensional deep crustal reflection

surveying.

The cross-line recordings also allow observations

which depend on the azimuth between the source

and the receiver. Fig. 7 shows azimuthal variations

of the P-wave velocity. Average velocities calcu-

lated by the distance between sources and receivers

and by the corresponding traveltime of the P-wave

first arrival are plotted against distance and azi-

muth. Although some scatter is visible because of

near-surface and topography effects, a clear rela-

tionship of velocities with azimuth and offset is

discernible. As expected, the velocities increase

with offset according to the increasing depth of

these diving waves, until they remain constant at

greater offset. This behaviour is also known from

numerous velocity measurements in the laboratory,

Fig. 7. Seismic anisotropy determined by the direct arrivals of P-waves a

offset, azimuth colour-coded. Right: Average velocity versus azimuth, off

when the confining pressure of the rock samples is

increased simulating greater depth by closing cracks

and microcracks. Additionally, velocities for waves

propagating in east–west direction (azimuth 908 and

2708) are systematically more than 10% higher than

velocities of waves travelling in north–south direc-

tion (azimuth 0 and 1808). This anisotropy is

compatible with microfabric observations made in

and around the Tauern Window (Lammerer and

Weger, 1998), which show an east–west elongation

of the rock texture caused by N–S compression and

E–W stretching. A similar observation has been

made by studying the azimuthal variations of cross-

line Q3 at the northern rim of the Tauern Window.

In this cross-line, S-wave splitting has been

observed in shot gathers recorded in east–west

direction, further direct proof of seismic anisotropy.

None of the other cross-lines exhibit such an

azimuthal variation of velocities. This is clear

evidence of intrinsic rock anisotropy due to tectonic

paleostrain constrained to the Tauern Window and

its surroundings.

t cross-line Q4. Left: Average velocity (distance/traveltime) versus

set colour-coded.


5. Interpretation and discussion

In the debate of the role of compression (Western

Alps, e.g. Pfiffner et al., 1997) versus lateral extrusion

tectonics (Eastern Alps, e.g. Selverstone, 1988;

Ratschbacher et al., 1991; Frisch et al., 1998), a

two-dimensional crustal section, as obtained from the

north–south oriented TRANSALP project, has its

inherent limitations. However, we have presented

three key observations which clearly emphasise

compression tectonics in terms of thrusting and

wedging as the dominant processes during mountain

building of the Eastern Alps. Compression tectonics

require much more stress than extension and extrusion

tectonics, and hence have the potential to produce

stronger seismic events. The key observations are the

following:

(1) Bi-vergent structure at whole crustal scale

beneath the Alpine axis. This bi-vergent pattern

is characterised by some asymmetry and culmi-

nates in a relatively narrow zone above the

crustal root zone. The more prominent south-

ward dipping reflections (Fig. 4) are grouped

subparallelly like a huge stack of stratified

medium, but show limited lateral extension.

One gets the impression that the most dramatic

events occurred here, close to the Alpine axis. If

projected to the surface, these reflections may be

linked to steeply southward dipping structures of

the Tauern Window. A direct link, particularly of

the most pronounced deeper structures, is not

possible. They seem to terminate along a steeply

northward-dipping line originating close to the

Periadriatic Lineament. Petrophysical data from

surface samples are available from Mazzoli et al.

(2002), who concluded that interfaces between

amphibolites and other rocks could provide

sufficient impedance contrasts to produce seis-

mic reflections. However, such sample-derived

characteristics are too sparse and do not take into

account other properties necessary to produce

reflections, such as some lateral continuity or

interferences by thin-layering. Instead, we prefer

to interpret the seismic reflections as a conse-

quence of physical alteration or weakening of

the rocks rather due to compositional effects.

This interpretation is justified by general expe-

rience obtained from deep seismic research

during the last few decades that the most

prominent seismic reflections are frequently

caused by fracture and fault zones (e.g. Harjes

et al., 1997). Therefore, we interpret the sub-

parallel reflections of the bi-vergent pattern as

fracture and slip surfaces within a deforming

crustal wedge. Fracturing produces thin low-

interval-velocity zones. This is presumably the

cause of the inverse polarity of strong reflections

observed at about 6 s traveltime in unprocessed

seismograms (Fig. 8). The subvertical sequence

of these reflections could be explained by the

contrast increasing effect of the intrusion of

fluids migrating from dewatering sediments

trapped at greater depth (e.g. ANCORP Working

Group, 1999) after the closure of the Penninic

Ocean.

A competing interpretation attributes the south-

dipping subparallel reflections to the Permian

evolution of the pre-Alpine Adriatic crust.

Underplating mafic lenses or sills may be found

at deeper parts of extended continental crust.

Reflectivity could be caused by mafic rocks

which interfinger with crustal restites. Sinigoj et

al. (1995) suggested that the deeper crust of the

Ivrea Zone was affected by magmatic under-

plating. An analogue model could be applied to

the Dolomites to explain the heat needed to

generate Permian granites and porphyrites.

(2) Transcrustal ramp-like structures (Figs. 5 and 6).

The bi-vergent pattern mentioned above appears

to be bounded at depth by a predominant, about

1–2 s (approximately 3–6 km) thick, reflection

pattern that originates at the Inn Valley. This

pattern dips to the south at an angle of about 308and is nearly 100 km long. It is particularly

strong and dominant just beneath the Tauern

Window. From there southward to greater depth,

the seismic signals of the presumed ramp change

to more widely distributed single individual

reflections dipping southward, which are visible

in the Alpine root zone. Shear surfaces within

this band are considered to be responsible for the

seismic impedance contrasts within the upper

part of the ramp, possibly due to rock anisotropy

caused by the elongated microfabric of the rocks

involved and by cataclastic deformation within

Fig. 8. Comparison of the polarity of seismic reflections with the first arrivals of the same traces within a non-processed Vibroseis record. Upper

left: selected traces, 1 s long, with first arrivals. Lower left: reflections at about 6 s TWT recorded on the same traces as above, corresponding to

the reflection in Fig. 4, left side. Lower right: same reflections with inverted polarity. Upper right: qualitative explanation using an embedded

low-velocity layer (approximately 500 m thickness and 10% decrease of velocity within a background medium of about 6 km/s). Note the

symmetric signal, containing one positive or negative maximum and two side lobes, typical for Vibroseis signals.


the brittle rheological regime. At greater depths,

most impedance contrasts vanish because of the

possible involvement of oceanic-type crustal

rocks and of ductile overprint, but may be

locally enhanced due to released fluids and/or

magmatic processes. A transcrustal ramp at a

similar position has already been proposed by

Lammerer and Weger (1998) based upon defor-

mation studies in the Tauern Window. The

internal folds, faults and the strain distribution

cannot explain the 20 km of rapid uplift of the

rocks of the Tauern Window during the last 20

Ma. At the surface, there are tectonic contacts

between lower Triassic resting upon Upper

Triassic. Some, but not all, of the thrust

deformation must be accommodated by this

outcropping trace. Other branches of this thrust

could be traced to the basal reflectors of the

Northern Calcareous Alps. Nevertheless, these

seem to be overthrusted by their former base-

ment, the Greywacke zone. The Inn Valley is

also known for its still active sinistrial move-

ment, due to lateral escape tectonics interfering

with compression tectonics in this area. This is

expressed by a relatively narrow belt of earth-

quake activity at shallow depth, within the brittle

regime, with thrust and strike slip focal mech-

anisms, according to TRANSALP seismological

studies of local earthquakes. Another belt of

active seismicity is located in the Southern Alps,

apparently conjugate to that of the Inn Valley.

Although in contrast, the Valsugana and pied-

mont fault belt clearly shows thrust-type focal

mechanisms (Slejko et al., 1989). We interpret

northward-dipping reflections in the middle crust

beneath the Dolomite Mountains in the context

of a crustal-scale backthrust system (Sub-Dolo-

mites-Ramp).

(3) Crustal root and asymmetric crustal structure.

The line-drawing of Fig. 9 presents an over-

view of the Alpine crustal structures, compiled

from Vibroseis and dynamite sections as well

as from cross-line recordings. Structure and

velocity models, the latter obtained from wide-

Fig. 10. Cartoon of two alternative structure and evolutional models. See discussion in the text and in TRANSALP Working Group (2002).


angle tomography (Bleibinhaus, 2003), exhibit

a remarkable asymmetry. The non-sedimentary

part of the European crust is continuously

downbending and almost transparent, except a

thin highly reflective layered lower crust. The

Adriatic–African crust, including the Tauern

Window as the hanging wall of the dSub-Tauern-RampT, is reflective at all levels. Its

lower crust is more than twice as thick as that

of the European crust. The middle–lower

Adriatic–African crust reveals two distinct,

almost subparallel, patterns of seismic reflec-

tions. This 25 km (or doubled?) thickness of

the lower crust is confirmed by corresponding

velocity peaks determined by previous seismic

refraction profiling and seems to be restricted

to the Southern Dolomites (Scarascia and

Fig. 9. Compilation of Bouguer anomaly of gravity (on top), simplified ge

and dynamite seismic sections as well as from alternative cross-line N–S se

tomography (from Bleibinhaus, 2003). Dotted line corresponds to the pre

Tauern Window has been upthrusted.

Cassinis, 1997). Receiver Functions as

obtained from the seismological studies

(TRANSALP Working Group, 2002; Kum-

merow, 2003) confirm the general crustal

configuration. The crustal root zone at 55–60

km depth (at line-km 160–170) lacks reflec-

tivity. This has been confirmed by the addi-

tional explosive experiment in 2001, which

was characterised by an extremely high signal-

to-noise ratio at the corresponding recording

times of 15–20 s. Vosteen et al. (2003)

estimated temperatures of about 800 8C at

55 km depth in the Alpine root zone from

steady-state forward and inverse simulation of

the conductive heat transport. This implies at

least partial anatexis in the deepest parts of the

European crust and offers an explanation for

ological section and manual line drawing, compiled from Vibroseis

ctions (middle), and seismic velocity model after wide-angle seismic

sumed dSub-Tauern-RampT, a tectonic shear zone along which the


the seismic transparency in that depth range.

The dSub-Tauern-RampT seems to separate two

different crustal domains. The velocity struc-

ture is asymmetric too; the European middle

and lower crust are characterised by signifi-

cantly lower velocities than the Adriatic–

African crust (Fig. 9). The crustal root, as

seen in the structure and velocity model, is

shifted to the south by about 50 km, with

respect to the main topographic crest and the

Bouguer anomaly of the gravity. The Bouguer

anomaly shows a well-pronounced, long wave-

length minimum of about �160 mGal just at

the Tauern Window and the main Alpine crest

with relatively smooth slopes towards the

foreland basins. This is remarkably different

from the Western Alps, where a strong local

maximum within the regional minimum corre-

lates with exhumed upper-mantle rocks at the

Ivrea Zone. A fit of the density model

(Ebbing, 2002) with the observed gravity can

Fig. 11. Restoring model A (bcrocodile modelQ). Sections A and B are at

length. Notice that the lower Adriatic–African crust is stacked during col

be achieved, if lower density is introduced into

the middle and lower European crust, consistent

with the velocity model. These asymmetries

indicate there is still a subduction polarity

towards the south, and that the dSub-Tauern-RampT may be regarded as the upward continu-

ation of the former subduction shear zone. Recent

high-resolution lithospheric tomography during

the TRANSALP campaign (Lippitsch, 2002)

shows an anomalous high-velocity, nearly 200

km long slab which dips towards southeast in the

Western Alps and plunges steeply in the Eastern

Alps lacking a clear polarity.

In a previous paper by the TRANSALP Working

Group (2002), two alternative models explaining the

general features of the transect have been proposed:

the dcrocodile modelT and the dlateral extrusion

modelT (Fig. 10). Both models are characterised by

the dominant role of transcrustal ramps and inter-

wedging at midcrustal levels. The dcrocodile modelT

shown at same scale. Section C is at different scale because of its

lision.


better explains the main mountain-building pro-

cesses, with respect to crustal thickness, producing

a mountain root, and the general seismic fabric

along the transect. Lateral extrusion, e.g. at the Inn

Valley fault in the north and the Periadriatic Linea-

ment in the south, is not excluded in the model, but

we consider it to be restricted to the nappe

complexes in the orogenic lid. The dlateral extrusionmodelT on the other hand focuses on the escape

deformation of the Tauern Window and attributes a

major importance to the Periadriatic Lineament (and

relating faults with dominant strike slip mechanism)

which dips to the north according to dip information

obtained at the surface. Laubscher (1971) and

Schmid et al. (1989) reported dextral offset of 300

km along the Periadriatic Lineament. In this model,

the evolution of the thick Adriatic lower crust is

attributed to Permian pre-Alpine magmatic under-

plating (Dal Piaz, 1993). Fig. 11 shows an attempt

to restore one of the above mentioned structure

models. It concentrates on the two-dimensional

behaviour and clearly demonstrates that there is no

obvious contradiction to this assumption. Addition-

ally, it explains the increased lower crustal thickness

on the Adriatic–African side by tectonic erosion

during subduction and subsequent stacking during

Alpine compression.

With the TRANSALP sections, we have contrib-

uted new and complete images of the Alpine structure.

The bi-vergent structure and the transcrustal ramps

are, in particular, more pronounced than in any section

from the Western Alps (Pfiffner et al., 1997).

Although we now possess a great wealth of seismic

sections for the whole Alpine chain, it is debatable

whether we are fully able to understand the Alpine

evolution in all three dimensions. We feel that this

cannot be solved by the integration of our new dataset

into existing models, but also requires the re-assess-

ment and reinterpretation of the older datasets.

Acknowledgements

The TRANSALP programme is jointly financed by

the Bundesministerium fqr Bildung und Forschung

(BMBF, Bonn), the Bundesministerium fqr Wissen-

schaft und Verkehr (BMWV, Vienna), the Consiglio

Nazionale delle Ricerche (CNR, Rome) and the

company ENI-AGIP (Milan). The TRANSALP group

comprises partners from universities of Munich,

Leoben, Bologna, Salzburg, Milan, Rome, Trieste

and Zurich and from GFZ Potsdam and company

ENI-AGIP. We thank our contractors, THOR Geo-

physikalische Prospektion, Kiel, Germany, DMT,

Essen, Germany, JOANNEUM Research, Leoben,

Austria, GEOITALIA, Milan, Italy and GEOTEC,

Campobasso, Italy, for their excellent and enthusiastic

work in difficult terrain and under difficult logistical

conditions, even during winter. We appreciate admin-

istrative services provided by the GFZ Potsdam. The

instruments for passive and active tomography were

provided by the Geophysical Instrument Pool of the

GFZ Potsdam and by the Universities of Munich,

Zqrich, Potsdam and Genoa. The seismic programme is

accompanied by other interdisciplinary research proj-

ects funded and coordinated by the Deutsche For-

schungsgemeinschaft (DFG), Osterreichischer Fonds

zur Ffrderung der wissenschaftlichen Forschung and

the Consiglio Nazionale delle Ricerche (CNR). We

thank Alex Nichols for his corrections of the manu-

script as well as the editor Fred Davey and the

reviewers for helpful comments and corrections.

References

ANCORP Working Group, 1999. Seismic reflection image

revealing offset of Andean subduction-zone earthquake loca-

tions into oceanic mantle. Nature 397, 341–344.

Auer, M., Eisbacher, G.H., 2003. Deep structure and kinematics of

the Northern Calcareous Alps (TRANSALP Profile). Int. J.

Earth Sci. 92, 210–227.

Berzin, R., Oncken, O., Knapp, J.H., Perez-Estaun, A., Hismatulin,

T., Yunusov, N., Lipilin, A., 1996. Orogenic evolution of the

Ural Mountains: results from an integrated seismic experiment.

Nature 274, 220–221.

Bleibinhaus, F., 2003. Entwicklung einer simultanen refraktions-

und reflexionsseismischen 3D-Laufzeittomographie mit Anwen-

dung auf tiefenseismische TRANSALP-Weitwinkeldaten aus

den Ostalpen. PhD thesis. University of Munich.

Blundell, D., Freeman, R., Mueller, St. (Eds.), 1992. A Continent

Revealed—The European Geotraverse. Cambridge Univ. Press,

Cambridge.

Castellarin, A., Cantelli, L., 2000. Neo-Alpine evolution of the

Southern Eastern Alps. J. Geodyn. 30, 251–274.

Dal Piaz, G.V., 1993. Evolution of Austro-Alpine and Upper

Pennidic basement in the northwestern Alps from Variscan

convergence to Post-Variscan extension. In: von Raumer, J.F.,

Neubauer, F. (Eds.), Pre-Mesozoic Geology in the Alps.

Springer Verlag, pp. 328–343.


Ebbing, J., 2002. 3-D Dichteverteilung und isostatisches Verhalten

der Lithosph7re in den Ostalpen. PhD thesis. Freie Universit7tBerlin.

ECORS Pyrenees Team, 1988. The ECORS deep reflection seismic

survey across the Pyrenees. Nature 311, 508–511.

Frisch, W., Kuhlemann, J., Dunkl, I., Brqgel, A., 1998.

Palinspastic reconstruction and topographic evolution of the

Eastern Alps during late Tertiary tectonic extrusion. Tectono-

physics 297, 1–15.

Harjes, H.-P., Bram, K., Dqrbaum, H.-J., Gebrande, H.,

Hirschmann, G., Janik, M., Klfckner, M., Lqschen, E.,

Rabbel, W., Simon, M., Thomas, R., Tormann, J., Wenzel,

F., 1997. Origin and nature of crustal reflections: results from

integrated seismic measurements at the KTB superdeep

drilling site. J. Geophys. Res. 102, 18267–18288.

Kummerow, J., 2003. Strukturuntersuchungen in den Ostalpen

anhand des teleseismischen TRANSALP-Datensatzes. PhD

thesis. Freie Universit7t Berlin.Lammerer, B., Weger, M., 1998. Footwall uplift in an orogenic

wedge: the Tauern window in the Eastern Alps of Europe.

Tectonophysics 285, 213–230.

Laubscher, H.P., 1971. Large-scale kinematics of the western Alps

and the northern Apennines and its palinspastic implications.

Am. J. Sci. 271, 193–226.

Linzer, H.-G., Ratschbacher, L., Frisch, W., 1995. Transpressional

collision structures in the upper crust; the fold-thrust belt of the

Northern Calcareous Alps. In: Neubauer, F., Ebner, F., Wall-

brecher, E. (Eds.), Tectonics of the Alpine–Carpathian–Panno-

nian Region. Tectonophysics, vol. 242, pp. 41–61.

Lippitsch, R., 2002. Lithosphere and upper mantle P-wave velocity

structure beneath the Alps by high-resolution teleseismic

tomography. PhD thesis. ETH Zqrich.Mandl, G., 2000. The Alpine sector of the Tethyan shelf—

examples of Triassic to Jurassic sedimentation and deforma-

tion from the Northern Calcareous Alps. Mitt. Osterr. Geol.

Ges. 82, 61–77.

Mazzoli, C., Sassi, R., Burlini, L., 2002. Experimental study of the

seismic properties of the Eastern Alps (Italy) along the Aurina–

Tures–Badia Valleys transect. Tectonophysics 354, 179–194.

Miller, H., Gebrande, H., Schmedes, E., 1977. Ein verbessertes

Strukturmodell fqr die Ostalpen, abgeleitet aus refraktionsseis-

mischen Daten unter Berqcksichtigung des Alpen-L7ngsprofils.Geol. Rundsch. 66, 289–308.

Pfiffner, A., 1992. Alpine orogeny. In: Blundell, D., Freeman, R.,

Mueller, St. (Eds.), A Continent Revealed—The European

Geotraverse. Cambridge Univ. Press, Cambridge, pp. 180–190.

Pfiffner, O.A., Lehner, P., Heitzmann, P., Mueller, St., Steck, A.

(Eds.), 1997. Results of NRP20-Deep Structure of the Alps.

Birkh7user Verlag, Basel.Ratschbacher, L., Frisch, W., Linzer, H.-G., Merle, O., 1991. Lateral

extrusion in the Eastern Alps. Tectonics 10, 257–271.

Roeder, D., 1989. South Alpine thrusting and trans-Alpine

convergence. In: Coward, M.P., Dietrich, D., Park, R.G.

(Eds.), Alpine Tectonics, Spec. Publ. - Geol. Soc. Lond., vol.

45, pp. 211–227.

Roure, F., Heitzmann, P., Polino, R. (Eds.), 1990. Deep structure of

the Alps. Mem. Soc. Geol. Fr., Paris, 156; Mem. Soc. Geol.

Suisse, Zqrich, 1; Vol. Spec. Soc. Geol. It., Roma, 1. 367 pp.

Scarascia, S., Cassinis, R., 1997. Crustal structures in the central-

eastern Alpine sector: a revision of the available DSS data.

Tectonophysics 271, 157–188.

Schmid, S.M., Aebli, H.R., Heller, F., Zingg, A., 1989. The role of

the Periadriatic Line in the tectonic evolution of the Alps. In:

Coward, M.P., Dietrich, D., Park, R.G. (Eds.), Alpine Tectonics,

Spec. Publ. - Geol. Soc. Lond., vol. 45, pp. 153–171.

Selverstone, J., 1988. Evidence for east–west crustal extension in

the Eastern Alps. Tectonics 7, 87–105.

Sinigoj, S., Quick, J.E., Mayer, A., Demarchi, G., 1995. Density

controlled assimilation of underplated crust, Ivrea zone, Italy.

Earth Planet. Sci. Lett. 129, 183–191.

Slejko, D., Carulli, G.B., Nicolich, R., Rebez, A., Zanferrari, A.,

Cavallin, A., Doglioni, C., Carraro, F., Castaldini, D., Iliceto,

V., Semenza, E., Zanolla, C., 1989. Seismotectonics of the

eastern Southern Alps: a review. Boll. Geofis. Teor. Appl. 31,

109–136.

Steer, D.N., Brown, L.D., Knapp, J.H., Baird, D.J., 1996.

Comparison of explosive and Vibroseis source energy pene-

tration during COCORP deep seismic reflection profiling in the

Williston Basin. Geophysics 61, 211–221.

TRANSALP Working Group, 2001. European orogenic processes

research transects the Eastern Alps. EOS Trans, AGU 82, 453

and 460–461.

TRANSALP Working Group, 2002. First deep seismic images of

the Eastern Alps reveal giant crustal wedges and transcrustal

ramps. Geophys. Res. Lett. 29, doi:10.1029/2002GL014911.

Vosteen, H.-D., Rath, V., Clauser, C., Lammerer, B., 2003. The

thermal regime of the Eastern Alps from inversion analyses

along the TRANSALP profile. Phys. Chem. Earth 28, 393–405.

Wessely, G., Liebl, W. (Eds.), 1996. Oil and Gas in Alpidic

Thrustbelts and Basins of Central and Eastern Europe, EAGE

Spec. Publ. vol. 5. Geological Society, London.

http://dx.doi.org/doi:10.1029/2002GL014911

orogenic structure of the eastern alps, europe, from transalp deep seismic reflection profiling

Documents