vrancea_99_hauser.pdf

8/14/2019 Vrancea_99_Hauser.pdf

1/24

VRANCEA99the crustal structure beneath the southeastern

Carpathians and the Moesian Platform from a seismic

refraction profile in Romania

F. Hausera,*, V. Raileanu b, W. Fielitz c, A. Balab, C. Prodehl a,G. Polonic d, A. Schulze e

aGeophysical Institute, University of Karlsruhe, Hertzstr. 16, D-76187 Karlsruhe, GermanybNational Institute for Earth Physics, P.O. Box MG-2, RO-76900 Bucuresti-Magurele, Romania

cGeological Institute, University of Karlsruhe, Kaiserstr. 12, D-76131 Karlsruhe, GermanydThe Institute of Geodynamics, 19-21 J.L. Calderon St., RO-70201 Bucuresti-32, Romania

eGeoForschungsZentrum, Telegrafenberg, D-14473 Potsdam, Germany

Received 2 May 2001; accepted 3 September 2001

Abstract

The VRANCEA99 seismic refraction experiment is part of an international and multidisciplinary project to study the

intermediate depth earthquakes of the Eastern Carpathians in Romania. As part of the seismic experiment, a 300-km-longrefraction profile was recorded between the cities of Bacau and Bucharest, traversing the Vrancea epicentral region in NNE

SSW direction. The results deduced using forward and inverse ray trace modelling indicate a multi-layered crust. The

sedimentary succession comprises two to four seismic layers of variable thickness and with velocities ranging from 2.0 to 5.8

km/s. The seismic basement coincides with a velocity step up to 5.9 km/s. Velocities in the upper crystalline crust are 5.96.2

km/s. An intra-crustal discontinuity at 1831 km divides the crust into an upper and a lower layer. Velocities within the lower

crust are 6.77.0 km/s. Strong wide-angle PmP reflections indicate the existence of a first-order Moho at a depth of 30 km near

the southern end of the line and 41 km near the centre. Constraints on upper mantle seismic velocities (7.9 km/s) are provided

by Pn arrival times from two shot points only. Within the upper mantle a low velocity zone is interpreted. Travel times of a PLP

reflection define the bottom of this low velocity layer at a depth of 55 km. The velocity beneath this interface must be at least

8.5 km/s. Geologic interpretation of the seismic data suggests that the Neogene tectonic convergence of the Eastern Carpathians

resulted in thin-skinned shortening of the sedimentary cover and in thick-skinned shortening in the crystalline crust. On the

autochthonous cover of the Moesian platform several blocks can be recognised which are characterised by different lithologicalcompositions.This could indicate a pre-structuring of the platform at Mesozoic and/or Palaeozoic times with a probable active

involvement of the Intramoesian and the CapidavaOvidiu faults. Especially the Intramoesian fault is clearly recognisable on

the refraction line. No clear indications of the important Trotus fault in the north of the profile could be found. In the central part

0040-1951/01/$ - see front matterD 2001 Elsevier Science B.V. All rights reserved.P I I : S 0 0 4 0 - 1 9 5 1 ( 0 1 ) 0 0 1 9 5 - 0

* Corresponding author. Tel.: +49-721-608-4592; fax: +49-721-71173.

E-mail address:[email protected] (F. Hauser).

www.elsevier.com/locate/tecto

Tectonophysics 340 (2001) 233256


2/24

of the seismic line a thinned lower crust and the low velocity zone in the uppermost mantle point to the possibility of crustal

delamination and partial melting in the upper mantle. D 2001 Elsevier Science B.V. All rights reserved.

Keywords: Vrancea earthquakes; Crustal structure; Refraction seismic; Eastern Carpathians; Dobrogea

1. Introduction

The Carpathian Orogen in Romania (Fig. 1) is the

result of Cretaceous and Neogene convergence and the

resulting mountain range is the site of still ongoing

neotectonic activity. This activity consists of near-sur-

face crustal deformation (compressional, extensional

and strike slip) and of strong seismicity at intermediate

depth (60180 km), which is concentrated in a small

area of the Eastern Carpathians in Romania called the

Vrancea zone (Fig. 2, Oncescu et al., 1998). This

localised seismicity is one of only three known places

worldwide of such a concentration of intermediate

depth earthquakes. The other two being the Bucara-

manga region in the Andes of Columbia (e.g. Taboada

et al., 2000) and a place in the HinduKushPamir

area of central Asia (e.g. Mellors et al., 1995). In all

these areas the geodynamic setting is not yet clear.

Subduction related processes have been suggested, in

particular, for the Vrancea area (Roman, 1970; Radu-

lescu and Sandulescu, 1973; Airinei, 1977; Fuchs et al.,1979; Constantinescu and Enescu, 1984; Constanti-

nescu et al., 1973; Oncescu, 1984; Linzer, 1996; Kovac

et al., 2000; Csontos, 1995; Girbacea and Frisch, 1998;

Nemcok et al., 1998; Seghedi et al., 1998).

The Vrancea zone is overlapping with the south-

eastern outer area of the Eastern Carpathian bend (Fig.

2). While the shallow seismic activity scatters widely

and has moderate magnitudes (Mw 5.6), the epicen-

tral region of the intermediate depth seismicity is

confined to an area of only about 40 km 80 km

(Oncescu et al., 1998). These earthquakes occurbetween 60 and 180 km depth (Figs. 3 and 4) within

an almost vertical elongated narrow zone and fre-

quently have large magnitudes (Mw 7.4) that have

caused a high toll of casualties and extensive damage

over the last centuries. Deeper events have also been

recorded, but show only small magnitudes. The depth

interval of the strong events is separated from the

crustal events by a zone of weak seismicity, which is

located between about 40 and 60 km depth (Fuchs et

al., 1979; Oncescu et al., 1998).

Because of the unknown relationship of this zone of

high seismicity to the geologic structures recognised at

the surface, a joint GermanRomanian research pro-

gram was set up. It comprises the Collaborative

Research Centre 461 (CRC 461) Strong Earth-

quakesa Challenge for Geosciences and Civil Engi-

neering at the University of Karlsruhe (Germany) and

the Romanian Group for Vrancea Strong Earthquakes

(RGVE) at the Romanian Academy in Bucharest

(Wenzel, 1997; Wenzel et al., 1998a).

The VRANCEA99 seismic refraction project (Fig.

1) presented in this paper is a contribution to this

research program. It was designed to study the crustal

and uppermost mantle structure to a depth of about 70

km underneath the Vrancea epicentral region. The

project was jointly performed by the Geophysical and

Geological Institutes of the University of Karlsruhe

(Germany), the GeoForschungsZentrum in Potsdam

(Germany) and the National Institute for Earth Physics

in Bucharest (Romania).

The crustal structure and seismic velocities obtainedby this project will further be used to refine the geo-

dynamic model for the Carpathian Orogen and to

calibrate the relative velocities obtained by a subse-

quently performed teleseismic tomography project

(Wenzel et al., 1998b). The detailed knowledge of the

velocity-depth structure and of the physical properties

of the crust and the upper mantle will help to under-

stand the propagation of seismic waves in the region. It

will, therefore, also contribute to the seismic risk

assessment for the metropolitan area of Bucharest,

due to its proximity of 100160 km to the Vranceaepicentral region with its preferably NNE SSW-direc-

ted wave propagation (Bonjer et al., 1998; Oncescu et

al., 1998; Wenzel and Lungu, 2000; Wenzel et al.,

1998a,c).

2. Geological and tectonic setting

The Eastern Carpathians are part of the Alpine

Carpathian orogenic belt and the result of the collision

F. Hauser et al. / Tectonophysics 340 (2001) 233256234


3/24

of several microplates with the European margin,

while closing the former Tethys Ocean (Dercourt et

al., 1993; Csontos, 1995; Channell and Kozur, 1997;

Stampfli et al., 1998; Linzer et al., 1998; Neugebauer

et al., 2001). The result of two main compressional

events, which took place during the Cretaceous and

the Neogene, are several tectonic units, which have

been accreted in the Carpathian area. During the

Cretaceous the Inner Carpathians or Dacides (after

the definition of Sandulescu, 1988, see Fig. 2 for

Fig. 1. Topographic map of Romania showing the geographical location of the VRANCEA99 seismic refraction lines. The contour lines indicate

the depth to Moho (after Radulescu, 1988). The dashed lines labelled with roman numbers are the early International Refraction Profiles of the

1970s.

F. Hauser et al. / Tectonophysics 340 (2001) 233256 235


4/24

Fig. 2. Geological overview of the Eastern Carpathian bend area and its foreland with the main crustal units, nappe structures, and faults. The lo

refraction lines are shown with their shot points. The thick solid lines indicate the location of the geological cross sections for Figs. 3, 4 and 10. C

in the text.


5/24

location) converged and in the Neogene the accreted

complex was transported to the present position (San-

dulescu, 1988; Royden, 1988; Csontos, 1995; Ma-

tenco and Bertotti, 2000). During the Neogene the

Outer Carpathians or Moldavides (after the definition

of Sandulescu, 1988, see Fig. 2 for location) got theirpresent-day geometry (Sandulescu, 1988; Ellouz et

al., 1994; Zweigel et al., 1998; Matenco, 1997).

The present-day Eastern Carpathians are made up

of several major tectonic units derived from this

activity (Fig. 2, Sandulescu, 1984; Badescu, 1998).

To the west the nappes of the Median Dacides re-

present a part of the Tisza Dacia microplate (after

Csontos, 1995). Depending on the geodynamic inter-

pre tation, the nappes of the Outer Dacides are

explained in two different ways: (1) as a Jurassic to

Cretaceous rift along the European margin (Sandu-

lescu, 1988) or (2) as a Neogene oceanic domain

between the Tisza Dacia microplate and the Euro-

pean margin (Csontos, 1995; Nemcok et al., 1998;

Stampfli et al., 1998; Linzer et al., 1998; Neugebauer

et al., 2001). They mark the transition to the Molda-vides, which are part of the European margin and

which grade into the foredeep. In the centre the

Dacidian nappes are partially covered by the Paleo-

gene to Neogene Transylvanian Basin, which was also

affected by some Neogene to Quaternary compres-

sional deformation (Ciulavu et al., 2000). Together

with an area of Late Pliocene to Quaternary intra-

mountain graben structures in the centre of the Carpa-

thian bend area, the foredeep is affected by the

youngest tectonic deformation. The front of the com-

Fig. 3. Pre-1999 geological section along the main NNE SSW VRANCEA99 seismic-refraction line with alternative interpretations of the

expected geological structures. Inside the Eastern Carpathians the section follows mostly the trend of main geological structures, whereas to the

south it becomes transverse to the main geological structures. For location see Fig. 2. Circles represent the foci of the intermediate depth

earthquakes of the Vrancea zone after Oncescu et al. (1998) projected onto this cross-section. Compiled from various sources given in the text.



6/24

pressional deformation is nowadays located inside the

foredeep and is represented by the Pericarpathian

Front (Fig. 2). These outermost nappes already thrust

over the foreland platform and are covered by youngerand undeformed sedimentary layers. This cover

extends further to the east and the south onto the

Moesian and Scythian platforms.

The Moldavidian zone is characterized by several

nappes with a fold-and-thrust belt geometry (Sandu-

lescu, 1984, 1988; Ellouz et al., 1994; Morley, 1996;

Badescu, 1998; Zweigel et al., 1998; Matenco and

Bertotti, 2000). The main nappes from west to east

are: the Inner Moldavide nappes, the Tarcau nappes, the

Marginal Folds and the Subcarpathian nappes (Fig. 2).

The Tarcau and Marginal Folds nappes are made up

mainly of Cretaceous marine basin sediments and

Paleogene to Neogene flysch and other clastic sedimen-

tary deposits, whereas the Subcarpathian nappe consistsmainly of molasse deposits. The lithology of the nappes

consists of shales and sandstones with subordinate

marls, limestones, tuff and conglomerates. The Mar-

ginal Folds and Subcarpathian nappes contain also

Neogene evaporitic formations like salt and/or gypsum

(Sandulescu, 1988; Matenco and Bertotti, 2000).

The thickness of the Moldavidian nappes is con-

strained by reflection seismic data, mainly from oil

exploration in the Subcarpathian nappe, but is only

interpolated through balanced cross-sections for the

Fig. 4. Pre-1999 geological section along the short W E VRANCEA99 seismic refraction line along the Putna valley with alternative

interpretations of the expected geological structures. The section is transverse to the main structures. For location see Fig. 2. Circles represent

the foci of the intermediate depth earthquakes of the Vrancea zone after Oncescu et al. (1998) projected onto this cross-section. Compiled from

various sources given in the text.



7/24

Marginal Folds and Tarcau nappes (Stefanescu and

Working Group, 1988; Ellouz et al., 1994; Morley,

1996; Matenco and Bertotti, 2000). These cross-sec-

tions are constrained by surface geology and boreholedata, but the structures of the deeper nappes are fairly

unknown and their balancing therefore less accurate.

The International Geotraverse XI (Fig. 1) which

crossed the Carpathians between Focsani and Targu

Secuiesc gives only some vague estimations on nappe

thickness (Radulescu et al., 1976). Magnetotelluric

data from Stanica and Stanica (1998) indicate a depth

of 8 km for the base of the Moldavidian nappes and of

about 1416 km for the basement in the area around

shot point D (Figs. 13).

The Carpathian foreland consists of older consoli-

dated crustal blocks with their pre-orogenic autochtho-

nous cover (Sandulescu, 1984; Visarion et al., 1988;

Ellouz et al., 1994; Zugravescu and Polonic, 1997;

Seghedi, 1998; Matenco and Bertotti, 2000). It is part

of the Moesian Platform to the south and the Scythian

Platform to the northeast (part of the Eastern European

platform), which are distinguished by geophysical

(mainly magnetic) anomalies in the crystalline base-

ment and lithologic differences in the sedimentary

cover (Seghedi, 1998 and references therein) and

which are separated by the Trotus fault (Fig. 2). The

crystalline basement of the two platforms is made up ofmetamorphic and intrusive magmatic rocks, which

sometimes provide a weak acoustic contrast to their

older sedimentary cover (Raileanu et al., 1994). The

sedimentary rocks, separated by several unconform-

ities, consist of Paleozoic and Mesozoic detritic and

carbonaceous deposits and a very thin undeformed

Neogen cover, which shows a slight dip towards the

central part of the foredeep (Raileanu et al., 1994). Near

the orogenic front of the Carpathians the platform

sediments are partly covered by the foredeep sedi-

ments. The sedimentary succession above the crystal-line basement has been especially well studied for oil

exploration by seismic reflection techniques and by

wells drilled in the southern part of the Moesian plat-

form.

In the autochthonous and overthrusted areas of the

Moesian platform the seismic refraction line is

expected to cross two major crustal faults, the Cap-

idavaOvidiu Fault (COF) and the Intramoesian Fault

(IMF). Whereas the COF is outcropping in the

Dobrogea area near the Black Sea, sediments and

nappes cover the supposed NW-prolongation of the

fault, as well as the IMF (Figs. 2 and 3; Visarion et al.,

1988; Polonic, 1996; Ellouz et al., 1994; Seghedi,

1998). The COF separates a greenschist basement tothe north from a higher-grade metamorphic basement

to the south (Seghedi, 1998 and references therein).

This crystalline basement extends to the south as far

as the IMF, which again separates two different parts

of the Moesian platform. They are characterized by

different structural orientations and lithologies, with

magmatic intrusions into the western block (Seghedi,

1998 and references therein). The IMF is an active

fault along which many low magnitude earthquakes

have been recorded (Radulescu et al., 1976; Cornea

and Polonic, 1979; Zugravescu and Polonic, 1997).

Mainly from seismic data all these crustal faults are

thought to prolongate down to respectively cross the

Moho discontinuity (e.g. Visarion et al., 1988; Radu-

lescu and Diaconescu, 1998).

Using published (Dumitrescu and Sandulescu,

1970) as well as unpublished (Fielitz, personal com-

munication) geological data, crustal cross-sections for

both refraction lines were constructed and are shown

in Figs. 3 and 4. They form the base for the inter-

pretation of the velocity-depth model obtained from

the seismic experiment (Figs. 9 and 10). Concerning

the nappe and the basement structures, two alternativeinterpretations are possible. In the interpretation of

Ellouz et al. (1994) the basement and the autochtho-

nous cover of the Moesian platform is affected by the

stacking of the Moldavidian nappes (Fig. 3, top and

Fig. 4, left). In the interpretation of Sandulescu (1984)

and Stefanescu and Working Group (1988) shortening

affects only the Moldavidian nappes, i.e. the cover

(Fig. 3, bottom and Fig. 4, right). Both cross-sections

also show the foci of the intermediate-depth earth-

quakes of the Vrancea zone for the last 10 years

(Oncescu et al., 1998) projected onto the profiles.Based on seismic reflection and geological data,

Matenco and Bertotti (2000) tried to estimate the

depth of the different nappes. Using their results,

the predictions for the area containing the seismic

refraction profile can be summarized as follows.

Between shot points B and C the Subcarpathian

nappe is about 6 km thick, while the thickness of

the Marginal Folds nappe around shot point D should

be 7 km. The thickness of the Tarcau nappe (between

shot points E and H) is expected to be about 67 km.



8/24

Near shot point K the Subcarpathian nappe should be

6 km thick, while the top of the Mesozoic platform

cover is expected to be about 5 km deep (Figs. 24).

3. Earlier geophysical investigations

The area crossed by the seismic refraction line

VRANCEA99 (Fig. 1) has already been investigated

by earlier geophysical work. While the flysch area

is only poorly explored, the Moesian platform has

been investigated in more detail by both seismic

reflection and refraction methods. The seismic

reflection investigations have revealed the structure

of the Neogene and Mesozoic sediments and to a

lesser extent the Paleozoic sediments and the crys-

talline basement.

The Neogene cover shows many reflections,

which can be correlated over long distances. One

such regional reflector, which can be observed over

the whole platform, is the Cretaceous erosional relief

unconformably overlain by Neogene sediments (Rai-

leanu, 1998; Raileanu and Diacunescu, 1998; Rai-

leanu et al., 1994). Across this interface a velocity

jump of more than 1 km/s generates strong reflec-

tions. On a short reflection line, about 20 km north-

east of shot point M, the P wave velocities in theNeogene sediments range from 2.0 to 3.5 km/s,

while the Mesozoic and underlying layers have Vp

values between 3.5 km/s and over 6.0 km/s (Raileanu

et al., 1993). For the Mesozoic and Paleozoic layers

with clastic sedimentary rocks the seismic P wave

velocities are less than 4 km/s, while they are higher

for the limestone and dolomite (Raileanu et al.,

1993).

In the Moesian area the reflectivity on seismic

reflection data decreases with depth. The contact of

the sedimentary cover with the crystalline basementdoes not generate reflected waves for near-vertical

angles of incident (Raileanu et al., 1994). In addition,

the crystalline crust is only weakly reflective all the

way down to the Moho. Some multiple waves,

originating in the Neogene or Mesozoic successions,

have been observed and may hide the deeper and

weaker reflected waves (Raileanu et al., 1994). The

base of the crust is marked by the complete disap-

pearance of reflectivity and lies at a depth of 35 38

km according to Raileanu et al. (1993).

Seismic refraction data collected on the eastern

part of the Moesian platform yielded P wave veloc-

ities of 5.9 6.2 km/s for the top of the basement

(Radulescu et al., 1976; Cornea et al., 1981). The baseof the crust was mainly detected by reflected waves at

wide-angle distances and to a lesser extent by head

waves, due to the short length of the seismic spreads.

The depth to Moho is given between 30 and 40 km.

A short (only 20 km long) seismic reflection line

in the Subcarpathians, just west of shot point B (Fig.

2), shows a weakly reflective upper and lower crust.

The main reflectors according to Raileanu et al.

(1994) are the seismic basement at 12 km depth, a

mid-crustal discontinuity at 23 km and the Moho at

43 km depth.

Results from the International Geotraverse XI (Fig.

1) on the Eastern Carpathians give the depth to base-

ment with 11 km. The depth to the middle crust is

given as 28 km, while the Moho discontinuity lies at

45 km depth (Radulescu et al., 1976).

Magnetotelluric data along a profile almost parallel

to the transverse line (shot points R and S in Figs. 1

and 2) have provided a crustal pattern of four layers

with alternating electric conductivity. For the area of

the refraction seismic line the first layer is assigned to

the Marginal Folds nappe with a thickness of 8 km.

The second one to the older sedimentary cover of theunderlying platform at 1416 km depth and the third

one to the base of the crystalline crust at 50 km depth.

The deepest layer was interpreted as upper mantle

(Stanica and Stanica, 1998).

The Bouguer anomaly map of Visarion (1998)

shows maximum negative values for the foredeep

region and the neighbouring flysch nappes. This is

caused by the thick folded and thrusted nappes, the fill

of the foredeep and the thickened crust. The Bouguer

anomaly is negative along the entire refraction line

with lowest values within the internal sector of theforedeep (i.e. from south of shot point G to shot point

L in Figs. 1 and 2). Towards both ends of the

refraction line the Bouguer anomaly increases to 0

mgal in the south and 50 mgal around Bacau in the

north.

Using a large amount of geological and geophys-

ical data Polonic (1996, 1998) compiled a map of the

crystalline basement for Romania. Along the VRAN-

CEA99 refraction line the depth to basement increases

from 6 km near Bucharest to 15 km under shot point



9/24

G. Further north it decreases again to 7 km around

Bacau.

Based on seismic refraction and reflection data, as

well as on gravimetric and other geophysical data,Radulescu (1988) and Enescu et al. (1992) produced

regional maps for the Conrad and Moho disconti-

nuities in Romania. The Conrad discontinuity varies

in depth from 14 km near the southern end of the

VRANCEA99 line to 28 km in the Focsani depres-

sion and 22 km near Bacau. The Moho contours

show about 30 km depth near Bucharest, about 50

km for the Carpathian bend area and about 40 km

near Bacau (Fig. 1). In addition to the above-men-

tioned maps, Enescu et al. (1992) also used travel

time information of some shallow seismic events (0

10 km depth) to derive empirical velocity-depth

functions (P and S waves) for the crust and for the

upper mantle down to 70 km depth. The reported P

wave velocity at the base of the crust is about 6.95

km/s. The mean P wave velocity for the entire crust

including the sediments is 6.18 km/s for the orogenic

area and 6.21 km/s for the Moesian and East Euro-

pean platforms.

4. The VRANCEA99 seismic experiment

In order to achieve the objectives discussed above,

the VRANCEA99 seismic refraction survey was car-

ried out in May 1999. The project consisted of two

lines intersecting each other at the northern edge of

the Vrancea epicentral region (Fig. 1).

The main line is 300 km long and runs from the city

of Bacau through Bucharest and onto the Danube

River, crossing the Vrancea region in a NNESSW

direction. It crosses the bending area of the Eastern

Carpathian Orogen in the Moldavidian zone, the fore-

deep and its eastern and southern foreland (Scythianand Moesian platforms). The line (Figs. 13) starts in

the Subcarpathian nappe, which outcrops between shot

points A and D. Along this segment of the seismic line

the Subcarpathian nappe overlies the Scythian plat-

form to the north of the Trotus Fault (TF) and the

Moesian platform to the south. Until about 5 km north

of shot point E the seismic line cuts across the

Marginal Folds nappe. The Tarcau nappe is crossed

until about 10 km south of shot point G. From shot

sites A to G the direction of the seismic line is almost

parallel to the direction of the folded structures, there-

after it is oblique to the geological structures. The next

segment of the seismic line, between shot points H and

K crosses again the Subcarpathian nappe. The fore-deep segment comprises shot points L and M, where

the basement is already part of the Moesian platform.

The southernmost shot point N south of Bucharest is

fully located on the Moesian Platform outside the

foredeep.

An additional, short refraction line transverse to the

geological structures passes along the Putna Valley,

north of the bend area intersecting with the main

refraction line at shot point D (Figs. 14). Two more

shot points are located on this line, one in the Focsani

foredeep to the east (shot point S) and one in the

Tarcau nappe east of Targu Secuiesc to the west (shot

point R).

Along the two segments a total of 140 recording

sites were occupied with an average station spacing of

2 km. Seismic recording equipment for the experiment

was provided by the GeoForschungsZentrum Potsdam

(Germany), by Leicester University (UK) and by the

NERC geophysical equipment pool (UK).

The spacing of shot points ranged from 12 to 14

km between shot points H K and E G to 29 30 km

between L M and G H, with an average of 22 km.

The charge sizes varied from 300 to 900 kg with thelarger shots being A and M (900 kg each) and B and L

(600 kg each), close to the end points of the main line

(Figs. 1 and 2). All shots were recorded simultane-

ously along the main line and the transverse line. For

more technical details on the project see Prodehl et al.

(2000) and Hauser et al. (2000).

As the 70 km long transverse line running perpen-

dicular to the main line along the Putna Valley (Fig. 2)

did not resolve the deeper levels of the crust, it is not

discussed further in this paper.

5. The seismic sections

The seismic record sections were compiled and

plotted using the SeismicHandler program package of

Stammler (1994). All record sections (Figs. 5 8)

presented in this paper are vertical component data

plotted with a reduction velocity of 6 km/s. Seismo-

grams are trace normalised, which means that the

amplitudes are scaled with respect to the maximum



10/24



11/24

amplitude per trace. A general band pass filter

between 3 and 14 Hz has been applied to the data

in order to improve the signal-to-noise ratio.

For this paper, the term travel time refers tothe reduced travel time of the seismic sections, the

term offset to the distance between source and

receiver, and the term distance to a location

along the refraction model with respect to shot

point A.

The data quality for the experiment is very

variable and seems to depend strongly on local

conditions, like structure and physical properties of

the corresponding shot sites and local receiver

conditions. The maximum offset to which first or

later arrivals can clearly be seen on the seismic

sections may be considered as a measure of data

quality. The northern segment of the refraction

profile generally has a low signal-to-noise ratio.

This may be due to the fact, that most of the shot

points (A K and R S) are located within the

Carpathians where nappes and faults cause strong

scattering and absorption of seismic energy and

therefore reduce the signal-to-noise ratio. Only for

shot point A (Fig. 5) can arrivals be picked over an

offset range of 200 km. Shot points L (Figs. 6 and

7) and M (Fig. 8) on the other hand, which are

located in the Carpathian foreland, have the highestsignal-to-noise ratios.

Not all phases have been identified on all record

sections, sometimes due to the limited offset, but

sometimes also because of the low signal-to-noise

ratio. We have only picked those phases, which are

coherent over several traces. A summary of the

phases observed on each section is given in Table 1.

rUp to four separate first-arrival refractions with appa-

rent velocities < 6 km/s could be identified on the

seismic sections on the basis of apparent velocity and

offset distribution. They have been named P1 to P4.Most of the data also show a clear Pg phase (a diving

wave through the upper crust) as first arrivals between

40 and 100 km offset. This phase is characterised by

strong undulations and sometimes very small ampli-

tudes, making picking generally difficult beyond 80 km

offset.

Within the deeper crust we could identify PcP

(reflection from the top of the lower crust) as asecondary arrival behind Pg up to 200 km offset.

The phase shows an apparent velocity of 6.1 6.3

km/s at furthest offsets. It varies from being prom-

inent and laterally coherent over 100 km offset to

being weak and only visible in the sub-critical

offset range. This could suggest that the mid-crustal

boundary is laterally not continuous along the

profile.

The reflection from the Moho (PmP) is observed

on several record sections, especially from shots at

both ends of the profile. The phase is characterised by

high lateral coherency, very strong amplitudes and has

an apparent velocity of 6.8 km/s at furthest offset.

This would suggest that the Moho is a laterally

continuous, sharp discontinuity or a thin transition

zone.

A diving wave through the upper mantle (Pn) can

only be seen for the southern shot points L (Figs. 6

and 7) and M (Fig. 8). It shows an apparent velocity of

8 km/s and can be picked out to the maximum offset

of over 200 km. For the same two shot points (L and

M) a mantle phase (PLP) can be seen in the data as

well. It is characterised by a high apparent velocity(8.6 km/s) and strong amplitudes (sometimes as large

as or larger than PmP). This phase is coherent

between 100 km offset and the end of the seismic

sections.

6. Interpretation techniques and the velocity model

There are several major steps in the modelling

procedure:

(1) Travel times and associated errors were pickedfor each of the seismic phases described above. The

integrity of the picked travel times and the consistency

of the phase identification were checked by compar-

ing reciprocal travel times where possible.

Fig. 5. (a) Trace normalised P wave record section and (b) synthetic seismograms for shot point A, reduced with 6 km/s. The calculated travel

times from the model in Fig. 9 are plotted on top of the data. Travel times are labelled as follows: P1P4, first-arrival phases refracted within the

sedimentary cover; Pg, diving wave through the upper crust; PcP, reflection from the top of the lower crust; PmP, reflection from the Moho; Pn,

diving wave through the upper mantle; PLP, reflection from the upper mantle.



12/24

Fig. 6. (a) Trace normalised P wave record section and (b) synthetic seismograms for shot point L observed to the north. For further explanations

see Fig. 5.



13/24

Fig. 7. (a) Trace normalised P wave record section and (b) synthetic seismograms for shot point L observed to the south. For further explanations

see Fig. 5.



14/24



15/24

(2) One-dimensional velocity-depth functions were

calculated for each shot point.

(3) The resulting 1-D models were merged into a 2-

D cross-section along the main profile, making due

allowance for the offsets in the different phases. Next

the data were interpreted by 2-D forward modelling,

initially using the structure deduced from the 1-D

models. The ray-tracing program SEIS83 of Cerveny

and Psencik (1983) was used for this step.

(4) Finally, when the velocity model obtained by

forward modelling was good enough to identify theobserved arrivals confidently, a travel time inversion

was carried out, using the method of Zelt and Smith

(1992). The misfit between the picked and calculated

travel times was minimised using a damped least-

squares inversion, which also allows the resolution for

individual model parameters to be quantified.

(5) The relative amplitudes of individual phases

were estimated qualitatively where possible and triedto match by varying the velocity gradient within

individual layers of the model. However, a detailed

trace-by-trace amplitude modelling was not attemp-

ted. Synthetic seismograms were calculated again

using the program SEIS83 (Cerveny and Psencik,

1983). The resulting synthetic seismograms are shown

in Figs. 58 together with the observed data.

6.1. The velocity structure of the crust and upper

mantle

The final 2-D velocity model derived using the

methods described is shown in Fig. 9a. It has a

multi-layered character with a mean velocity for the

whole crust of 6.2 km/s. The main structures crossed

by the seismic line can be separated into three

groups. (1) The sedimentary cover showing veloc-

ities < 6 km/s. (2) The crystalline crust down to

Moho. (3) An upper mantle structure at sub-Moho

level.

The sedimentary succession along the main line

consists of two to four layers with velocities ranging

from 2.0 to 5.8 km/s. The acoustic basement coincideswith a velocity step up to 5.9 km/s. The crustal

velocity is fairly constant lateral direction and in-

creases gradually to 6.2 km/s at the base of the upper

crust. An intra-crustal discontinuity, defined by sec-

ondary PcP reflections is present and divides the crust

into an upper and a lower layer. Velocities within the

lower crust again seem fairly constant in a horizontal

direction and increases vertically from 6.7 to 7.0 km/s

at the Moho.

Strong wide-angle Moho reflections (PmP) indicate

the existence of a first-order crustmantle boundary.The depth to Moho increases from 38 km at the north-

ern end of the profile to 41 km between shot points F

and L. South of L it decreases to 30 km under shot point

N (Fig. 9a). A constraint on upper mantle seismic

Fig. 8. (a) Trace normalised P wave record section and (b) synthetic seismograms for shot point M observed to the north. For further

explanations see Fig. 5.

Table 1

Summary of the phase correlation

P1 P2 P3 P4 Pg PcP PmP Pn

A S XXX XXX XXX XX X XB N XXX XXX XX

B S XXX XX XX XX X

C N XXX XXX XX XX

C S XXX XXX XX X

D N XX XX XX X

D S XXX XXX XX

E N XXX XXX XX X

E S XX XX XX

F N XX XX XX X

F S XX XX

G N XX XX XX X

G S XXX XXX XX X

H N XXX XXX XXX XX XX X X

H S XXX XXX XX XX X X XK N XXX XXX XX XX XX X X

K S XXX XXX XX XX X X X

L N XXX XXX XXX XXX XX XX XX XX

L S XXX XXX XX XX XX XX XXX

M N XXX XXX XXX XX XX XXX XX

M S XXX XX XX XX XXX

N N XXX X X

N S XXX X X

XXX indicates an easy correlation, XX a less clear correlation,

while X a very difficult correlation. A blank field means that this

phase was either not present or could not be correlated.



16/24

velocities (7.98.0 km/s) is provided by Pn arrival

times picked from shot points L (Fig. 6) and M (Fig. 8).

Based on PLP reflections from the same two shot

points, a low-velocity zone with a velocity of 7.6 km/s

was modelled within the upper mantle. The base of

this velocity inversion lies at a depth of 55 km. The

velocity beneath this interface must be at least 8.5 km/

s in order to match the amplitudes and critical dis-



17/24

tances of the PLP reflections as seen on shot points L

(Fig. 6) and M (Fig. 8).

6.2. Model resolution and uncertainties

In a qualitative sense it is obvious that velocities

derived from refracted phases are certainly more reli-

able than those derived from reflected phases. It is also

intuitive that the reliability to the depth of a reflector

increases with the number of rays reflected from aninterface. The advantages of using an inversion algo-

rithm like the one described by Zelt and Smith (1992)

are the ability to construct models that satisfy a large

number of shots and receivers and to assess the model

resolution, uncertainty and non-uniqueness of the

derived model in a more objective way.

The reliability of the model was quantified during

the inversion procedure, where estimates of the reso-

lution and the absolute parameter uncertainties were

calculated. Travel time fits are assessed using the

normalised form of the misfit parameter v2 (Beving-

ton, 1969). In general a value of v2, as close to 1 as

possible, is sought as this indicates that the computed

travel times fit the data to within their assigned

uncertainties. Values much less than 1, indicate that

the data have been overfit. As a result the model will

contain structure not required by the data. Final values

of v2 much greater than 1 generally indicate the data

have sampled small-scale heterogeneities they cannot

resolve (Zelt and Smith, 1992). In practice, final v2

values greater 1 are acceptable if the parameter

resolution is high (see below) and if it is possible to

trace rays to all stations. The root mean square travel

time residual for the whole model is 0.12 s, and the

normalised v2 value is 1.339. The values for individ-

ual phases are shown in Table 2. While the final

model obtained through travel time and amplitude

modelling is shown in Fig. 9a, ray coverage and travel

time fits for all shots and all phases are shown in Fig.

9b and c, respectively.

Fig. 9. (a) Final 2-D velocity-depth model along the main VRANCEA99 line, starting from the city of Bacau in the north, passing through

Bucharest and ending near the river Danube. The profile is traversing the Vrancea epicentral region in a NNE SSW direction. Labelled dots at

the top of the model indicate the shot points, while numbers indicate the P wave velocities in km/s. Thick solid lines indicate areas which are

constrained by reflections and/or refractions. (b) Ray coverage through the final model connecting all shot and receiver pairs. (c) Comparison of

observed and calculated travel times for all shots and all phases. Vertical bars indicate observed data with height representing pick uncertainties.

Solid lines indicate calculated travel times.



18/24

The diagonal elements of the resolution matrix

range between zero and one. While a value close to

1 indicates a well resolved parameter, a low value (i.e.


19/24

It shows some topography, and two steps are obvious

in this layer. The first one occurs between shot points

B and C and has a vertical offset of about 1 km. It is

located just a few kilometres south of the Trotusvalley and can be interpreted as either being the

Trotus Fault or a local fault striking transverse to the

nappe structures and separating two blocks of the

same nappe as a transfer fault. The second step is

located south of shot point G and has a vertical offset

of about 4 km. This coincides with a slight increase in

thickness, but a decrease in velocity. Furthermore,

halfway between shot points G and H this layer

plunges beneath layer 1 and pinches out near shot

point M, south of the Pericarpathian Front. This

second step corresponds to the onset of the foredeep

cover and the pinching out of the layer further south

coincides with the location of the Intramoesian Fault.

Near the fault, layer 2 corresponds to the undeformed

pre-Sarmation beds of the foredeep. These same beds

have been folded and thrusted further north until the

front of the Subcarpathian nappe between shot points

K and L (the Pericarpathian Front in Fig. 2). Velocities

in this layer range from 3.43.9 km/s for the Molda-

vian nappes to 3.33.5 km/s for the subsided part of

the Subcarpathian nappes.

The third seismic layer starts from the northern end

of the model, but pinches out south of shot point M(Fig. 9a). It consists of the Subcarpathian, the Mar-

ginal Folds, and the Tarcau nappes as well as of

foredeep rocks. The P wave velocities in this layer

range from 4.0 to 4.8 km/s. The marked decrease in

velocity under shot point G could indicate a change in

petrological composition towards less consolidated

rocks. The layer is at a constant depth of 4 km at

the northern end of the line, but deepens to 8 km

between shot points G and H. South of shot point H

its depth decreases again, while the velocity increases

until it pinches out.Geological cross-sections for the area (Matenco

and Bertotti, 2000) show the occurrence of Miocene

salt deposits between shot points A and F at 4 km

depth, which could explain the observed velocity of

4.7 km/s at the base of this layer. Therefore, it is

very likely that between those shot points the

seismic energy was reflected from the top of the

salt beds. South of shot point F, where according to

Matenco and Bertotti (2000) the salt beds disappear

due to facies changes in the more inner parts of the

Moldavidian nappes, the reflector deepens and the

velocities decrease. In this part of the seismic

profile the base of the third seismic layer would

correspond to the base of the Moldavidian nappes.The velocity increase south of shot point H can be

interpreted with the reappearance of the salt beds

facies in the again more outer parts of the Molda-

vidian nappes.

The fourth seismic layer is the only sedimentary

layer that can be observed along the entire seismic

refraction profile (Fig. 9a). It probably represents the

autochthonous Palaeozoic, Mesozoic, and maybe, the

very thin Cenozoic sedimentary cover rocks of the

Moesian platform. The velocities within this layer

vary from 4.7 to 5.8 km/s. The greatest thickness of

about 9 km is reached between shot points F and G,

but it decreases towards both ends to 5 km.

The marked lateral velocity changes within this

layer allow a sub-division into several blocks. (1) A

southern block with highest velocities (5.35.8 km/

s). (2) A central block with lowest velocities (4.7

5.2 km/s) between shot points L and G, which also

coincides with the position of a minimum in the

Bouguer anomaly. (3) This is followed by a smaller

high-velocity block (5.35.7 km/s) and (4) a north-

ern block with intermediate velocities of 5.0 5.5

km/s.These observations led us to the conclusion that the

Moesian platform is made up of several blocks with

different lithological compositions. These blocks

could correspond to some kind of Mesozoic and/or

Palaeozoic pre-structuring of the platform. The Intra-

moesian (IMF) and the Capidava Ovidiu (COF)

faults could have played an important role, because

of their proximity to the described block boundaries.

It is possible that the observed lateral velocity

changes across the COF correlate also with known

differences in the basement, which is made up near itstop of greenschist to the north and crystalline rocks to

the south of the fault. This would point to different

geological evolutions of the two blocks. No velocity

changes have been observed for this seismic layer

across the Trotus Fault (TF).

7.2. The crustal structure

The next two seismic layers make up the crystalline

crust, which consists of an upper (layer 5) and a lower



20/24

(layer 6) crustal layer. The crystalline crust shows some

thickness variations, while the lateral velocity structure

along the entire seismic line remains constant. The

upper crustal velocity of 5.9 km/s is low and may beinterpreted as a low-grade metamorphic sedimentary

composition.

The total thickness of the crustexcluding sedi-

mentsis 2830 km for the northern and central part

of the model. Further to the south it decreases to 25 km

under shot point N. The marked decrease in thickness

corresponds to the location of the Intramoesian Fault.

An intra-crustal boundary was recognised from wide-

angle reflections. It separates an upper crust with

velocities of 5.96.2 km/s from a lower crust with

velocities of 6.7 7 km/s.

The upper crust reaches its greatest thickness of 20

km in the centre of the model. Towards both ends it

decreases to 13 km. On the other hand the thickness of

the lower crust gradually increases from 10 km near the

centre of the seismic model towards both ends where it

reaches 12 kmin the south and about 16 km in the north.The fact that the velocity structure of the crust shows

no lateral variations while the thickness does so could

suggest the existence of an originally homogeneous

crust of approximately constant thickness north of the

Intramoesian Fault (IMF) and a thinner crust of com-

parable composition south of it. Later on the crustal

segment north of the IMF was deformed at mid-crustal

level by shortening near the centre of the seismic line.

The overthrusted body must have had the same phys-

ical properties and/or lithological composition as the

undeformed crust, because the present crust has the

same velocity. As the thickened upper crust coincides

with the area of emplacement of the Moldavidian

Fig. 10. Interpreted geological section along the main VRANCEA99 seismic refraction line between Bacau and Bucharest from the 2-D seismic

model of Fig. 9a. Inside the Eastern Carpathians the section follows mostly the trend of the main geological structures, whereas to the south it

becomes transverse to the main geological structures due to the bending of the Carpathians. For location see Fig. 2. Circles represent the foci of

the intermediate depth earthquakes of the Vrancea zone after Oncescu et al. (1998) projected onto this cross-section.



21/24

nappes, it is suggested that this thickening is due to the

same shortening process, with internal thrusting of the

crystalline crust, which could have been favoured by

the lower rheologic strength of the middle crust. TheCapidavaOvidiu and the Intramoesian faults could

bound this thrust sheet, respectively.

The thinned lower crust between the Capidava

Ovidiu and the Intramoesian faults correlates well

with a low velocity layer at the top of the upper

mantle (Figs. 9a and 10). An explanation for this

could be delamination of the lower crust, which

allowed partial melting of the uppermost mantle.

7.3. Moho and upper mantle structure

The wide-angle Moho reflections indicate the exis-

tence of a first-order discontinuity. The depth to Moho

increases from 38 km under shot point A to 41 km

between shot points F and L. Further to the south it

decreases again to 30 km under shot point N. These

values are shallower than shown in previous studies

(Enescu et al., 1992), with the maximum depth shifted

somewhat to the south (Fig. 1). It is also interesting to

note that the deepest part of the Moho lies between the

Intramoesian and the CapidavaOvidiu faults, while

the hypocenters of the intermediate depth earthquakes

project onto the profile slightly to the north of the laterfault zone (Fig. 10).

The uppermost mantle has a velocity of 7.9 km/s,

which suggests a homogeneous lithological composi-

tion. As has been pointed out above we modelled a 4-

km-thick layer below the Moho with a small velocity

gradient, which is underlain by a low velocity zone.

The travel times of the PLP reflections define an inter-

face in the mantle that lies at a depth of 55 km. The

velocity beneath this interface must be at least 8.5 km/s.

The depth interval of the low velocity zone coincides

well with the seismic gap between crustal and inter-mediate depth earthquakes in the Vrancea zone (Fig.

10, Oncescu et al., 1998). Fuchs et al. (1979) proposed

a zone of low strength in the upper mantle in order to

explain this seismic gap. Such a zone of low strength

could show up as a sub-crustal low velocity zone,

which was also observed for the area by other authors

(Lazarescu et al., 1983; Fan et al., 1998).

The low velocity of 7.6 km/s could suggest a

mixture of crustal and upper mantle rocks or a process

of eclogitisation, i.e. a transition from crustal to mantle

rocks. Another interpretation could be a zone of partial

melting in the uppermost part of the mantle. This could

coincide with a delamination of the lower crust as

described above. A sub-crustal low-velocity zone alsoplays a crucial role in the deep lithospheric model

proposed by Chalot-Prat and Girbacea (2000) for the

Eastern Carpathians. These authors suggest that slab

rollback and break-off induced delamination of the

European mantle lithosphere and upwelling of the

asthenosphere into the newly created space.

8. Conclusions

A 300-km-long NNE SSW trending seismic

refraction line was carried out in Romania in order

to study the lithosphere underneath the Vrancea epi-

central region within the SE Carpathians. The inter-

pretation of the data by forward and inverse modelling

gave the following results.

The sedimentary succession is up to 13 km thick

and can be sub-divided into two to four layers. It

comprises the Moldavidian nappes, the Neogene infill

of the foredeep and cover of the Moesian and Scy-

thian platforms, as well as the autochthonous Meso-

zoic and Palaeozoic sedimentary rocks of the Moesian

and Scythian platforms.The underlying crystalline crust shows thickness

variations, but at the same time the lateral velocity

structure along the seismic line remains constant. An

intra-crustal boundary separates an upper crust with

velocities of 5.96.2 km/s from the lower crust with

velocities of 6.77 km/s. Within the upper mantle we

observe a low velocity zone down to a depth of 55 km.

Using the VRANCEA99 seismic refraction experi-

ment in addition to other geophysical data, a crustal

cross-section can be derived (Fig. 10). In the centre of

the model the crystalline crust is thickened andcovered by the Moldavidian flysch nappes and the

Carpathian foredeep up to the Pericarpathian Oro-

genic Front. The deepest segment is situated approx-

imately between shot points L and G under the

Carpathian foreland and is also confirmed by both

Bouguer and isostatic gravity anomalies (Visarion,

1998; Rosca, 1998). It seems that the Neogene tec-

tonic convergence resulted in thin-skinned shortening

of the sedimentary cover and in thick-skinned short-

ening in the deeper part of crust. This would correlate



22/24

with the first alternative interpretation of the geo-

logical cross-sections from Figs. 3 and 4. On the

Moesian platform itself several blocks with different

lithological compositions can be recognised, whichpoints to a Mesozoic and/or Palaeozoic pre-structur-

ing of the platform, where the Intramoesian and the

CapidavaOvidiu faults probably played an important

role, because of their proximity to two of the

described block boundaries. Especially the Intramoe-

sian fault is evident from interpretations of the seismic

refraction data. A reactivation of these crustal faults

during the Neogene shortening is probable. However,

no clear indications of the important Trotus fault in the

north have been found.

A thinned lower crust between the Capidava

Ovidiu and the Intramoesian faults correlates well

with a low velocity layer at the top of the upper

mantle. This could be explained by delamination of

the lower crust, which may have allowed partial

melting of the uppermost mantle.

Acknowledgements

This investigation was only enabled by the

continuous effort of many volunteers. In particular

we thank all additional participants in the field work:G. Danci, M. Diaconescu, A. Hlevca, D. Mateciuc, L.

Munteanu, V. Nacu (NIEP, Bucharest), V. Dumitrescu

and M. Georgescu (Geotec, Bucharest), T. Orban

(University, Bucharest), J. Bribach (Potsdam), P.

Denton and A. Myers (Leicester), G. OBrien (Dublin),

H. Raue (Wurzburg), S. Bourguignon, A. Goertz, P.

Heidinger, C. Jaeger, I. Koglin, H-M. Rumpel and C.

Weidle (Karlsruhe). The National Institute for Earth

Physics (NIEP) provided the logistics for the fieldwork

in Romania. We express our sincere thanks to

Professor D. Enescu and Dr. G. Marmureanu, GeneralDirectors. Dr. Mihaela Rizescu and Mihaela Popa

ensured the recordings of the NIEP network. We are

extremely grateful to Professor C. Dinu and Dr. V.

Mocanu (University of Bucharest) for providing the

facilities at the Geological Institute of the University of

Bucharest. We thank Dr. M. Melinte (Geological

Institute of Romania), Dr. D. Badescu, Dr. L. Matenco,

and Dr. D. Ciulavu (University of Bucharest), who

contributed much with their discussions about the

Romanian geology. Dr. Mihail Ianas, President of the

National Agency for Mineral Resources issued the

general permit for the fieldwork. Col.eng. Neculai

Cioanca, Deputy chief of the Military Topographic

Department, provided permission to obtain maps in thescale 1:100,000, revised and re-published in 1996. The

governmental forestry offices at Casin, Tulnici, Nereju

and Gura Teghii as well as many other institutions and

individuals provided logistical support for suitable

sites. The Romanian exploration company PROSPEC-

TIUNI, Bucharest, carried out the drilling and shooting

operations. Dr. V. Varodin and Eng. M. Milea,

Technical Directors, co-ordinated the whole operation,

from permitting through property access formalities to

the drilling and shooting works. Data were collected

using the seismic equipment of the GeoForschungs-

Zentrum Potsdam (120 units) as well as of the

Department of Geology, Leicester University, UK

(12 units) and of the NERC Geophysical Equipment

Pool, UK (8 units). Dr. J. Bribach (GFZ Potsdam)

trained the German participants before the experiment

proper. Some figures were created using the Generic

Mapping Tools (GMT) software of Wessel and Smith

(1995). The project was funded through the Deutsche

Forschungsgemeinschaft (German Research Society)

by providing the funding for the Sonderforschungsber-

eich 461 (CRC 461) at the University of Karlsruhe,

Germany: Strong Earthquakes a Challenge forGeosciences and Civil Engineering. The NATO

Science Collaborative Research Linkage Grant no.

EST.CLG 974792 assisted the project by additional

funding of travel and living expenses for data

interpretation, and the NATO Computer Network

supplement EST.CNS 976375 supported the labora-

tory of NIEP. We also appreciate helpful comments by

G. Fan and L. Ratschbacher.

References

Airinei, St., 1977. Lithosphere microplates on the Romanian terri-

tory reflected in regional gravimetric anomalies (in Romanian).

St. Cerc., Geol., Geofiz., Geogr., Geofiz. 15 (1), 19 30.

Badescu, D., 1998. Geology of the East Carpathiansan overview.

CERGOP South Carpathians Monogr. 7 (37), 49 69,

Warszawa.

Bevington, P.R., 1969. Data Reduction and Error Analysis for the

Physical Sciences. McGraw-Hill, New York.

Bonjer, K.-P., Oncescu, M.C., Rizescu, M., Driad, L., 1998. A note

on empirical site responses in Bucharest areas. In: Wenzel, F.,

Lungu, D., Novak, O. (Eds.), Vrancea Earthquakes: Tectonics,



23/24

Hazard and Risk Mitigation. Kluwer Academic Publishing, Dor-

drecht, Netherlands, pp. 149162.

Cerveny, V., Psencik, I., 1983. 2-D seismic ray tracing program

package. Charles University, Prague.

Chalot-Prat, F., Girbacea, R., 2000. Partial delamination of conti-nental mantle lithosphere, uplift-related crust-mantle decou-

pling, volcanism and basin formation: a new model for the

Pliocene-Quaternary evolution of the southern East-Carpathians,

Romania. Tectonophysics 327, 83 107.

Channell, J.E.T., Kozur, H.W., 1997. How many oceans? Meliata,

Vardar, and Pindos oceans in Mesozoic Alpine paleogeography.

Geology 25 (2), 183 186.

Ciulavu, D., Dinu, C., Szakacs, A., Dordea, D., 2000. Neogene

kinematics of the Transylvanian basin (Romania). AAPG Bull.

84 (10), 1589 1615.

Constantinescu, L., Enescu, D., 1984. A tentative approach to pos-

sibly explaining the occurrence of Vrancea earthquakes. Rev.

Roum. Geol., Geophys., Geogr., Geophys. 28.

Constantinescu, L., Cornea, S., Lazarescu, D., 1973. An approach to

the seismotectonics of the Romanian Eastern Carpathians. Rev.

Roum. Geol., Geophys., Geogr., Geophys. 17 (2), 133143.

Cornea, S., Polonic, G., 1979. Data concerning the seismicity and

seismotectonics of the Eastern Moesian Platform (in Romanian).

St. Cerc., Geol., Geofiz., Geogr., Geofiz. 17 (2), 167 176.

Cornea, I., Radulescu, F., Pompilian, A., Sova, A., 1981. Deep

seismic soundings in Romania. PAGEOPH 119, 1144 1156.

Csontos, L., 1995. Tertiary tectonic evolution of the Intra-Carpathi-

an area: a review. Acta Vulcanol. 7, 113.

Dercourt, J., Ricou, L.E., Vrielinck, B., 1993. Atlas Tethys, Paleo-

environmental Maps. Gauthier-Villars, Paris, 307 pp.

Dumitrescu, I., Sandulescu, M., 1970. Tectonic map of Romania at

a Scale 1:1000,000. Inst. Geol. Rom., Bucharest.Ellouz, N., Roure, F., Sandulescu, M., Badescu, D., 1994. Balanced

cross-sections in the Eastern Carpathians (Romania): a tool to

quantify Neogene dynamics. In: Roure, F., Ellouz, N., Shein,

V.S., Skvortsov, I. (Eds.), Geodynamic Evolution of Sedimen-

tary Basins. Proceedings Volume, International Symposium in

Moscow, pp. 305325.

Enescu, D., Danchiv, D., Bala, A., 1992. Lithosphere structure in

Romania II. Thickness of the Earth crust. Depth-dependent

propagation velocity curves for the P and S waves. Stud. Cercet.

Geol. Geofiz. Geogr., Geofiz. 30, 319.

Fan, G., Wallace, T.C., Zhao, D., 1998. Tomographic imaging of

deep velocity structure beneath the Eastern and Southern Carpa-

thians, Romania: implications for continental collision. J. Geo-

phys. Res. 103 (B2), 2705 2723.Fuchs, K., Bonjer, K., Bock, G., Cornea, I., Radu, C., Enescu, D.,

Jianu, D., Nourescu, A., Merkler, G., Moldoveanu, T., Tudor-

ache, G., 1979. The Romanian earthquake of March 4, 1977: II.

Aftershock and migration of seismic activity. Tectonophysics 53

(3 4), 225 247.

Girbacea, R., Frisch, W., 1998. Slab in the wrong place: lower

lithospheric mantle delamination in the last stage of the Eastern

Carpathians subduction retreat. Geology 26, 611614.

Hauser, F., Raileanu, V., Prodehl, C., Bala, A., Schulze, A., Denton,

P., 2000. The Seismic-Refraction Project VRANCEA-99, Open-

File Report. Geophysical Institute, University of Karlsruhe.

Kovac, M., Marton, E., Sefara, J., Konecny, V., Lexa, J., 2000.

Miocene development of the Carpathian chain and the Panno-

nian Basin: movement trajectory of lithospheric fragments, sub-

duction and diapiric uprise of asthenospheric mantle. Slovak

Geol. Mag. 6 (23), 7784.Lazarescu, V., Cornea, I., Radulescu, F., Popescu, M., 1983. Moho

surface and recent crustal movements in Romania; geodynamic

connections. Annu. Inst. Geol. Geofiz., 8391.

Linzer, H.G., 1996. Kinematics of retreating subduction along the

Carpathian arc, Romania. Geology 24, 167 170.

Linzer, H.-G., Frisch, W., Zweigel, P., Girbacea, R., Hann, H.-P.,

Moser, F., 1998. Kinematic evolution of the Romanian Carpa-

thians. Tectonophysics 297, 133 156.

Matenco, L., 1997. Tectonic evolution of the outer Romanian Car-

pathians. PhD thesis, Vrije Universiteit, Amsterdam, 160 pp.

Matenco, L., Bertotti, G., 2000. Tertiary tectonic evolution of the

external East Carpathians (Romania). Tectonophysics 316,

255286.

Mellors, R.J., Pavlis, G.L., Hamburger, M.W., Al-Shukri, H.J.,

Lukk, A.A., 1995. Evidence for a high-velocity slab associated

with the Hindu Kush seismic zone. J. Geophys. Res. 100,

40674078.

Morley, C.K., 1996. Models for relative motion of crustal blocks

within the Carpathian region, based on restorations of the outer

Carpathian thrust sheets. Tectonics 15, 885904.

Nemcok, M., Pospisil, L., Lexa, J., Donelick, R.A., 1998. Tertiary

subduction and slab break-off model of the Carpathian Panno-

nian region. Tectonophysics 295, 307340.

Neugebauer, J., Greiner, B., Appel, E., 2001. Kinematics of the

Alpine-West Carpathian orogen and palaeogeographic implica-

tions. J. Geol. Soc. Lond. 158, 97110.

Oncescu, M.C., 1984. A three-dimensional distribution of the seis-mic wave velocities under the Carpathians region. PhD thesis,

Institute of Atomic Physics, Bucharest.

Oncescu, M.C., Bonjer, K.-P., Rizescu, M., 1998. Weak and strong

ground motion of intermediate depth earthquakes from the Vran-

cea region. In: Wenzel, F., Lungu, D., Novak, O. (Eds.), Vrancea

Earthquakes: Tectonics, Hazard and Risk Mitigation. Kluwer

Academic Publishing, Dordrecht, Netherlands, pp. 27 42.

Polonic, G., 1996. Structure of the crystalline basement in Romania.

Rev. Roum. Geophys. 40, 5770.

Polonic, G., 1998. The structure and morphology of the crystalline

basement in Romania. CERGOP South Carpathians Monogr.

7 (37), 127131, Warszawa.

Prodehl, C., Raileanu, V., Hauser, F., Bala, A., Rumpel, H.-M.,

Schulze, A., 2000. EUROPROBE PanCarDi project: the seis-mic-refraction project VRANCEA-99. EUROPROBE News 13,

1518.

Radulescu, F., 1988. Seismic models of the crustal structure in

Romania. Rev. Roum. Geol. Geophys. Geogr. Ser. Geophys.

32, 13 17.

Radulescu, F., Diaconescu, M., 1998. Deep seismic data in Roma-

nia. CERGOP South Carpathians Monogr. 7 (37), 177 192,

Warszawa.

Radulescu, D.P., Sandulescu, M., 1973. The plate-tectonics concept

and the geological structure of the Carpathians. Tectonophysics

16, 155161.



24/24

Radulescu, D.P., Cornea, I., Sandulescu, M., Constantinescu, P.,

Radulescu, F., Pompilian, A., 1976. Structure de la croute ter-

restre en Roumanie, Essai dinterpretation des etudes sismiques

profondes. Ann. Inst. Geol. Geofiz., L 50, 5 36.

Raileanu, V., 1998. Studiul unor parametri fizici ai litosferei in unelezone din Romania. PhD Thesis, Inst. Fizica Atomica, Bucharest,

232 pp.

Raileanu, V., Diaconescu, C., 1998. Seismic signature in Romanian

crust. Tectonophysics 288, 127136.

Raileanu, V., Talos, D., Vorodin, V., Stiopol, D., 1993. Crustal

seismic reflection profiling in Romania on the UrziceniMizil

line. Tectonophysics 223, 401409.

Raileanu, V., Diaconescu, C., Radulescu, F., 1994. Characteristics of

Romanian lithosphere from deep seismic reflection profiling.

Tectonophysics 239, 165185.

Roman, C., 1970. Seismicity in Romaniaevidence for the sin-

kung lithosphere. Nature 2, 1176 1211.

Rosca, V., 1998. The isostatic anomaly maps. CERGOP South

Carpathians Monogr. 7 (37), 139155, Warszawa.

Royden, L.H., 1988. Late Cenozoic tectonics of the Pannonian

basin system. In: Royden, L.H., Horvath, F. (Eds.), The Panno-

nian Basin, a Study in Basin Evolution. AAPG Memoir, vol. 45,

pp. 27 48.

Sandulescu, M., 1984. Geotectonics of Romania. Technical Publish-

ing House, Bucharest, 336 pp. (in Romanian).

Sandulescu, M., 1988. Cenozoic tectonic history of the Carpathi-

ans. In: Royden, L.H., Horvath, F. (Eds.), The Pannonian

Basin, a Study in Basin Evolution. AAPG Memoir, vol. 45,

pp. 17 25.

Seghedi, A., 1998. The Romanian Carpathian foreland. CERGOP

South Carpathians Monogr. 7 (37), 2148, Warszawa.

Seghedi, I., Balintoni, I., Szakacs, A., 1998. Interplay of tectonicsand neogene post-collisional magmatism in the intracarpathian

region. Lithos 45, 483497.

Stammler, K., 1994. SeismicHandler programmable multichannel

data handler for interactive and automatic processing of seismo-

logical data. Comput. Geosci. 19 (2), 135140.

Stampfli, G.M., Mosar, J., Marquer, D., Marchant, R., Baudin, T.,

Borel, G., 1998. Subduction and obduction processes in the

Swiss Alps. Tectonophysics 296, 159 204.

Stanica, D., Stanica, M., 1998. 2D modelling of the geoelectric

structure in the area of the deep-focus Vrancea earthquakes.


Warszawa.

Stefanescu, M., Working Group, 1988. Geological Cross-sections at

Scale 1:200,000 A9 A14. Inst. Geol. Geofiz., Bucharest.

Taboada, A., Rivera, L.A., Fuenzalida, A., Cisternas, A., Philip, H.,

Bijwaard, H., Olaya, J., Rivera, C., 2000. Geodynamics of the

northern Andes: subductions and intracontinental deformation(Colombia). Tectonics 19 (5), 787 813.

Visarion, M., 1998. Gravity anomalies on the Romanian territory.


Warszawa.

Visarion, M., Sandulescu, M., Stanica, D., Veliciu, S., 1988. Con-

tributions a la connaissance de la structure profonde de la plat-

forme Moesienne en Roumanie. St. Tehn. Econ., Ser. Geogiz., D

15, 211222.

Wenzel, F., 1997. Strong earthquakes: a challenge for geosciences

and civil engineeringa new collaborative research center in

Germany. Seismol. Res. Lett. 68, 438443.

Wenzel, F., Lungu, D., 2000. Earthquake risk mitigation in Roma-

nia. Proceedings Volume, 2nd EuroConference on Global

Change and Catastrophe Risk Management, Luxembourg.

Wenzel, F., Lungu, D., Novak, O. (Eds.), 1998a. Vrancea Earth-

quakes: Tectonics, Hazard and Risk Mitigation. Selected papers

of the First International Workshop on Vrancea Earthquakes,

Bucharest, November 1 4, 1997. Kluwer Academic Publishing,

Dordrecht, Netherlands, 374 pp.

Wenzel, F., Achauer, U., Enescu, D., Kissling, E., Russo, R., Mo-

canu, V. and Mussachio, G., 1998b. The final stage of plate

detachment; International tomographic experiment in Romania

aims to a high-resolution snapshot of this process. EOS, 79: 589,

592594.

Wenzel, F., Oncescu, M.C., Baur, M., Fiedrich, F., Ionescu, C.,

1998c. 25 seconds for Bucharest. In: EWC98 Group (Eds.),

Proceedings Volume of the International IDNDR Conferenceon Early Warning Systems for the Reduction of Natural Disas-

ters, Potsdam, Sept. 7 11, GFZ Potsdam, Germany.

Wessel, P., Smith, W.H.F., 1995. Free software helps map and display

data. EOS Transactions, Am. Geophys. Union 72, 445446.

Zelt, C.A., Smith, R.B., 1992. Seismic traveltime inversion for 2-D

crustal velocity structure. Geophys. J. Int. 108, 1634.

Zugravescu, D., Polonic, G., 1997. Geodynamic compartments and

present-day stress on the Romanian territory. Rev. Roum. Geo-

phys. 41, 3 24.

Zweigel, P., Ratschbacher, L., Frisch, W., 1998. Kinematics of an

arcuate fold-thrust belt: the southern East Carpathians (Roma-

nia). Tectonophysics 297, 177 207.


vrancea_99_hauser.pdf

Documents