You are on page 1of 18

CRUSTAL MODELS IN ROMANIA II.

MOLDAVIAN PLATFORM AND ADJACENT AREAS


V. RAILEANU, D. TATARU, B. GRECU and A. BALA
National Institute for Earth Physics, P.O. Box MG-2, RO-077125 Bucharest-Magurele, Romania,
E-mail: raivic@infp.ro

Abstract. A new crustal model for the central and north-eastern part of Romania is relying on
the largest database available at this moment. In comparison with previous crustal models the new
model takes into account all known data: old and new seismic refraction data, deep seismic reflection
data and seismology data recorded by the broadband stations belonging to the Romanian seismic
network for monitoring of earthquakes. The new crustal model reveals the topography of the Moho
and top of the lower crust discontinuities as well as the areal distribution of the mean P-wave seismic
velocities across the Moldavian and Scythians platform and partially for the Eastern Carpathians. The
new configuration of the Moho topography map brings some details in the central study area.
Key words: upper and lower crust, Moho, top of lower crust, mean P-wave seismic velocity,
Moldavian and Scythian platforms, Eastern Carpathians orogen.

1. INTRODUCTION
This paper continues a previous work on the crustal models in Romania [1]
this time being focused on the Moldavian and Scythian platforms and on the
Eastern Carpathians orogen and its foredeep.
Along the last decades more attempts to build up crustal models across this
area have been done. A first map for the base of the crust (Moho) for the whole
country based on gravimetry data were achieved in the 1960s [2]. Deep seismic
investigations performed in the 1970s [3] allowed a better estimation of the crustal
thickness. Based on deep seismic refraction data a couple of maps for the Moho
and Conrad (top of the lower crust) discontinuities were published in 1988 [4]. A
synthesis of the seismic and earthquake data resulted in an updated version of the
Moho map of Romania in 1992 [5]. At the beginning of the 2000s an other Moho
map [6] is sketched for the south-eastern half of Romania using previous crustal
data and recent data provided by the Vrancea 99 and Vrancea 2001 seismic
refraction experiments [7, 8] and CALIXTO experiment [9]. A rougher map (1x1
degree resolution) of the crustal thickness [10] is relied on the surface wave
tomography in SE Carpathians. Some crustal models at the European scale released

in the last years are mainly based on the same data used by [6]: Eu-CRUST-07
[11], ESC Moho [12] and EPCrust [13]. The last one [13] is the most recent and it
was used for comparison with our model. At the local scale a Moho model based
on local earthquakes and seismic refraction data is build up for the south-eastern of
Romania [14].
It should be mentioned the recently above crustal models for the Romanian
territory are based on seismic refraction and earthquake data collected in the last
decades, but before 2010. Those models covered some areas of country with no
measured data (e.g. mountain areas) where only extrapolated data was used. Our
crustal model presented in the next pages is a local model based on a larger
database including both old and new refraction data and in addition to previous
models the reflection and earthquake data recorded in the Eastern Carpathians and
Moldavian platform. This is why it is the most complete model performed until
now it bringing new and unknown details.

2. GEOLOGICAL AND GEOPHYSICAL SETTINGS


The study area covers about a rectangle zone from 25.00E to 28.50E and from
45.4 N to 47.80N, Fig. 1, and it is overlapped with the Moldavian and Scythian
platforms and partially with Eastern Carpathians orogen and Focsani basin in the
east and south.
Moldavian platform is a fragment of the cratonic area of the East European
Platform extended over more than a half of the European continent. The age of its
basement is older than the Upper Rifean or at least the Vendian times and with a
consolidation in the Carelian or Gothians times [15]. Moldavian platform is
located at west and south-west of Ukrainian Shield. Its western border is not clear
delimited from the Scythian Platform, Fig.1. The contact of two platforms
(Moldavian and Scythian) having different ages is covered by a thick cover of
Palaezoic and Mesozoic sediment stack or by the Eastern Carpathian nappes.
Geological and geophysical data showed the platform basement is of the Ukrainian
Shield type from the east to at least the Solca fault, Fig.1 [16]. Towards south the
Moldavian platform is separated by two major faults: Vaslui and Bistrita, the first
being the northern margin of a lowered step of the paltform and the second the
contact with Barlad depression (part of the Scythian platform). Both faults are
extended towards south-east up to the continental platform of the Black Sea.
The Moldavian platform basement is folded and metamorphosed with
gneissic rocks and granito-gneissic migmatites and pierced by granitic intrusions.
The base of sedimentary cover is of detrital type and starts with Vendinian rocks
underlying several sedimentary cicles: Cambrian, Ordovician-Silurian, Devonian,
Upper Jurassic-Eocretacious, Upper Cretacious and locally with Palaeogen and
Neogene piles[16]. The thickness of sediments increases from the east to west,
0

from <1 km in north-east to 3-4 km in the neighbourhood of the Solca and Bistrita
faults [17].
The geological and geophyiscal data suggest a ruptural structure of the
basement cut by two fracture systems: one on a north-south or north-west southeast direction with old fractures along which the platform is stepwise descending
towards the west begining probably from the Palaezoic times and the second on the
east-west or east-south-east west-north-west directions which crosses the first
system and plays an important role in underthrusting of the platform under the
Eastern Carpathians orogen. The marginal areas of the platform display the highest
levels of fracturing [16].
Scythian platform is bordered by the East European platform to the north
and the Alpine range (the Great Caucasus, Southern Crimea and North Dobrogea)
in the Black Sea region to the south. The Scythian platform is a continental block
running from the north-west south-east to the west-east and it is located in
between the Bistrita and Trotus faults, Fig.1 [15]. On the Romanian territory it is
overalpped with the Predobrogean depression which includes the Barlad depression
as well. At west and north-west the Scythian platform is represented by a crustal
block delimited by the Solca and Campulung-Bicaz faults, Fig. 1. It is an uplifted
block in comparison with the adjacent blocks to the east and west. Internal
structure is less known due to the thick stack of the Tertiary sediments of the
Barlad depression and the Carpthians nappes which mask it. The sediment
thickness increases from 3 km in the Barlad depression to 10 km on the block
placed in between the Solca and Campulung-Bicaz faults [17]. The basement of
Predobrogean depression had an active tectonics during the Hercynian Orogeny
resulting in a structure with narrow compression grabens [15].
To the west of Campulung-Bicaz fault a rift (aulacogen) zone is present with
thick piles of Palaeozoic, Mesozoic and even Tertiary sediments of the platform
cover. It marks the south-eastern extension under the Carpathians nappes of the
Miechow depression which belongs to the Central European platform. The
Miechow, Barlad and Predobrogean depressions are Permian-Mesozoic structures
overlapped with the Dano-Polono-Dobrogean basement. It is assumed the East
European platform overrides the Central Europe platform [15].
The Eastern Carpathians orogen is a complex structure of folded and thrust
nappes which override each other from the west to east and all together cover the
foreland platforms (Moesian and Scythian), Fig.1, [15]. The main tectonic units
from the west to east are the followings: Neogene Vulcanic zone, Middle Dacides,
Outer Dacides, Moldavides and foredeep [15]. Its outer part on the Romanian
territory is marked by the foredeep margin. The foredeep covers the platforms
located in the front of Carpathians [15].

Fig.1 - Layout of the Moldavian and Scythian platforms and the Eastern Carpathians orogen with the
main fault systems (grey lines) and location of the data points (full triangles) used for crustal
mapping. Dotted lines follow the Pericarpathians line in the east and the western limit of Carpathians
orogen in the west. Tectonic information is compiled from the Tectonic map of Romania [15].

3. CRUSTAL MODEL OF THE STUDY AREA


3.1 Data base
The main classes of data used in this study belong to the deep seismic
refraction and reflection lines which are mainly located on the western flank of the

studied area and the earthquake data collected at the local stations of the Romanian
seismic network.
Deep seismic data were recorded on seismic lines with lengths of a few
hundred km for refraction studies and 20-30 km for reflection studies and resulted
in 2D seismic sections. For mapping of crustal parameters such as the main
discontinuities (Moho or top of lower crust) or mean P-wave velocities in the upper
and lower crust a sampling of seismic sections in 1D crustal models is needed. The
low density of the seismic lines across the study area imposed us a rate of sampling
in 1D models for each 30-50 km distance along the seismic sections. This was
successfully applied for the refraction lines, where the chosen points are coincident
with the shot points. The shorter length of reflection lines (about 20 km) resulted in
only one 1D crustal model for each line. Using two different techniques for data
processing (receiver functions and surface wave inversions) for earthquake and
noise data recorded by the broadband seismic stations several 1D crustal models
were built up.
1D crustal model comprises the geographic coordinates and the depth to the
top of the crystalline upper crust (basement), top of lower crust, base of crust
(Moho) and the mean P-wave seismic velocity in the upper and lower crust. Each
crustal layer is defined by a range of P-wave velocities (Vp): from 5.8 km/s to 6.7
km/s for the upper crystalline crust and from 6.7 km/s to 7.5 km/s for the lower
crust, in agreement with [18, 19, 20].
Seismic refraction data are provided by the following seismic lines:
International Geotraverse XI Galati-Oradea [21], International Geotraverse VII
Vinitza-Iasi-Vrancioaia-North Bulgaria [22], Bicaz-Roman [23], Vrancea 99 [7]
and Vrancea 2001 [8]. On the map (Fig.1) are displayed only the position of points
where 1D models were extracted: Xi-ga, XI-Foc, XI-Nar, XI-TgS, VII-Ias, BcRoPN, A99, B99, C99, D99, E99, F99, R01, S01, T01, U01, W01 and X01. The
refraction data provides information on the geometry of the main crustal
discontinuities (basement, top of lower crust and Moho) for all seismic lines and on
the P-wave seismic velocity (Vp) across the upper and lower crust for the newer
seismic lines only (Vrancea 99 and Vrancea 2001). If for the older refraction lines
no uncertainty is specified for the depths and Vp, the Vrancea 99 and Vrancea
2001 provide uncertainties as follows: depths from 0.5-1.0 km for the shallower
layer (top of basement) to 1.0-1.5 km for the deeper interfaces (e.g. Moho) while
the Vp is in the uncertainty range of 0.1 km/s for the upper and middle crust and
0.15-0.20 km/s for the lower crust and upper mantle. Where the upper or lower
crust is made of two or more layers a weighted average of P-wave velocities for the
individual layers was assigned to the major crustal layer.
Seismic reflection data has extended the database with useful crustal
information for mapping of the geometry of the main discontinuities. Instead they
could not provide valid information on the Vp in the crystalline crust. Two seismic

lines were available for this study: East Comanesti (Com [24] on Fig.1) and
Falticeni (Fal [25, 26]).

Fig.2 - Velocity models: two way travel times (2T) versus depth (H) derived from 1D models for the
A and B shot-points of the Vrancea 99 seismic refraction line.

On the reflection section we could correlate the top of crystalline basement,


top of the lower crust and Moho in terms of two-way-travel times. In order to
transform the position of the main interfaces from travel-time in terms of depth we
used the Vp models for the A99 and B99 points of the Vrancea 99 seismic line,
Fig.2 [10]. The accuracy of depth in this case is at least 1.0-1.5 km. We assigned
one 1D model for each reflection line at Com and Fal (see Fig.1), reported to the
middle point of seismic line.
The earthquake data completed crustal information in the areas without other
previous knowledge on the deep structure, like Moldavian Platform and northern
part of the Eastern Carpathians and was redundant information in the Vrancea area
besides the seismic refraction data. Data processing and interpretation relied on the
receiver functions technique (RF) and surface wave inversion (SuW). RF analyses
the teleseismic events occured in a certain epicentral window [27]. A travel time
difference between the P-wave and S-wave converted at the Moho discontinuity
gives a measure of the crustal thickness and of the P- to S-wave velocity ratio
across the crust. Finally, an inversion technique is applied and a 1D crustal/upper
mantle model is obtained. The SuW technique is based on the analysis of the noise
records from two distant stations [28, 29]. The larger distances between seismic
stations favour the increase of the investigation depth. By cross-correlating the
seismic noise at two stations a Rayleigh Greens function is generated for the
medium between the two stations. Finally a crustal model is derived by an
inversion technique applied to the dispersion curve of the Greens function. The
resulted 1D model reports some mean crustal parameters between the two stations.
Such a model is assigned to a point at the half distance between the stations: BUR-

IAS, BUR-KIS and KIS-VRI, Fig.1 and Fig.3. An improved result is obtained by
applying a simultaneous inversion of the RF and SuW data sets. This is the case for
the BUR33, TESR, IAS and LEOM broadband stations, Fig.1 and Fig.4.

Fig.3 - P-wave seismic velocity vs. depth for the pair of stations: BUR-IAS, BUR-KIS and KIS-VRI
derived from the inversion of the dispersion curves; for locations see Fig. 1.

This study used data recorded by the permanent stations of the Romanian
seismic network: BUR33, TESR, IAS, LEOM, VRI, KIS as well as the mobile
stations deployed during the CALIXTO experiment [9]: A02, A04, A06, C03, E02,
E03, E05, E13, E17, E18, E21, F01, F02, F04, F05, F06, F10, S03, S07.
About 45 data points have been provided by the deep seismic
reflection/refraction investigations and by earthquake data for the mapping of
crustal data, Fig.1.
We focused on the mapping of the two major discontinuities: top of the lower
crust and Moho as well as the distribution of mean P-wave seismic velocities in the
upper and lower crust. The mapping of the top of crystalline basement has been
discarded because the existing map of the crystalline basement [17] is based on a
larger database (seismic prospecting and well data) in respect with our poorer
database used in this study.
3.2. Building up of the crustal model
For each data point a 1D crustal model with depths to the top of basement,
the top of lower crust and Moho, and the mean P-wave seismic velocities in the
upper crystalline and lower crust is assigned. The layout map of Fig.1 points out an
unequal distribution of data points across the target area: a better coverage in the
south and south-east, a poorer coverage in the north and a lack of data in the northeastern corner.

Fig.4 - P-wave seismic velocity vs. depth for the stations BUR, TESR, IAS and LEOM derived from
simultaneous inversion of the RF and SuW data sets; for location see Fig.1.

In order to build up the maps with crustal parameters a two steps procedure
was applied: a linear Kriging method for interpolation [30, 31] and a filtering of the
resulted grid by a spatially Gaussian filter with 50 km wide [32]. For interpolation
we used the Surfer 10 Golden Software program (www.GoldenSoftware.com) that
offers more interpolation technics: kriging, minimum curvature, polynomial
regression, radial basis function, triangulation with linear interpolation, moving
average, etc. Some of technics interpolate data sets more effectively than others as
a function of number of observation data. For data sets with <250 observations the
Kriging technics with linear variogram is recommended as producing a good
representation of data. For larger data sets other technics such as: triangulation with
linear interpolation or minimum curvature are pointed. We have chosen the kriging
method because our data set comprises about 45 observations. A mean distance of
data observations reported to the study area is about 50 km. Taking into account
the unequal distribution of data and the large gradients between some close data
points a smoothing processing was needed. For smoothing we used the GMT
software package [32] which filtering a 2-D gridded file in the space domain. Out
of the options of this software we choice the Gaussian function filtering with 50 km
width. It weights the data values by a Gaussian function. This filter is based on
convolution methods and its action is a lowpass frequency filtering, removing the
high spatial frequencies from a map. The Gaussian filter shows advantage in
comparison with other filters (e.g. median filter) because it is fairly confident on
the range of spatial frequencies which are still present in map after filtering. For
more information see http://homepages.inf.ed.ac.uk/rbf/HIPR2/gsmooth.htm.

Two maps with the depths to the top of lower crust and Moho, one map with
the thickness of the lower crust and two maps with the mean P-wave seismic
velocity in the upper crystalline crust and lower crust across of the study areas were
achieved. They are displayed in the Figs.5-9.
3.3. Discussions
The five maps illustrate the crustal features of the study area which comprises
both platform and orogen structures. As expected the crustal maps reflect the
peculiarities of each tectonic unit that differ both in age/evolution and structure.
The map at the top of the lower crust shows variable depth across of the
study area, from 20 km within the Piatra Neamt-Moinesti area to almost 30 km in
the south near Nehoiu, Fig. 5. The depths to the top of lower crust decrease towards
north-west and increase to east and south where Focsani depression is located. Our
data shows in comparison with the EPcrust [13] more details and different degrees
of agreement. The differences between the two models are due to the data base
used: a more complete data base for our model and a poorer one for the EPCrust. A
shallower top of the lower crust seems to follow the eastern flank of the Scythian
platform under the Eastern Carpathians (20-22 km depth) while the centre of the
Barlad depression is deeper (~25 km). The Moldavian platform shows depths down
to 25 km in its southern part and 22.5 km on its eastern margin.
The Moho map, Fig. 6, points out crustal thicknesses from 30 km in the
eastern-central part to 37.5-42.5 km in the south and south-east, and 40 km toward
east. Two uplifted crustal blocks within the Miercurea-Ciuc Sf.Gheorghe area
with 30-35 km thickness and the Bacau-Roman area with about 37.5 km can be
noticed. They point out a thinned crust like on the Vrancea 2001 seismic section
[8] for the Miercurea-Ciuc Sf.Gheorghe area or on receiver function data (TSR
station and [9]) in the Bacau-Roman area.
The first thinned crust area could be connected to the Volcanic Neogene zone
of the Eastern Carpathians where the base of crust has suffered an erosion process
during the magmatic activity. The second seems to be located within the central
segment of the Scythian platform and at the western margin of the Moldavian
platform. The thickness of the Scythian platform does not exceed 40 km while the
Moldavian platform is thicker than 40 km. A thicker crust (>40 km) is observed for
the Focsani basin and the North Dobrogea orogen as well. Our model is closer to
the model of [14] than to the [13] model. As we showed the uncertainties of depths
to the Moho are at least 1.5 km and they could be as much as 2.0 km where a low
density of data was used. The observed difference from our model to [14] is as
much as the uncertainties.
The relatively different shape of the isobaths is due to the additional data
used by us in comparison with [14]. The same argument is valid for the differences
between our model and [13] the latter using a poorer data base than us and [14].

Fig.5 - Map with the depth (km) at the top of lower crust from our data with black line and from [13]
with dashed black line for comparison. Grey lines mark the crustal faults and the dotted line the
Pericarpathians fault [15]. CL-BiF - Campulung-Bicaz fault, Sol-F Solca fault, Si-F Siret fault,
VaF Vaslui fault, BiF Bistrita fault, TrF Trotus fault, NTrF the New Trotus fault, Pe-CaF
Pecceneaga-Camena fault.

An isopach map of the lower crust is displayed on the Fig. 7. It points out
thicknesses from ~7.5 km in the south-east to 20 km in the central part of the map.
The thinnest lower crustal layer is located on the Tisza-Dacia crustal block [8] and
supports the hypothesis of the lower crust delamination [33]. The thickest crustal
block is on a direction from Moinesti Piatra-Neamt - Campulung Moldovenesc
coincident to the Scythian platform. A thinner lower crust (~15 km) is within the

eastern-central part on a direction from Harlau to Vaslui and Focsani while a


thickening trend towards east is noticed.

Fig.6. Map with the depths (km) to the Moho discontinuity. Thick black line Moho derived from
our data, dotted line Moho from [14] and dashed line Moho derived by EPCrust model [13].
Crustal faults are drawn with full grey line. For the name of faults see caption of the Fig. 5.

Two mean P-wave seismic velocity maps in the upper crystalline crust and
lower crust are displayed in the Fig.8, 9. The uncertainties of P-wave seismic
velocities derived from refraction data are in the range of 0.1 km/s for the upper
and middle crust and 0.15-0.20 km/s for the lower crust and upper mantle [8]. The
crustal model derived from earthquake data is affected by uncertainties as much as
or greater than the refraction data.

The map in the Fig.8 shows a range of the mean P-wave velocity from 6.00
km/s in the north-west corner and in the central-southern part to 6.25 km/s in the
central-northern part. Even if the relatively low variation (6.00-6.25 km/s) is a little
larger than the uncertainties of data (0.15-0.20 km/s), it suggests the trend of the
mean P-wave seismic velocity, taking into account the seismic velocity is a
function of mineralogy and tectonothermal influence of the Alpine orogeny [24].
Our data are measured data and not exptra- or interpolated and thus gives us a
justified confidence.

Fig.7. Map with the thickness (km) of the lower crust, black lines. Faults are represented with grey
lines. For additional information see the caption of Fig.5.

Fig.8. Mean P-wave seismic velocity (thick lines) in the upper crystalline crust. Black dotted lines are
iso-velocities compiled for comparison from [13]. Grey dashed/dotted lines are crustal faults.

The lowest values are coincident with the Focsani basin and the northern part
of the Eastern Carpathians, both units being younger and/or tectonothermal worked
areas [24] in comparison to the Moldavian platform (an older tectonic unit) where
some larger values are noticed. In comparison with EPcrust model [13] our models
shows about the same velocity range but a more detailed shape of isovelocity lines
contouring two minimum areas in north-west and central-southern.
The mean P-wave seismic velocities in the lower crust are displayed in Fig.9.
They are in a narrow range from 6.85 km/s to 6.95 km/s. The highest values of 6.95
km/s or more are on a north-west south-east direction at the western and southern
margins of the Moldavian platform in between the Siret and Solca fauts and in
between the Vaslui and Bistrita faults. Relatively lower values ( 6.85 km/s) are on
the both flanks of the highest velocity zone they overlapping the Carpathians

orogen and its foredeep in west and the Moldavian platform in east. The EPcrust
model [13] based on extrapolated data points out some lower velocity with an
average of 6.85 km/s. The latter model did not benefit of the earthquake data like
our model and it gives a much more simplified image of the Vp distribution.

Fig.9. Mean P-wave seismic velocity (black lines) in the lower crust.
Thin lines are iso-velocities compiled for comparison from [13]. Grey dashed line are crustal faults.

4. CONCLUSION
The Moldavian and Scythian platforms have been less investigated by deep
seismic sounding in comparison with the adjacent areas (e.g. Vrancea region). New
information derived from the earthquake data recorded by the broadband stations
located in this area is now available. A synthesis of the old data provided by
seismic refraction studies released in the 1970s along with the new data (seismic
refraction/reflection and earthquake data) allowed us to build up an improved
crustal model. The main refraction data were supplied by the several seismic
refraction lines: Geotraverse XI Galati-Oradea, Vrancea 99 Bacau-Vrancea zoneGiurgiu and Vrancea 2001 Tulcea-Vrancea zone-Medias. Two short deep seismic
reflection lines: East Comanesti and Falticeni have also been used for the new
crustal model. The earthquake data were provided by the National Institute for
Earth Physics network: BUR33, TESR, IAS, VRI or outside of it: LEOM and KIS
(Republic of Moldavia) and by the mobile stations deployed during the CALIXTO
experiment [9]: A02, A04, A06, C03, E02, E03, E05, E13, E17, E18, E21, F01,
F02, F04, F05, F06, F10, S03, S07.
About 45 data points with 1D crustal models were used for mapping of
topography of the (1) top of lower crust, (2) Moho, (3) thickness of the lower crust
and the mean P-wave seismic velocities in the (4) upper crystalline crust and (5)
lower crust. A 1D crustal model derived from seismic data comprises: depths to the
top of crystalline basement, top of lower crust and Moho, and the mean P-wave
seismic velocity in the upper crystalline crust and lower crust. For interpolation of
data and smoothing of results we used a linear Kriging method and a filtering of
the resulted grid by a spatially Gaussian filter with 50 km wide.
The five maps bring new details on the crustal structure in comparison with
the precedent models of the investigated area due to the use of new data (mainly
earthquake data) in the areas with a low coverage or lack of crustal data.
The map to the top of lower crust illustrates a variable depth across region:
20 km within the Piatra Neamt-Moinesti area, to almost 30 km in south around the
Nehoiu village. A trend of decreasing of depth towards north-west and an
increasing within the Focsani basin is remarked. A shallower top of the lower crust
is found along the eastern flank of the Scythian platform under the Eastern
Carpathians (20-22 km depth). The centre of the Barlad depression is deeper (~25
km) and the Moldavian platform shows depths down to 25 km in its southern part
and 22.5 km on its eastern margin.
Moho topography shows depths from 30 km in the eastern-central part to
37.5-42.5 km in south and south-east and 40 km towards east. Two elevated crustal
blocks are located in the Miercurea-Ciuc Sf.Gheorghe area (30-35 km) and the
Bacau-Roman area (37.5 km). The first block overlaps with the Volcanic Neogene
zone of the Eastern Carpathians where the base of crust has suffered an erosion
process during magmatic activity. The second block is coincident with the central

segment of the Scythian platform and with the western margin of the Moldavian
platform. In the Scythian platform crustal thickness does not exceed 40 km while in
the Moldavian it is thicker than 40 km. A thickened crust (>40 km) is found within
the Focsani basin and in the North Dobrogea orogen.
The thickness of the lower crust increases from ~7.5 km in the south-east to
20 km in the central part of the map. The thinnest lower crustal layer is located on
the Tisza-Dacia crustal block being in agreement with the delamination hypothesis
of the lower crust[33]. The thickest crustal block (~20 km) is observed on the
Moinesti-Piatra Neamt-Campulung Moldovenesc direction and overlaps with the
Scythian platform. A thinner crust (~15 km) appears to the east on a direction from
Harlau to Vaslui and Focsani. To the east of it a thickening trend is displayed.
The mean P-wave seismic velocities in the upper crystalline and lower crust
display changes in a narrow range. For the upper crystalline crust it starts from 6.00
km/s in the north-western corner and in the central-southern part (6.05 km/s) to
6.25 km/s in the central-northern part, suggesting the trend of variation as a
function of mineralogy and tectonothermal influence of the Alpine orogeny. Once
again we mention the used data are derived from field data and not from extra- or
interpolated data what confers a justified confidence of the new model. The lowest
values overlap with the Focsani basin and the northern part of the Eastern
Carpathians, younger and/or tectono-thermal worked areas in comparison with the
older Moldavian platform where some higher Vp (6.25 km/s) are noticed.
The Vp in the lower crust varies in a narrow range from 6.85 km/s to 6.95
km/s. The highest values ( 6.95 km/s) are located in between the Siret and Solca
faults and in between the Vaslui and Bistrita faults. Lower values ( 6.85 km/s)
overlap with the Carpathians orogen and its foredeep in the west and the
Moldavian platform in the east.
In comparison with other recent crustal models [6, 10, 13, 14] our model
brings more details on the Moho and lower crust topography and on the mean Pwave seismic velocities in the crust by using a larger and extended database than
the previous models. In addition to those models we have attempted to interpret our
crustal model in correlation with tectonics of region.
Acknowledgements. The authors thank the anonymous reviewer for the
suggestions done for improvement of paper. This paper reveals a part of the results
achieved within the VELOROM project (contract 32153/2008, Partnership
Program), the project PN09 02 01 (contract 03 01PN 09-30, Nucleu Program) and
the project PN II-RU 120/2010 (UEFISCDI Human Resources Program), all of
them financed by the Romanian Ministry of Education, Research Youth and Sport.

REFERENCES
1.V. Raileanu, D. Tataru, and B. Grecu, Crustal models in Romania. I-Moesian Platform. Rom.Rep.
Phys, Manuscript Acceptance for Publications.
2. M. Socolescu, D. Popovici, M. Visarion and V. Rosca, V. Structure of the Earth crust in Romania
as based on the gravimetric data, Rev.Roum.Geophys., 8, 3-11, 1964.
3. D. Radulescu, I. Cornea, M. Sandulescu, P. Constantinescu, F. Radulescu et A. Pompilian,
Structure de la croute terrestre en Roumanie, Essai dinterpretation des etudes sismiques
profondes, Rev.Roum.Geophys., 20, 5-32, 1976.
4. F. Radulescu, Seismic models of the crustal structure in Romania, Rev.Roum.Geol.Geophys.
Geogr.Ser.Geophys., 32, 13-17, 1988.
5. D. Enescu, D. Danchiv, A. Bala, Lithosphere structure in Romania.II. Thicckness of the crust.
Depth- dependent propagation velocity curves for the P and S waves, St.cerc. GEOPFIZICA,
30, 3-19, 1992.
6. M. Martin, J.R. Ritter and the CALIXTO Working Group, High-resolution teleseismic body-wave
tomography beneath SE Romania I. Implications for three-dimensional versus
onedimensional crustal correction strategies with a new crustal velocity model. Geophys. J.
Int. 162, 448460, 2005.
7. F. Hauser, V. Raileanu, W. Fielitz, A. Bala, C. Prodehl, G. Polonic, A. Schulze, VRANCEA'99-the
crustal structure beneath SE Carpathians and the Moesian Platform from a seismic refraction
profile in Romania, Tectonophysics 340, 233256, 2001.
8. F. Hauser, V. Raileanu, W. Fielitz, C. Dinu, M. Landes, A. Bala, C. Prodehl, Seismic crustal
structure between Transylvanian Basin and the Black Sea, Romania, Tectonophysics, 430, 125, 2007.
9. T. Diehl, J.R. Ritter and CALIXTO group, The crustal structure beneath Romania from teleseismic
receiver functions, Geophys.J.Int.,163, 238-251, 2005.
10. R. Raykova and G.F. Panza, Surface waves tomography and non-linear inversion in the southeast
Carpathians, PEPI, 157, 164-180, 2006.
11. M. Tesauro, M.K. Kaban, S.A.P.L. Cloetingh, EuCRUST-07: a new reference model for the
European crust. Geophys. Res. Lett., 35, L05313, 2008.
12. M. Grad,T. Tiira and ESC Working Group, The Moho depth map of the European Plate,
Geophys. J. Int. 176 (1), 279292, 2009.
13. I. Molinari and A. Morelli, EPcrust: A reference crustal model for the european plate. Geophys.
J. Int. 185 (1), 352364, 2011.
14. M. Ivan, Crustal thickness in Vrancea area, Romania from S to P converted waves, J.Seismol. 15,
317-328, 2011.
15. M. Sandulescu, Geotectonics of Romania, Tech.Publ. House, Bucharest , pp. 336, 1984.
16. M. Sandulescu, M. Visarion, M., La structure de la plate-formes situees dans lavant-pays ety
au-dessous des nappes du flyschdes Carpathes Orientales, I.G.G., St.Tehn.Ec., Ser.D,
Prosp.Geof.,15, 61-68, 1988.
17. G. Polonic, Structure of the crystalline basement in Romania, Rev. Roum. Geophysique, 40, 5770, 1996.
18. N.I. Christensen and W.D. Mooney, Seismic velocity structure and composition of the continental
crust; a global view. J. Geophys. Res. 100 (B6), 97619788, 1995.

19. H.J.A.V. Avendonk, W.S. Holbrook, G.T. Nunes, D.J. Shillington, B.E. Tucholke, K.E. Louden,
H.C. Larsen, J.R. Hopper, Seismic velocity structure of the rifted margin of the Eastern Grand
Banks of Newfoundland, Canada. J. Geophys. Res. 111, B11404, 2006.
20. W.D. Mooney, C. Prodehl, N.I. Pavlenkova, Seismic velocity Structure of the Continental
Lithosphere from Controlled Source Data, in International Handbook of Earthquake&
Engineering Seismology, Eds: W.H.K. Lee, H. Kanamori, P.C.Jennings, C. Kisslinger,
Academic Press, Amsterdam, 887-910, 2002.
21. D. Radulescu, I. Cornea, M. Sandulescu, P. Constantinescu, F. Radulescu et A. Pompilian,
Structure de la croute terrestre en Roumanie, Essai dinterpretation des etudes sismiques
profondes, Rev.Roum.Geophys., 20, 5-32, 1976.
22. F. Radulescu, Inregistrari seismice din punctul de explozie Vinnita (URSS) in : Cercetari asupra
structurii scoartei terestre si a mantalei superioare in partea sudica si vestica a R.S.Romania,
Raport de faza, contract CSEN 31.81.3, Arhiva CFP Bucuresti, 1982.
23. A. Pompilian, F. Radulescu, M. Diaconescu, M. Bitter, M., A. Bala, Refraction seismic data in the
eastern side of Romania, Romanian Reports in Physics, 7-8, 613-621, 1993.
24. B. Stanchievici, Descifrarea structurii profunde a teritoriului Romniei, Raport la tema A3/19911992, subtema A3C, Arhiva IGG, Bucureti, 1992.
25. V. Raileanu, C.C. Diaconescu, and F. Radulescu, Characteristics of Romanian lithosphere from
deep seismic reflection profiling,Tectonophysics, 239, 165-185, 1994.
26. V. Raileanu, Study of some physical parameters of lithosphere in some zone of Romania, Ph.D.
Thesis, Inst.Fizica Atomica, Bucharest, 232 pp, 1998.
27. J.P. Ligorria, C.J. Ammon, Iterative deconvolution of teleseismic seismograms and receiver
function estimation, Bull. Seism. Soc. Am., 89, 1395-1400, 1999.
28. M. Campillo, A. Paul, Long-range correlations in the diffuse seismic coda, Science, 299, 547549, 2003.
29. N.M. Shapiro, M. Campillo, Emergence of broadband Rayleigh waves from correlations of the
ambient seismic noise, Geophysical Research Letters, 31, L07614, doi:10.1029/
2004GL019491, 2004.
30. I. Clark, Practical Geostatistics, Geostok Limited. Alloa Business Centre, Whins Road, Alloa,
Central Scotland FK10 3SA, UK, pp.119, 2001.
31. J. Davis, Statistics and Data Analysis in Geology, 3rd Edition. John Wiley & Sons, 2002
32. P. Wessel and W.H.F.Smith, New, improved version of the Generic Mapping Tools released. Eos
Trans. AGU 79, 579, 1998.
33. M. A. Fillerup, J.H. Knapp, C.C. Knapp, V. Raileanu, Mantle earthquakes in the absence of
subduction? Continental delamination in the Romanian Carpathians , Lithosphere, October
2010, 2, 333-340, doi:10.1130/L102.1.

Short title: Crustal model of Moldavian platform and adjacent areas

You might also like