Tectonoelectric zonation in the Hellenic Arc

1Letter

Tect~n~ph.vs;cs, 143 (1987) 337-342 Elsevier Science Publishers

B.V., Amsterdam

- Printed

Tectonoelectric

in The Netherlands

ronation in the Hellenic Arc

M. LAZARIDOU-VAROTSOU

’ Department

Section 1

’ and D. PAPANIKOLAOU

2

’ Knossou ST. 36, Ano GIyfado, Athens 165 61 (Greece) ofAthens,Fanepistim~oupo~is Zografou, Athens 157 84 (Greece)

of Geology, University (Received

March

5.1987;

revised version accepted

September

15, 1987)

Abstract Lazaridou-Varotsou,

M. and Papanikolaou,

D., 1987. Tectonoelectric

zonation

in the Hellenic

Arc. Tectonophysics,

143:

337-342. The electrical Greece

(apparent)

and in two directions,

allow the folIowing

E-W

p were measured

and N-S.

by means of the magnetotelluric In most of the sites, paw differs considerably

three zones to be ~stinguish~:

intermediate

zone along

the northern

Aegean

the main mountain

chain

(1) an external in continental

zones are compared

which all show a similar

to data

geometry

zone along western Greece with ~nw > PNs; (2) an where pN5 > ~nw and (3) an internal zone in

Greece

from geotectonic,

of zonation

Introduction

Since March 1981 a systematic study of the variations in the electric field of the earth has been carried out at various sites in Greece. The basic scope of this study was the detection of transient variations in the electric field that precede earthquakes (Varotsos et al., 1982; Varotsos and Alexopoulos, 1984). These variations (hereafter called Seismic Electric Signals; SES) depend on the epicentral distance and the magnitude of the impending event and are not accompa~ed by a detectable variation in the magnetic field. Varotsos and Alexopoulos (1984) mentioned that the so-called magnetotelluric variations (MT), i.e. variations in the electric field induced by disturbances of the magnetic field, render the identification of the SES more difficult, especially during periods of magnetic storms. However, when measuring the magnetotelluric response function of each station (i.e. by estimating the corresponding impedance tensor elements) the subtraction of 0040-1951,‘87/$03,50

effect at twenty sites in from pNS. These results

where pEW > pNS.

The above resistivity m~surements

resistivities

0 1987 Elsevier Science Pubhshers

B.V.

almost

neotectonic,

parallel

seismotectonic

to the active Hellenic

and in situ stress Arc.

the magnetotelluric disturbances from the recordings of the electrotelluric stations is straightforward (Varotsos and Alexopoulos, 1984). One should emphasize that the fundamental difference between SES and ma~etotellu~c disturbances is that the latter were detected (almost simultaneously) at all stations in the network, whereas the SES were recorded only at a restricted number of stations. This difference mainly arises from the so-called ““directivity effect” and distance dependence of the SES. The study of magnetotellu~c ~sturb~ces can lead to a good estimate of the “apparent resistivities” (p) at each site. If x and y refer to the measuring axes in the N-S and E-W direction, respectively, then the two horizontal electric field components E, and E,, are connected to the two horizontal magnetic field components H, and H,, through the relation (in the frequency domain):

338

where Z,,, Zx,, ZYX and ZYY are elements so-called

impedance

resistivity

structure

tensor

2, and depend

underneath

of the on the

a particular

sta-

tion. Usually,

instead

ized impedance defined

Z, the normal-

of the tensor

tensor

T is used;

the latter

four sites: Thiva, Patra, Komotini and Syros. For the other sites the resistivity of the one direction is appreciably larger than the other. The phenomenon is more intense for the following five sites: Astakos,

is

as follows:

Chalkida.

Rentina,

exceeds the corresponding of magnitude eration).

the angular

frequency

and pLothe

magnetic permeability in a vacuum. The apparent . . . . resrsttvmes pNS and pEW are defined by: PNS

=

1TxyI2

and

PEW=

iq_x

Gorgopotamos

and

p,,-vahe

by one order

or more (the exact value of the ratio

pEW/pNS depends where w denotes

Anchialos,

For the first three sites the value of pEW

on the frequency

under

consid-

On the other hand in the case of Gorgo-

potamos and Chalkida, pNs exceeds pEw by one order of magnitude. For the other sites investigated, we have the following values (for T = 10 s): for Assiros, Thisvi, Alfioussa and Kalamata, the ratio pEW/pNS is around 2 while for loannina it

I*

The apparent resistivities pNS and pEw at each station can be determined from an analysis of the

lies between

6 and 20. On the other hand the ratio

magnetotelluric data. They should not be confused with the so-called “relative effective resistivities” defined by Varotsos

and Alexopoulos

(1984)

which are drawn from analysis of SES data without considering any magnetotelluric disturbance. Twenty sites were investigated and most of them showed a large difference in their resistivities in the two measuring directions, i.e. E-W and N-S. A detailed analysis of the p-variations versus the frequency will soon be published separately. The scope of the present paper is to draw attention to the fact that sites exhibiting similar “electrical anisotropy”, in the sense that (pEW/pNS) > 1 or (pEW/pNS) -c 1, form zones that show a significant correlation with tectonic data. Experimental

results and discussion

Details of the experimental procedure can be found elsewhere (Lazaridou-Varotsou, 1986). An example of the analog recordings is given in Fig. 1. The twenty sites at which measurements were carried out are marked in Fig. 2. Note that since the end of 1982 most of these sites have been telemetrically connected to Athens by telephone line and electrotelluric

data (sampling

rate:

three

samples per set) are continuously collected (Varotsos and Alexopoulos, 1986; Varotsos et al., 1986). Approximate equality of the two resistivities pEW and pKs was found only for the following

Fig. 1. Typical example of variations in the electrotelluric field at the MEG station.

339

Sigeritsa, PNS

zone

’

Nafplio,

Megara

and Chalkida)

where

PEW,

ZZZ: northeastern

Aegean,

Greece

comprising

tina, Anchialos

four

and Thisvi)

around

the North

sites (Assiros,

Ren-

where pEW > pNS.

The site of Patra (where pEW = pNS) lies on the boundary

between

zones Z and ZZ, and the site of

Thiva (where pEW = pNS) lies on the boundary between zones ZZ and ZZZ. It is known semble

that resistivity

of inhomogeneities

structure

at different

is an enscales and

that in some cases large structures may have MT responses comparable to those of small structures. However it seems that the observed zonation of electrical resistivity cannot be attributed to smallscale phenomena, such as the influence of special lithologies, 100 km

very Fig. 2. Schematic (mean oriented east-west arrows

representation

skin depth double

because

there is a great variety of rocks

participating in each zone and the composition of the upper crust in the similar zones Z and ZZZ is

of the order

arrow

oriented

of the results of 10 km):

corresponds

arrow

indicates

for T= 10 s

the north-south

to pNS > pEW whereas that

pEW z pNS;

the both

are used where pNS = pEw.

pNs/pEW is around 2 for Nafplio, 4 to 5 for Sigeritsa, Kefalonia and Megara, 6 for Iraklio and between 5 and 9 for Veria. A more or less systematic distribution of the electrical anisotropy (in the sense that pEW # pNS) can be seen in Fig. 2. In this figure the N-S-directed double arrow identifies cases where the predominant resistivity is in the N-S-direction, i.e. pNS > pEW, and the east-west oriented double arrow indicates that pEW > pNS. For the four sites at which pEW=pNS both symbols are used. A close look at the figure indicates the following electrical resistivity zones, excluding the cases of the islands of Kefalonia, Crete and Syros (where the so-called “island effect” influences the measurements of the electric field; see LazaridouVarotsou, 1986): zone I: western Greece, comprising four sites (Ioannina, Astakos, Alfioussa and Kalamata) where PEW ’ PNS, zone ZZ: the central zone of continental Greece, comprising five sites (Veria, Gorgopotamos,

different.

What

is remarkable

is that

“electrical zonation” observed for areas tinental Greece has almost a N-S-direction,

the

of conwhich

is parallel to the geotectonic trend of the Hellenides. Thus, a comparison of the electric zonation with the available data on neotectonics, seismotectonics and in situ stress measurements may allow us to draw conclusions regarding tectonic

stresses and electric resistivity.

Neotectonics

and seismotectonics

The neotectonic research on faults in Greece activated during the last few million years (Mercier et al., 1979; Angelier, 1979; Lyberis, 1984; Mariolakos et al., 1985) can be summarized as follows: (1) in western Greece there is a compression in the ENE-WSW direction (maximum horizontal stress oi) as indicated by thrust faults, (2) in continental Greece and the southern Aegean there is a predominance of normal faults with extension in various directions depending on the area. In the northwestern Peloponnese and Sterea E-W faults with N-S extension predominate whereas in the southeastern Peloponnese, Attica, south Evia and Cyclades NW-SE faults with NE-SW direction of extension (minimum principal horizontal stress as) predominate, and (3) in northeastern Greece and the northern Aegean

340

there is a predominance dextral

and sinistral,

Seismotectonic Kenzie,

1978;

Drakopoulos

faults, faults.

research

(Ritsema,

Nakamura

and

and Delibasis,

in western quake

of strike-slip and also normal

Greece

mechanisms

both

1974;

Uyeda,

Mc1980;

1982) has shown

the dominant

shallow

show horizontal

that earth-

compression

seismotectonic

data,

it is possible c, across

three

zones,

a, b and

which

can

be correlated

tinguished

to distinguish the Hellenic

arc

to the previously

electric zones Z-III

dis-

(Fig. 3).

A comparison of the neotectonic and seismotectonic zones of the Hellenic arc with the electric zones

shows

the same

geometry,

indicating

not

with the slip vector trending

in the ENE direction.

only that the general

The same focal mechanisms

are observed

south-

similar before,

to the geotectonic structure, as was shown but also that each electric zone corre-

200 km

sponds

to a tectonic

modern

as the overall

earthquakes

occurring

ern Greece south

below

along a Benioff

of Attica

and

western

for deep

and

zone reaching

Cyclades

in

the

direction

of the zonation

zone of a special nature

stressfield

is

as far

and the position

of the

volcanic arc whereas the focal mechanisms of shallow earthquakes are tensional in the same area. The existence of strike-slip movements is also

principal axes ul, a2 and u3 are concerned. Thus, a predominant resistivity in the E-W-direction is

confirmed

a and c) whereas the N-S-direction

by fault plane solutions

in the northern

Aegean sea, especially at the westward prolongations of the north Anatolian fault zone in the Saros basin. Thus, summarizing

observed

where ur is in the ENE direction

(zones

the predominant resistivity is in where ur is vertical (zone b).

Zn-situ stress measurements the

neotectonic

and Paquin et al. (1982) carried out a survey in Greece using the overcoring method in drillholes

i

Fig. 3. Summary resulting

are compared (broken reverse

faults;

compression

(I, b and

to the electric

lines).

3 = vertical deduced

of the neotectonic

in three zones

I = horizontal 2 = horizontal

zones

and seismotectonic c (continuous I, ZI and

compression extension

shear zones with strike-slip with slip in the E-W

III

with with

faults;

of Fig. 2 thrust

normal

and faults;

4 = area under

to ENE-WSW

from the fault plane solutions.

data,

lines) which

direction

Fig. 4. The data of the in-situ al., 1982) distinguished tinuous

lines),

separating

extension,

compared

2 (broken

lines).

stress measurements

in four zones, areas

(Paquin

A, B, C and

of actual

et

D (con-

compression

and

to the electric zones I, II and III of Fig.

341

to measure

the actual stressfield

in the horizontal horizontal

and the directions

plane of maximum

stress (compressive

and minimum

or extensive).

From

their data, four zones can be distinguished, C and

D, each characterized

or extensional correlated

stressfield.

to the electric

also to the neo-

and

A, B,

by a compressional

Three

of them

can be

zones I, ZZ and ZZZ and

seismo-tectonic

zones

and c (Fig. 4). Their stress zone D, in Thraki, outside

a, b lies

the area dealt with in this paper.

The comparison of the zonation of the in-situ stress measurements with the electric zonation shows a remarkable similarity and verifies the result of the comparison between the electric zonation and the neo- and seismo-tectonic zonacomprestion with pEW > pNS where horizontal sion along the ENE or E-W direction occurs (zones zontal

A and C) and with pNS > pEW where horiextension along various directions occurs

(zone B). Concluding remarks

Fig. 5. The tectonoelectric ternal

The electric anisotropy of resistivity identified in Greece shows a systematic distribution which permits the distinction of electric zones. These zones and the zones which can be drawn on the basis of geotectonic, neotectonic, seismotectonic and in situ stress measurements all show a similar

zone with

resistitity pression

subparallel (aj

vertical);

tion of electric vertical);

zonation

(dominance

direction

to the o1 direction d-intermediate

excess resistivity

3-internal

of the Hellenic

of) E-W

zone

excess resistivity

subparallel

The black arrows

indicate

with

com-

zone with N-S

direc-

direction

to horizontal the direction

of electric

of horizontal

and horizontal E-W

arc. I -ex-

shear

extension

(0,

of electric

( oz vertical).

of plate movement.

geometry of zonation, parallel to the active Hellenic arc. The three main tectono-electric zones are (Fig. 5): (I) An external zone along western Greece, adjacent to the Hellenic trench, characterized by excess resistivity in the E-W-direction; it is subparallel to the horizontal E to ENE compression deduced from the focal mechanisms of shallow earthquakes, from in situ stress measurements and from neotectonic observations, (2) An intermediate zone along the main mountain chain of continental Greece which is characterized by larger resistivity values in the N-S-direction and extension in various directions. In the northern Peloponnese and central Sterea E-W normal faults predominate with ug in the N-S direction whereas in the southeastern Peloponnese and the south Aegean NW-SE faults predominate with u3 in the NE-SW direction,

(3) An internal zone at the core of the Hellenic arc in the northern Aegean, characterized by excess resistivity in the E-W-direction, which is subparallel to the horizontal compression deduced from in-situ stress measurements and to the main direction of strike-slip faulting. Thus, the general result seems to be a parallel development of the electric excess resistivity in the E-W or ENE-WSW horizontal axis of compression in the case of the external and internal tectonoelectric zones parallel to the direction of the plate movement. In the intermediate tectonoelectric zone, where cri is vertical and extension prevails in the horizontal plane, an excess resistivity is developed in the N-S-direction.

342

References Angelier, J., 1979. N~otecto~que de I’arc Egeen. Sot. G&of. Nord. Publ., 3: I-417. Drakopoulos, J. and Delibasis, N., 1982. The focal mechanism of earthquakes in the major area of Greece for the period 1947-1981. Univ. Athens, Seismol. Lab., Publ. 2: 72 pp. Lazaridou-Varotsou, M., 1986. A study of electric polarity from changes of the earth’s electric field in Greece. Thesis, Univ. of .4thens, Athens, pp. l-224. Lyberis, N., 1984. Geodynamique du domaine egeen depuis le Miodne superieur. Thesis, Univ. P. et M. Curie, Paris, pp. l-367. Mariolakos, I., Papanikolaou, D. and Lagios, E., 1985. A neo-tectonic geodynamic model of Peloponnesus based on morphotectonics, repeated gravity measurements and seismicity. Geol. Jahrb., Reihe B, 50: 3-17. McKenzie, D., 1978. Active tectonics of the Alpine-Himalayan belt, the Agean Sea and surrounding regions. Geophys. J.R. As&on. Sot.. 55(l); 217-254. Mercier, J.L., Delibassis, N., Gauthier, A., Jarrige,J.J., Lemeille, F.. Philip, H., Sebrier, M. and Sorel, D., 1979. La nriotectonique de I’Arc Egeen. Rev. G&of. Dyn. Geogr. Phys., 21 (I): 67-92.

Nakamura, K. and Uyeda, S., 1980. Stress gradient in arcback-arc regions and plate subduction. J. Geophys. Res., 85 (32 1): 6419-6428. Paqum, C., Froidevaux, C., Bloyet, J., Ricard, Y. and Angelidis, C., 1982. Tectonic stresses on the mainland of Greece: in-situ measurements by overcoring. Tectonophysics, 86: 17-26. Ritscma, A.R., 1974. The earthquake mechanisms of the Baikan region. R. Neth. Meteor. Inst., De Bilt, Sci. Rep., 74 (4): l-36. Varotsos, P. and Alexopoulos, K., 1984. Physical properties of the variations of the electric fieid of the earth preceding earthquakes. Tectonophysics, 110: 73-88, 99-125. Varotsos, P. and Alexopoulos, K., 1986. Recent developments in the detection of preseismic electric signals. In: P. VarotSOSand K. Alexopoulos, Therm~~a~cs of Point Defects and their Relation to Bulk Properties. North-Holland, Amsterdam, pp. 410-411, 417-420. Varotsos, P., Alexopoulos, K. and Nomicos, K., 1982. Electrotelluric precursors to earthquakes. Proc. Akad. Athenon, 57: 341-363. Varotsos, P., Alexopoulos, K., Nomicos, K. and Laxaridou, M., 1986. Earthquake prediction and electric signals. Nature, 322: 120.