920

Bulletin of the Seismological Society of America, Vol. 94, No. 3, pp. 920–929, June 2004

Tectonic Deformation and Microseismicity of the Saronikos Gulf, Greece

by J. Makris, J. Papoulia, and G. Drakatos

Abstract We deployed an on/offshore seismic network consisting of 36 three-component stand-alone digital stations in the Saronikos Gulf and surrounding areaand observed the microseismic activity from 17 August to 9 December 2001. Thegeometry and location of the network were designed to observe the offshore activityto connect known active faults onshore with those offshore. An average of 15 eventsper day were recorded by a minimum of six stations. The focal parameters of 739events could be defined. Within the network area we located 306 events with an rmstravel-time residual of about 0.2 sec. The epicenters in the Saronikos Gulf are alignedmainly along an east–west trending fault with a length of nearly 30 km. This faultis located north of Aegina and could produce earthquakes as large as Ms 6.5. It isthe offshore continuation of a seismic zone located in east Corinthia. A second activefault was identified east of Aegina, trending northeast–southwest. It is probablylinked to the northeast–southwest striking zone of the Salamis-Fili seismic region,activated by the September 1999 event of Athens. Outside the network we mappedan active area at the Parnis mountain, still active after the Ms 5.9 event, and a secondone further northeast at the western coast of Evia. This zone is close to Chalkis andthe recently active Psachna area. The network also mapped intense activity at thenortheastern coast of Evia, close to the city of Kymi and the area of Skyros, whichwas seriously damaged by an Ms 5.8 earthquake in 2001. The onshore/offshore localnetwork identified several regions of very high seismic activity not detectable byregional networks, and demonstrated once more the advantage of deploying localseismic networks for mapping active deformation in very short periods of time.

Introduction

The region of Attica and, in particular, that of the cityof Athens have long been considered aseismic (Sieberg,1932). This was mainly attributed to the fact that the his-torical center of the City, built on schist of high rigidity, isnot vulnerable to earth tremors. However, since Athens hasbeen constantly expanding, densely urbanized areas have en-croached into seismogenic regions and Athens has becomevulnerable to earthquakes. Today, according to estimates ofPapadopoulos and Arvanitides (1996) the seismic risk ofAthens is among the highest in Greece. This was verifieddramatically by the Ms 5.9 earthquake of 7 September 1999,which was the costliest natural disaster of Greece in recenthistory.

Prior to this event and during the instrumental period ofGreek seismology, no significant earthquake activity hadbeen reported. The strongest shocks (Ms 6.0) to affect Athenswere located in the Thiva-Oropos zone at a distance of morethan 50 km from the historic center of the city (Papazachosand Papazachou, 1997). Events of rather small magnituderecorded from time to time in the vicinity of Athens, likethose of 3 April (Ms 4.2) and 4 December (Ms 3.5) 1965 and

a cluster of events ranging from 3.8 to 4.5 in November 1997[see the National Observatory of Athens (NOA) bulletins],did not attract public interest. It was therefore assumed thata detailed examination of the microseismicity of the Athensarea was unnecessary.

The 1999 event, however, clearly demonstrated that thelow rate of seismicity alone is not a safe criterion to assessthe seismic potential of a region. Only a detailed tectonicanalysis and mapping of active faults and soil conditionscontribute toward a reliable seismic risk assessment. Hadsuch studies been undertaken before 1999, they certainlywould have revealed the potential risk of the active tectonicstructures that ruptured during the 7 September 1999 seismiccrisis.

Public awareness and the need for a thorough under-standing of the regional seismic potential after the 7 Septem-ber 1999 event motivated the Earthquake Planning andProtection Organization of Greece to support an onshore-offshore microseismicity survey in the region of the Saron-ikos Gulf, whose purpose was to identify active faults on-and offshore and to define their seismic potential. A local

Tectonic Deformation and Microseismicity of the Saronikos Gulf, Greece 921

seismic network was deployed consisting of eight ocean bot-tom seismographs (OBS) and 28 land stations (Fig. 1). Thisnetwork operated for 45 days.

Both marine and land stations were equipped withSEDIS seismic recorders of GeoPro, Hamburg (Makris andMoeller, 1990; GeoPro technical reports). These instrumentsare capable of recording six seismic channels on hard discwith 10-GB capacity. Onshore, an integrated Global Posi-tioning System (GPS) receiver provides coordinates and tim-ing at each location. Marine stations are housed in a 17-inchglass sphere and record seismic signals directly on the seafloor. A gimbal-mounted three-component geophone systemand a deep sea hydrophone are used for this purpose. Theinstruments are autonomous and can be acoustically ad-dressed by a surface-mounted system. Timing is providedby an internal quartz clock. By measuring the quartz tem-perature continuously and GPS time at the beginning and endof the survey, time-drift corrections are obtained for eachinstrument.

Regional Tectonics

The region of Attica as presented in the 1:50,000 scalegeological maps of the Institute of Geology and Mineral Ex-ploration of Greece (IGME) (1989), belongs to an autoch-thonous metamorphic basement of Paleozoic-Mesozoic age,consisting of marbles and schists (see also Jacobshagen,1986). Remnants of a “tectonic cover” have been identifiedin several localities. A major tectonic boundary, dominatingthe area, separates the mountains of Parnis to the northwestfrom those of Pendeli and Immitos to the southeast. Thesetwo are characterized by metamorphic rocks, mainly marblesand schists belonging to the Cyclades Massif, which dip tothe northwest, below the nonmetamorphic limestones ofnorthwestern Attica. Normal faults along the Parnis moun-tain are considered to be the eastward extension of the Cor-inthiakos Gulf fault system and are of roughly east–westorientation. During historic times and the more recent periodof instrumental earthquake observations in Greece, consid-erable earthquake activity has been observed along this zone,whereas the southeastern part of Attica appears to be nearlyaseismic (Galanopoulos, 1960, 1961; Makropoulos et al.,1989; Papazachos and Papazachou, 1997).

The Saronikos Gulf is a neotectonic basin, divided by avery shallow north–south-oriented platform into a westernand an eastern part. The islands of Methana, Angistri, Ae-gina, and Salamis (see Fig. 1) are situated on this platform.The western part of the Gulf consists of two basins. TheEpidaurus Basin is of west-northwest–east-southeast orien-tation and more than 400 m deep, whereas the Megara Basinis of east–west orientation with relatively shallow depth, lessthan 250 m. The marginal extensional faults of the EpidaurusBasin strike parallel to the bathymetry and have a 350-mthrow, forming a symmetric graben. The Megara Basin isformed by east–west to east-northeast–west-southwest mar-

ginal faults. The Eastern Saronikos gulf is dominated bynorthwest–southeast faults of relatively small throws and al-ternating shallow basins and plateaus (Papanikolaou et al.,1988).

Several volcanoes and volcanic outcrops of Plio-Quaternary age are also situated in the Gulf and are part ofthe western active Aegean volcanic arc. The westernmostgroup of volcanic outcrops are those of the Loutraki-Sousakiarea. They are locally more than 70 m thick (Freyberg,1973). The peninsula of Methana consists mainly of a largevolcanic dome and volcanic rocks of Pliocene and Quater-nary age which are widely exposed (Gaitanakis and Dietrich,1995). The most recent eruption is dated around 250 B.C.and was desrcibed by Strabo (Stothers and Rampino, 1983).To the northwest they appear to truncate all but the upper-most 100 m of Quaternary sediments (Papanikolaou et al.,1988). Volcanic rocks are also found to the south of theisland of Poros (Fytikas et al., 1986). Clearly, volcanic ac-tivity in the Gulf is extensive, recent, and very close to Ath-ens. It is poorly understood and its hazard has never beenassessed.

From displacements at faults and the distribution of sed-iments it is shown that the Saronikos Gulf is more active inthe west than in the east. This has also been verified by themicroseismic activity, as will be presented subsequently.

The most seismically active area in the vicinity of Ath-ens is the Corinthiakos Gulf, which exhibits deformationtypical of the extension that dominates central Greece andthe Peloponnese (Jackson et al., 1982; Collier et al., 1992;Armijo et al., 1996), and an annual extension rate on theorder of 1 to 7 mm. About 20% of this extension is linkedto east–west-striking normal faults of eastern Corinthia (Pa-pazachos and Kiratzi, 1996). The continuation of this faultsystem to the east was also mapped by high-resolution,single-channel reflection seismic profiles (Papanikolaouet al., 1988) offshore in the Saronikos Gulf. A significantpart of the observed seismicity coincides with this system offaults, as will be discussed subsequently.

Microseismicity

Within 45 days of field observations we recorded 739microearthquakes above a threshold magnitude of ML 0.3(Fig. 2), which were observed by at least six stations.

Magnitudes were defined by the coda length of the seis-mograms, calibrated by using earthquakes recorded also bythe Seismograph Network of the NOA. For the magnitudeestimates, we derived a mean function from four stationshaving similar site response and applied this relationship tothe full temporary network (Fig. 3).

The locations of the hypocenters were obtained usingHYPOINVERSE software (Klein, 1989). HYPOINVERSEallows the application of spatially varying local velocitymodels for the hypocenter location. The velocity model used

922 J. Makris, J. Papoulia, and G. Drakatos

Figure 1. Location of the on and offshore seismic network used to record mi-croearthquake activity in the area of Saronikos gulf, Greece. Stations are representedby triangles.

was determined by a controlled source seismic experimentbetween Kythnos and Eratini, crossing the Saronikos Gulffrom southeast to northwest (Makris et al., 2002). However,to avoid local geological complexities, we adopted a meanvelocity model, as presented in Table 1.

Within the network most travel-time residuals after lo-cation have an rms 0.2 sec. This corresponds to an epicentralaccuracy of the order of �2 km, depending on the localvelocity model used, whereas the hypocentral depths are af-fected by an error twice this value. Outside the network,errors are significantly higher (Fig. 4).

The frequency-magnitude distribution of the recordedevents is shown in Figure 5. Within the network the thresh-old magnitude of completeness is ML 1.7. Outside the net-work this increases to ML 2.3.

To demonstrate the advantage of using local networksover regional ones in assessing the seismic activity in thisarea, we present in Figure 6 a comparison between historicevents (Fig. 6a), instrumentally recorded events publishedby the NOA regional network (Fig. 6b), and events recordedby this temporary local network (Fig. 6c). The advantage of

using a local network of densely spaced stations is obvious.Even a short period of observations is enough to resolveactive structures that a regional network would require tensof years to identify due to the different threshold of com-pleteness.

Within the network and its immediate vicinity we lo-cated 306 events (Fig. 7). Most of them are within the upper15 km of the crust, whereas only a few were found at depthsexceeding 30 km. The distribution of the foci is illustratedin Figure 8. The southwest–northeast oriented profile (Fig.8) extends from eastern Peloponnese (point A) crossing theSaronikos Gulf to central Evia (point B). Hypocenters con-tained in a 100-km-wide zone, parallel to line AB, were pro-jected in the surface defined by AB and the z axis extendingto 40 km depth. Their distribution shows some distinct zonesof foci concentration that can be associated with knownfaults. At 90 km (see Fig. 8) the hypocenters are located atthe “Aegina fault” as presented in Figures 2 and 7. The lineardistribution of the foci along the fault extends to a depth ofmore than 15 km and marks a deep fracture of the thinnedSaronikos crust. The next concentration, 20 km further

Tectonic Deformation and Microseismicity of the Saronikos Gulf, Greece 923

Figure 2. Microseismic activity recorded by the Saronikos seismic network in a 45-day period. The map shows the epicentral positions of 739 events, each having beenidentified by a minimum of six stations. Depth of the hypocenters is indicated by thecolor scale.

northeast coincides with the faults east of the island ofSalamis, which were reactivated by the “Athens Event” of7 September 1999. They do not seem to have a deep exten-sion and probably originate at the thrust front between theCyclades Massif and west Attica. In the Parnis area, 30 kmfurther northeast, we encounter the Fili fault zone. The seis-

micity shows that all faults around the Parnis epicenter of7 September 1999 are still active and that the activity extendsthrough the complete crust.

At 45-km distance to the northeast we cross an activeseismic zone that is associated with the Chalkis area (seeFig. 2). The Psachna region nearby is presently very active

924 J. Makris, J. Papoulia, and G. Drakatos

Figure 3. Calibration functions used to define lo-cal magnitudes by the coda-duration method. The av-eraged values were derived from functions defined atfour different location of comparable recording con-ditions and therefore influenced by similar site effects.

Table 1Local Velocity Model Used in the Location of Events

(Makris et al., 2002)

Velocity Vp (km/sec) Depth (km)

4.7 0.05.7 3.06.8 7.08.1 17.0

and has hosted several Ms 4.5 events. Another 35 km to thenortheast we encounter the Kymi fault zone with intensecrustal seismicity. This system of faults separates the north-ern part of the island of Evia from the Skyros basin andtruncates the entire crust. Large-magnitude events have beenreported from this zone (Makropoulos and Burton, 1984;Papazachos and Papazachou, 1997). Finally, around theSkyros area, 65 km to the northeast off the coast, we en-counter a seismically active region with a diffuse distributionof foci. This area is notorious for its intense seismicity andhas hosted several large earthquakes (see, e.g., Makropoulos,1978; Makropoulos et al., 1989). The large distance of theSaronikos network from the island of Skyros and the limitedperiod of monitoring does not permit us to associate therecorded microseismicity with the existing faults. A specialstudy with a local network is needed because the positionand size of the active faults around Skyros are not yet clearlymapped.

In general, below Peloponnese and particularly the Sa-ronikos gulf, the foci are located much deeper than belowAttica, Evia, and the Aegean domain (see Fig. 8). This phe-nomenon points to the existence of two regional stress fieldsthat dominate the deformation of the Hellenides and the Ae-gean region. Deeper seismicity, at subcrustal levels, as ex-

pressed below the Saronikos Gulf, is driven by the subduc-tion of the lonian oceanic plate below Greece. Shallowseismicity, on the other hand, concentrated at crustal levels,is associated with extensional stresses due to deformation ofthe upper crust. The oceanic lithosphere is subducted to thenortheast, whereas the continental crust and thinned litho-sphere are uplifted and stretched westward. The crustal de-formation is very complex involving rotation of blocks,normal faulting, and crustal fragmentation. Significant west-ward motion of the Hellenic napes creates a complex topog-raphy that complicates even more the tectonic phenomena,defining a third level of deformation at shallow depth, asrecognized by GPS observations (Kahle et al., 1995). Thiscomplex deformation has been extensively discussed in thegeological and geophysical literature (see, e.g., Le Pichonand Angelier, 1979; McKenzie, 1978; Rotstein, 1985) andis also suggested on Figure 8.

Correlation of Seismicity with Tectonic ElementsWithin the Saronikos Network

To correlate the seismicity with the tectonic elementswe used two sources of tectonic information: one is the Seis-motectonic Map of Greece, published by IGME (1989) forthe onshore areas, and the other is the submarine neotectonicmap of Saronikos (Papanikolaou et al., 1988). Although ex-tensive literature exists on the regional deformation of theHellenides (e.g., Papazachos and Delibasis, 1969; Papaza-chos and Comninakis, 1971; Taymaz et al., 1991; Stiros,1993) it is the two sources mentioned here that best describethe Saronikos area.

The first observation worth mentioning is that the east–west-oriented fault between the islands of Aegina and Sa-lamis is also marked by significant seismic activity of shal-low events. The seismicity can be followed onshore into eastCorinthia, delineating an active fault of more than 30 kmlength. Such a fault can produce events on the order of mag-nitude Ms 6.5 (see also Kiratzi et al., 1985). The fact thathistoric catalogs (e.g., Papazachos and Papazachou, 1997;Ambraseys and Jackson, 1990) do not report events fromthis offshore area (see Fig. 3) is due to bias of observationsbased on land-locked locations only.

Significant activity was also identified at the easterncoast of Aegina, along a northeast–southwest-trending lin-eament, mapped by Papanikolaou et al. (1988). At the vol-canic areas of Methana and Aegina, on the other hand, wedid not map shallow seismicity worth mentioning. Althoughmany foci locations below the Saronikos Gulf are subcrustaland associated with the subducting oceanic slab below theHellenides (see also Leydecker, 1975; Makropoulos, 1978;Makris and Roever, 1986), we could not associate the seis-mic activity with magma movements or hydrothermal pro-cesses in the volcanic zone. This is probably due to the in-adequate density of the local network, which did not permitthe identification of events of so small magnitude and vol-canic noise.

Tectonic Deformation and Microseismicity of the Saronikos Gulf, Greece 925

RMS Distribution Within the Network

0

20

40

60

80

100

120

140

160

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

RMS (sec)

Nu

mb

er o

f E

ven

ts

RMS Distribution Outside the Network

0

10

20

30

40

50

60

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

RMS (sec)

Nu

mb

er o

f E

ven

ts

ERH Distribution Within the Network

0

20

40

60

80

100

120

140

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

ERH (Km)

Nu

mb

er o

f E

ven

ts

ERH Distribution Outside the Network

0

20

40

60

80

100

120

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

ERH (Km)

Nu

mb

er o

f E

ven

ts

ERZ Distribution Within the Network

0

20

40

60

80

100

120

140

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

ERZ (Km)

Nu

mb

er o

f E

ven

ts

ERZ Distribution Outside the Network

0

20

40

60

80

100

120

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

ERZ (Km)

Nu

mb

er o

f E

ven

ts

Figure 4. rms distribution within and outside the Saronikos Gulf seismic network.

Of interest is the area of Mount Parnis where the 1999,Ms 5.9 “Athens Event” occurred. We observed a great con-centration of microearthquakes close to the 1999 epicentrallocation of the mainshock, indicating that the aftershock ac-tivity is still ongoing and the stresses responsible for thisevent have not yet been completely relaxed. Also the areaof Spata, in southeastern Attica is seismically active along afault of northeast–southwest orientation, extending offshorein the southern Evoikos Gulf (see Figs. 2, 6c, and 7).

Outside the network we recorded high seismicity inthree areas (see Fig. 2). One is the northeastern part of the

Corinth Basin, toward Thiva. This area is well known toseismologists for its intense activity and, in particular, thedestructive earthquake sequence of 1981 that severely dam-aged this region. The second area is the northeastern part ofcentral Evia (Fig. 2), north of Kymi as mentioned previ-ously. This coastal part of the island and its immediate in-terior are seismically very active. We could not define thefaults precisely, since the network is located at a distance of80 to 100 km. The lateral velocity variations in the crust andsediments are intense and cause significant systematic errorsin locating the hypocenters. Assuming a lateral homoge-

926 J. Makris, J. Papoulia, and G. Drakatos

Cumulative Magnitude Distribution Within the Network

1

10

100

1000

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2

3.6

Magnitude (ML)

log

N(M

)

Cumulative Magnitude Distribution Outside the Network

1

10

100

1000

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2

3.6

Magnitude (ML)

log

N(M

)

Limit Magnitude ofCompleteness: ML 1.7

Limit Magnitude ofCompleteness: ML 2.3

Figure 5. Frequency-magnitude distributionwithin and outside the Saronikos Gulf seismic net-work.

neous velocity model and reducing the three-dimensionalstructure to one dimension only, location uncertainties in-crease and errors may easily become systematic.

The third region we want to refer to is that to the eastand northeast of the island of Skyros (see Fig. 2). Here thehigh seismicity recorded is associated with the aftershocksequence of the Ms 5.8 “Skyros Earthquake” of 2001. Aprecise definition of the associated active faults is also notpossible for the reasons stated previously.

Conclusions

The Saronikos local seismic network, despite its limitedoperation of 45 days, produced important results permittingseveral conclusions on the existence and location of activefaults previously unknown or inaccurately mapped. Theshallow seismicity within the network is, to a great extent,associated with the deformation and opening of the Corin-thiakos Gulf, which is one of the most rapidly deforming

areas of Greece. This is particularly true for the east–westfault extending from Corinthia across the Saronikos Gulf,between the islands of Salamis and Aegina. The deep, sub-crustal seismicity, extending to a minimum of 100 km depth,is obviously linked to the interaction of a subducted oceanicIonian Sea lithosphere with the mantle below Corinthia, theSaronikos Gulf, and western Attica. The subducting litho-sphere is limited to the north by the volcanic area of Li-chades, in the northern Evoikos Gulf. Deep seismic eventsreported by the Greek Seismological Services and our ob-servations indicate that this is the northernmost limit of thesubducting slab. All the remaining seismicity is crustal. It iscontrolled to the east, in the central and north Aegean Sea,by transtension and normal faulting along the Corinthia Riftdue to rearrangement of crustal blocks escaping the com-pression along the western Hellenides.

Use of a local velocity model obtained by a controlled-source seismic experiment (Makris et al., 2002) significantlyincreased the location accuracy. The seismicity defined byobservations of the seismological network of Greece delin-eates active faults by the concentration of large-magnitudeevents observed over long periods of time. Local networkscan significantly improve the picture by detecting and betterlocating small-magnitude events over short periods of time.This permits rapid assessment of the seismic hazard of anarea. The necessity for an accurate three-dimensional veloc-ity model for the territory of Greece cannot be stressedenough. It is mandatory if we wish to more accurately definethe active faulting and provide the engineering communitywith the reliable seismic hazard parameters needed.

Acknowledgments

We are grateful to two anonymous reviewers for their useful com-ments, which contributed to the improvement of our article. We acknowl-edge the assistance provided by GeoPro-Hamburg for the field operationsand the instruments used in the network. We thank Captain C. Handras andthe crew of the R/V AEGAIO for their support during the marine opera-tions. We specially thank Mr. D. Becker and Mr. S. Stoyanov from GeoPro-Hamburg for the data processing and preparation of figures. Our fieldoperations were financed by the Earthquake Planning and ProtectionOrganization of Greece.

References

Ambraseys, N. N., and J. A. Jackson (1990). Seismicity and associatedstrain of central Greece between 1890 and 1988, Geophys. J. Int. 101,663–708.

Armijo, R., B. Meyer, G. King, A. Rigo, and D. Papanastasiou (1996).Quaternary evolution of the Corinth rift and its implications for thelate Cainozoic evolution of the Aegean, Geophys. J. Int. 126, 1,11–53.

Collier, R. E., M. R. Leeder, P. J. Rowe, and T. C. Atkinson (1992). U-series dating and tectonic uplift of upper Pleistocene marine sedimentsfrom the Corinth and Megara basins, Greece, Tectonics 11, 1159–1167.

Freyberg, B. V. (1973). Geologie des Isthmus von Korinth, Erlanger Geol.Abh. 95, 183 p.

Fytikas, M., F. Innocenti, N. Kolios, P. Manetti, and R. Mazzuoli (1986).

Tectonic Deformation and Microseismicity of the Saronikos Gulf, Greece 927

Figure 6. A comparison among historical(Papadopoulos et al., 2000: 450 B.C. to 1889,light-gray circles; Makropoulos et al., 1989and Papazachos and Papazachou, 1997: 1914–1964, dark-gray circles and NOA bulletins1965–1999: black circles), (a), instrumentalreadings from NOA (b), and presently recordedseismicity (c) in the Saronikos Gulf and adja-cent areas.

928 J. Makris, J. Papoulia, and G. Drakatos

Figure 7. Location of events recorded within the Saronikos network. A minimumof six records per event has been used to locate each: earthquake. Events with an rmsvalue exceeding 0.5 sec have been discarded. Active faults identified by our observa-tions are marked in red. A discussion of the map is presented in the text.

The Plio-Quaternary volcanism of Saronikos area (western part of theAegean volcanic arc), Ann. Geol. Pays Hell. 33, 23–45.

Gaitanakis, P., and A. D. Dietrich (1995). Geological map of Methanapeninsula, scale 1:25 000, Swiss Federal Institute of Technology,Zurich.

Galanopoulos, A. G. (1960). A catalogue of shocks with Io � VI or M �

5 for the years 1801–1958, Seismological Laboratory, Athens Uni-versity, Athens, Greece, 1–119.

Galanopoulos, A. G. (1961). A catalogue of shocks with Io � VI for theyears prior to 1800, Seismological Laboratory, Athens University,Athens, Greece, 1–19.

Institute of Geology and Mineral Exploration (1989). Seismotectonic mapof Greece, scale 1:500,000.

Jackson, J., J. Gagnepain, G. Houseman, G. King, P. Papadimitriou, C.Soufleris, and J. Virieux (1982). Seismicity, normal faulting, and thegeomorphological development of the Gulf of Corinth (Greece): theCorinth earthquakes of February and March 1981, Earth Planet. Sci.Lett. 57, 377–397.

Jacobshagen, V. (1986). Geologie von Griechenland, Beitr. RegionalenGeologie der Erde, 363 p.

Kahle, H. G., M. V. Muller, A. Geiger, G. Danuser, S. Meuller, G. Veis,H. Billiris, and D. Paradisis (1995). The strain field in NW Greeceand the Ionian islands: Results inferred from GPS measurements, Tec-tonophysics 249, 41–52.

Kiratzi, A. A., G. F. Karakaisis, E. E. Papadimitriou, and B. C. Papazachos(1985). Seismic source parameter relations for earthquakes in Greece,Pageoph 123, 27–41.

Klein, F. (1989). User’s Guide to HYPOINVERSE, a program for VAXcomputers to solve for earthquake locations and magnitudes, U.S.Geol. Surv. Open-File Rep. 89-314.

Le Pichon, X., and J. Angelier (1979). The Hellenic arc and trench system:a key to the neotectonic evolution of the eastern Mediterranean area,Tectonophysics 60, 1–42.

Leydecker, G. (1975). Seismizitatsstudien im Bereich des Peloponnes auf-grund von Prazisionsherdbestimmungen, Ph.D. Dissertation, Univer-sity of Frankfurt, Frankfurt, Germany.

Makris, J., and L. Moeller (1990). An ocean bottom seismograph for gen-eral use. Technical requirements and applications, in Proceedings ofSymposium “Europe and the Sea,” Hamburg, J. Hoefeld, A. Mitzlaff,and S. Polomsky (Editors), sponsored by the EC.

Tectonic Deformation and Microseismicity of the Saronikos Gulf, Greece 929

Figure 8. Seismicity along a northeast-southwest-oriented profile with all foci con-tained in a 100-km-wide and 40-km-deep block, projected on the surface shown in thefigure. For further information, see text.

Makris, J., and P. Roever (1986). Struktur und heutige Dynamik der Lith-osphaere in der Agais, in Geologie von Griechenland, V. Jakobshagen(Editor), Gebruder Borntraeger, Berlin, 241–256.

Makris, J., J. Papoulia, G. Drakatos, and V. Karastathis (2002). Seismo-logical investigations in the Saronikos gulf, Athens, XXVII GeneralAssembly of the European Seismological Society, Nice (Abstracts).

Makropoulos, K. C. (1978). The statistics of large earthquake magnitudeand an evaluation of Greek seismicity, Ph.D. Thesis, University ofEdinburg, 1–193.

Makropoulos, K. C., and P. W. Burton (1984). Greek tectonics and seis-micity, Tectonophysics 106, 275–304.

Makropoulos, K. C., D. Drakopoulos, and J. Latousakis (1989). A revisedand extended earthquake catalogue for Greece since 1900, Geophys.J. Int. 98, 391–394.

McKenzie, D. P. (1978). Active tectonics of the Alpine-Himalayan belt:the Aegean Sea and surrounding regions, Geophys. J. R. Astr. Soc.55, 217–254.

Papadopoulos, G. A., and A. Arvanitides (1996). Earthquake risk assess-ment in Greece, in Earthquake Hazard and Risk, V. Schenk (Editor),Kluwer, Dordrecht, The Netherlands, 221–229.

Papadopoulos, G. A., A. Vassilopoulou, and A. Plessa (2000). A Catalogueof Historical Earthquakes for Central Greece: 480 B.C.–A.D. 1910,Institute of Geodynamics, National Observatory of Athens, Publ.no. 11.

Papanikolaou, D., V. Lykousis, G. Chronis, and P. Pavlakis (1988). A com-parative study of neotectonic basins across the Hellenic arc: the Mes-siniakos, Argolikos, Saronikos and Southern Evoikos Gulfs, BasinRes. 1, 167–176.

Papazachos, B. C., and P. Comninakis (1971). Geophysical and tectonicfeatures of the Aegean arc, J. Geophys. Res. 76, 8517–8533.

Papazachos, B. C., and N. D. Delibasis (1969). Tectonic stress field andseismic faulting in the area of Greece, Tectonophysics 7, 231–255.

Papazachos, B. C., and A. A. Kiratzi (1996). A detailed study of the activecrustal deformation in the Aegean and surrounding area, Tectono-physics 253, 129–154.

Papazachos, B. C., and C. Papazachou (1997). The Earthquakes in Greece,Ziti Ed., Thessaloniki, 304 pp.

Rotstein, Y. (1985). Tectonics of the Aegean block: rotation, side arc col-lision on crustal extension, Tectonophysics 117, 117–137.

Sieberg, A. (1932). Erdbebengeographie, Verlag von Gustav Fischer, Ber-lin, 102 pp.

Stiros, S. (1993). Kinematics and deformation of central and southwesternGreece from historical triangulation data and implications for the ac-tive tectonics of Aegean, Tectonophysics 220, 283–300.

Stothers, R. B., and M. R. Rampino (1983). Volcanic eruptions in the Med-iterranean before A.D. 630 from written and archeological sources,J. Geophys. Res. 88, 6357–6371.

Taymaz, T., J. Jackson, and D. McKenzie (1991). Active tectonics of thenorth and central Aegean Sea, Geophys. J. Int. 106, 433–490.

GeoPro GmbH20457 St. Annenufer 2Hamburg, Germany

(J.M.)

Institute of OceanographyHellenic Centre for Marine Research19013 AnavyssosAttica, Greece

(J.P.)

Geodynamic InstituteNational Observatory of Athens11810 ThissionAthens, Greece

(G.D.)

Manuscript received 9 October 2002.