Konstantinos Makropoulos

Two arrays equipped with broadband sensors were installed for a period of 10 months, in order to study the seismic activity in the area of Santorini (Thira) volcano. During these periods, about 330 earthquakes were recorded and located within a radius of 50 km from the center of the caldera. An iterative damped traveltime inversion procedure yielded a local 1-D Ρ-wave velocity model and improved locations with an accuracy better than 5 Km in both horizontal and vertical components for 135 earthquakes. Those are mainly distributed within a depth range 5-18 Km, in the vicinity of the submarine Kolumbo Reef (NE of Santorini Island). Signal analysis of the recorded volcanic earthquakes including typical Fourier transformations and several operations in the time-frequency domain, allowed their dominant frequency determination and their classification into three groups based on waveform appearance and frequency content: (1) highfrequency events; (2) low-frequency events; and (3) volcanic tremor. Frequencytime analysis of tremor, detected at three stations, revealed two kinds of harmonic tremor with one sharp peak, at 3-5 Hz and 8.5-10 Hz.

Strong motion data recorded during the 15-year operation of the CORinth Gulf Soft Soil Array (CORSSA) in the highly seismic region of Aegion have been homogenized and organized in a MySQL database. In the present work we describe the contents of the database and the web portal through which these data are publicly accessible. CORSSA comprises one surface and four downhole 3-D broadband accelerometric stations. It was installed in 2002, in the framework of European project CORSEIS, aiming at gathering data for studying site effects, liquefaction, and non-linear behaviour of soils, as well as earthquake source properties. To date, the array has recorded 549 local and regional events with magnitudes ranging from 1.1 to 6.5. Although the vast majority of the recorded events caused weak ground motion at the CORSSA site, the scientific value of the data set pertains to the sparsity of this kind of infrastructure in most parts of the world.

The Gulf of Corinth, located in central Greece, is characterized by normal faulting and by high seismicity since the antiquity. On 4 February 2008 a seismic swarm burst in NW Peloponnesus characterized by the occurrence of two moderate-size earthquakes of moment magnitudes 4.7 and 4.5, respectively. Analysis of the data revealed the existence of shear-wave splitting. The events that were chosen for the anisotropy study fit the selection criteria, having clear and impulsive S wave arrival phases on the horizontal components. In addition, the amplitude of the S wave phase on the vertical component was smaller than on the horizontal ones. The representations that were used to determine the polarization direction of the fast split shear wave, the time delay between the two split shear waves and the polarization direction of the source were the polarigram and the hodogram. The uniform fast shear wave polarizations, irrespective of the azimuth of each event, are consistent with the general NNE-SSW direction of extension in the Gulf and, therefore, in agreement with the extensive dilatancy anisotropy (EDA) model. Finally, a decrease of the time delay values was observed after the occurrence of the first moderate earthquake, implying changes of the medium’s properties.

Ground motion prediction equations, widely known as attenuation relations, are common input for probabilistic and deterministic seismic hazard studies. The construction of a ground motion model to describe such a complex phenomenon as the effects of seismic wave propagation is highly dependable on a number of parameters. The quality and the distribution of strong motion data, which is the original input for the calculation of any ground motion model, can be thought as one of the main parameters that heavily influence the form of ground motion prediction equations. The selected processing scheme, involving significant choices about a series of adjustments and filter specifications, implemented to remove low and high frequency noise, is related with the credibility of the calculated ground motion parameters such as the spectral ordinates. Once a set of response variables for a number of predictors is available, the researcher’s interest is related with the mathematical definition of the ground motion model, in terms of selecting the appropriate parameters and the determination of their coefficients of the equation. Another significant part involves the selection of the optimum solver in order to achieve high confidence level coefficients and a computationally inexpensive solution. Each method should be evaluated through statistics but the researcher should bear in mind that residual analysis and statistical errors, although they can adequately represent the efficiency of the mathematical equations, do not always provide information about where our efforts should lie in terms of further improvement. The scope of this paper is to point out the multi-parametric nature of the construction of ground motion prediction equations and how each of the aforementioned development stages influences the credibility of the proposed attenuation relations.

The source parameters of all the moderate and strong events that occurred in the broader area of Zakynthos Island for the period 1997–2009 are determined using two different techniques depending on epicentral distance. For the strong events we employed teleseismic body-wave modeling between 30° and 90°, while for moderate events a regional moment tensor inversion approach was used. In both cases we used broadband waveform data with a good signal-to-noise ratio. The calculated focal mechanisms are representative of the displacement and strain fields of the broader area. Those associated with the Cephalonia transform fault are consistent with dextral strike-slip motion. The area between the island of Zakynthos and the Western coast of Peloponnese is also characterized by strike–slip faulting, while reverse faulting is mainly observed south of Zakynthos Island. Using both techniques, we also retrieved the source parameters of the biggest events of the seismic sequence of April 2006 near the southern coast of Zakynthos Island. The depths of the located events of the sequence vary mainly between 10 and 25 km. The fault plane solutions revealed thrust type faulting in all cases, something which can be related with upward motions in the southern part of the Island observed by DGPS measurements.

The National Observatory of Athens has begun installing permanent GPS stations on February 2006 including a EUREF permanent station in Attica, NOA1. Currently the National Observatory of Athens operates 20 continuous GPS stations around Greece all sampling at 1-s and transmitting real-time data to Athens. Several stations also sample at 0.2-s (5 Hz) and record the data in the ring buffer for a period of 1-2 days. Their location is carefully selected so that both geological and seismotectonic criteria are fulfilled. All stations are situated close to major seismogenic structures of Greece such as the Cephalonia Transform Fault (CTF) in the Ionian Sea (VLSM, PONT, SPAN, KIPO), and the two North Anatolian Fault branches in the North Aegean Sea (PRKV, LEMN ). We describe the CGPS data archiving and processing procedures, used to combine into a uniform velocity solution the observations of all the NOANET stations, accounting for the seasonal (annual and semiannual) signals, and considering the off-sets in the coordinate time-series.

The most important natural hazard in Greece is earthquake. The earthquake phenomenon can be explained using the theory of Plate Tectonics. Greece lies in the middle of the collision between two major tectonic plates, the Eurasian and the African, resulting in a very fragmented geotectonic regime. From the point of energy released, half of the European seismic energy is released within the Greek territory. Thus, the ways and means of reducing the seismic risk, that is the consequences from an earthquake, is for Greece of vital importance. The seismic risk is the convolution of the seismic hazard and the vulnerability of the specific area. From those factors, the vulnerability, which expresses the weakness or the sensitivity of the system and the value at risk during an earthquake, is the only parameter that can and should be minimized. The accomplishment of such an important task requires a combination of top-down and bottom-up approaches. In terms of the top-down approach, the Earthquake Planning and Protection Organisation’s, (E.P.P.O.), main target is to plan the national policy for earthquake protection, as well as to coordinate the public and private resources for the implementation of this policy, through issuing regulations, guidelines for emergency situation or for strengthening existing buildings including monuments of cultural heritage value e.t.c. EPPO also has a strong educational/training focus, targeting inter alia schools and hospitals. Of importance are also bottom up approaches, often at the personal level, which include useful measures concerning the proper behaviour before, during and after a destructive earthquake. These approaches are also part of the EPPO mandate and focus.

The Corinth Rift in Central Greece has been studied extensively during the past decades, as it is one of the most seismically active regions in Europe. It is characterized by normal faulting and extension rates between 6 and 15 mm yr⁻¹ in an approximately N10E° direction. On 2013 May 21, an earthquake swarm was initiated with a series of small events 4 km southeast of Aigion city. In the next days, the seismic activity became more intense, with outbursts of several stronger events of magnitude between 3.3 and 3.7. The seismicity migrated towards the east during June, followed by a sudden activation of the western part of the swarm on July 15th. More than 1500 events have been detected and manually analysed during the period between 2013 May 21 and August 31, using over 15 local stations in epicentral distances up to 30 km and a local velocity model determined by an error minimization method. Waveform similarity-based analysis was performed, revealing several distinct multiplets within the earthquake swarm. High-resolution relocation was applied using the double-difference algorithm HypoDD, incorporating both catalogue and cross-correlation differential traveltime data, which managed to separate the initial seismic cloud into several smaller, densely concentrated spatial clusters of strongly correlated events. Focal mechanism solutions for over 170 events were determined using P-wave first motion polarities, while regional waveform modelling was applied for the calculation of moment tensors for the 18 largest events of the sequence. Selected events belonging to common spatial groups were considered for the calculation of composite mechanisms to characterize different parts of the swarm. The solutions are mainly in agreement with the regional NNE–SSW extension, representing typical normal faulting on 30–50° north-dipping planes, while a few exhibit slip in an NNE–SSW direction, on a roughly subhorizontal plane. Moment magnitudes were calculated by spectral analysis of S waves, yielding b-values between 1.1 and 1.2 in their frequency–magnitude distribution. The seismic moment release history indicates swarm-like activity during the first phase, which could have acted as a preparatory stage for the second phase (after 12 July) that presented a more typical main-shock–aftershock behaviour. The spatiotemporal analysis reveals that the swarm has occurred in a volume that is likely related with the extension at depth of the NNE-dipping Pirgaki normal fault, outcropping ∼8 km to the south. The slow velocity of eastward migration of the epicentres during June implies triggering by fluids. The situation appears different in the second phase of the sequence, which was probably triggered by a build-up of stress during the first one. The relatively deep hypocentres of the 2013 swarm, compared to the shallower seismic layer within the rift, and their coincidence with the steeply dipping Pirgaki fault, favour an immature rift detachment model. Previous results from instrumental data indicate that approximately the same region had been activated during July–August 1991. The availability of the dense permanent seismological network data thus allowed for a detailed analysis of this crisis, a better understanding of its mechanical context and of the earlier events.