Regional moment tensor determination in the European–Mediterranean area — initial results
Introduction
The location (hypocenter) and size (magnitude) of an earthquake are basic parameters in observational seismology. The distribution of earthquakes in terms of epicenter, depth, and size is also an important parameter for seismotectonic analyses and seismic hazard assessment (Giardini, 1999). The epicentral coordinates are usually relatively well determined by standard location procedures, while hypocenter depth and earthquake size are often only insufficiently well known. Moment tensor inversion (Gilbert, 1971) that uses the complete information contained in the seismograms can constrain hypocenter depth and seismic moment. In addition, moment tensor analysis reveals the deformation styles (fault plane solution).
For larger earthquakes globally (moment magnitude; Hanks and Kanamori, 1979, MW≥5.5), source parameters are routinely determined by the Harvard Centroid Moment Tensor (CMT) project Dziewonski et al., 1981, Dziewonski and Woodhouse, 1983 and by the United States Geological Service (USGS) Sipkin, 1982, Sipkin, 1986, Sipkin and Needham, 1993. These two methods use teleseismic waveforms and only stronger events that generate enough signals at distances several thousand kilometers from the source can be analyzed.
Analysis of moderate earthquakes (MW≥3.0) requires the use of local or regional data. At short epicentral distances, modern broadband seismic stations record complete seismograms with a high signal-to-noise ratio over a broad frequency band even for relatively small (MW≈3.0–4.0) earthquakes. At regional distances of up to about 2000-km epicentral distance, broadband stations can record MW≥4.5 earthquakes with a good signal-to-noise ratio particularly for the large-amplitude surface wave part of the seismograms. The growing network of broadband three-component stations (Fig. 1) offers the opportunity to routinely determine the earthquake source parameters (fault plane solution, source depth, and seismic moment) of moderate-to-strong earthquakes in the European–Mediterranean area.
Previous attempts to model European–Mediterranean earthquakes (e.g., Giardini et al., 1993, Braunmiller et al., 1994, Dufumier et al., 1997) using regional waveforms studied stronger events that in most cases could also be analyzed with teleseismic data. Here we determine moment tensor solutions for a large number of earthquakes too small for teleseismic analysis. Routine regional waveform analysis of moderate earthquakes is widely performed in the western United States Romanowicz et al., 1993, Thio and Kanamori, 1995, Braunmiller et al., 1995 where event–station distances commonly are on the order of hundreds of kilometers. In the European–Mediterranean area, average event–station distances are on the order of 1000–2000 km for several earthquake source regions, like northern Africa, Greece, and Turkey, hampering routine analysis. The only previous attempt for routine analysis covering the European area has been presented by Arvidsson and Ekström (1998).
Our goal is twofold. For larger earthquakes, we want to provide rapid moment tensor solutions within a few hours after an earthquake. Data sources are broadband stations accessible in near-real-time. The rapid solutions provide first reliable estimates of earthquake size, depth, and faulting style that could be of importance for disaster relief agencies, the general public, and Earth scientists. Our second goal is to build a moment tensor catalog for the entire European–Mediterranean area. The catalog should contain a large number of moderate events too small for teleseismic analysis. The source parameters have to be of high quality useful for detailed seismotectonic studies, seismic hazard analyses, quantification of earthquake size, and mapping of strain rates.
Section snippets
Data
Regional broadband seismograms are routinely available from several international (ORFEUS, Geofon, Geoscope, USGS, IRIS) and national (Austria, Germany, Israel, Switzerland) data centers and individual stations (TRI). Some data are rapidly available via AutoDRM Kradolfer, 1993, Kradolfer, 1996 (for example, from Austria, Czech Republic, Israel, and Switzerland), others are available within hours-to-days after a larger event (through ORFEUS, IRIS, Geofon, and ReNaSS), while data from most
Rapid analysis
As an example for rapid regional moment tensor analysis, we present our results for the MW=7.2 Dücze earthquake. The earthquake occurred in the early evening hours of November 12, 1999. The epicenter was located on the North Anatolian fault in western Turkey immediately east of the rupture zone of the devastating MW=7.5, August 17, 1999, Izmit earthquake. Several thousand people lost their lives in these two earthquakes and the destruction in the epicentral area was immense Delouis et al., 2000
Fault plane solutions
Moment tensor analysis covering only about 1 year of larger earthquakes already begins to reflect long-term seismicity in the European–Mediterranean area (Fig. 4). Most events occurred in Greece and western Turkey, the seismically most active areas in the study region (Jackson and McKenzie, 1988). Thrust-faulting mechanisms south of Crete and the Peloponnesus are associated with subduction at the Hellenic trench. Normal faulting events in Greece and southwestern Turkey are part of Aegean
Conclusions
The broadband seismic network in the European–Mediterranean area allows routine regional moment tensor analysis of moderate-to-strong earthquakes. Analysis of moderate (MW≈4.5–4.8) earthquakes is possible, because broadband stations can record such events with a good signal-to-noise ratio even at far regional distances of 1200–2000 km. At teleseismic distances signals are more attenuated limiting teleseismic source parameter determination to larger (MW≥5.0–5.3) earthquakes. Regional data thus
Acknowledgements
We thank the data centers and seismic station operators for providing high-quality broadband data that made this study possible. Namely, we thank the BGR (Germany), Geofon (Germany), Geoscope (France), GII (Israel), Gräfenberg (Germany), IRIS (USA), ORFEUS (The Netherlands), ReNaSS (France), USGS (USA), ZAMG (Austria) data centers, and the operators of station TRI (Italy). The people involved in setting up and running our own Swiss and the MIDSEA networks deserve our praise. Special thanks go
References (40)
- et al.
Preliminary reference Earth model
Phys. Earth Planet. Inter.
(1981) - et al.
Active crustal deformation from the Azores triple junction to the Middle East
Tectonophysics
(1995) Estimation of earthquake source parameters by the inversion of wave-form data: synthetic waveforms
Phys. Earth Planet. Inter.
(1982)- et al.
Moment–tensor solutions estimated using optimal filter theory: global seismicity, 1991
Phys. Earth Planet. Inter.
(1993) - et al.
Global CMT analysis of moderate earthquakes, MW≥4.5, using intermediate-period surface waves
Bull. Seismol. Soc. Am.
(1998) - et al.
The Swiss digital seismic network (SDSNet)
Orfeus Electron. Newsl.
(2000) - et al.
Earthquakes in Switzerland and surrounding regions during 2000
Eclogae Geol. Helv.
(2001) - et al.
Automatic regional moment tensor determination in the Mediterranean region
The complete synthesis of seismic crustal phases at regional distances
J. Geophys. Res.
(1982)- Braunmiller, J., 1998. Seismotectonics of the Explorer region and of the Blanco transform fault zone. PhD thesis,...
Source mechanism of the 1992 Roermond earthquake from surface-wave inversion of regional data
Geophys. J. Int.
The 1993 Klamath Falls, Oregon earthquake sequence: source mechanisms from regional data
Geophys. Res. Lett.
A deep earthquake under South Spain, 8 March 1990
Bull. Seismol. Soc. Am.
Earthquakes in Switzerland and surrounding regions during 1999
Eclogae Geol. Helv.
Joint inversion of InSAR and teleseismic data for the slip history of the 1999 Izmit (Turkey) earthquake
Geophys. Res. Lett.
Regional structure modelling and source inversion for the 1992 Roermond earthquake
J. Seismol.
An experiment in the systematic study of global seismicity: centroid moment-tensor solutions for 201 moderate and large earthquakes of 1981
J. Geophys. Res.
Determination of earthquake source parameters from waveform data for studies of global and regional seismicity
J. Geophys. Res.
The Global Seismic Hazard Assessment Program (GSHAP)—1992/1999
Ann. Geofis.
Moment–tensor inversion from Mednet data: 2. Regional earthquakes of the Mediterranean
Geophys. Res. Lett.
Cited by (95)
Towards fast focal mechanism inversion of shallow crustal earthquakes in the Chinese mainland
2024, Earthquake Research AdvancesRotational ground motion measurements for regional seismic moment tensors: A review
2021, Advances in GeophysicsInsights into the fragmentation of the Adria Plate
2016, Journal of GeodynamicsCitation Excerpt :Focal mechanisms (FMs) in the studied area are available from various global and regional catalogues based on seismogram modelling. We gathered focal mechanisms from: (i) the Global Centroid Moment Tensor project (GCMT, Dziewonski et al., 1981; Ekström et al., 2005), (ii) the National Earthquake Information Center (NEIC, Sipkin, 1982, 1994) of the U.S. Geological Survey, (iii) the Eidgenössische Technische Hochschule Zürich (ETHZ, Braunmiller et al., 2002), (iv) the European-Mediterranean Regional Centroid Moment Tensor (RCMT, Pondrelli et al. (1999, 2001, 2006, 2011) and quick RCMT on-line available at: http://autorcmt.bo.ingv.it/quicks.html) project of the Italian Istituto Nazionale di Geofisica e Vulcanologia (INGV), (v) the Time Domain Moment Tensor (TDMT, Scognamiglio et al., 2009 and quick TDMT on-line available at: http://cnt.rm.ingv.it/tdmt.html), and (vi) the database collecting first-pulse solutions in literature (the Earthquake Mechanisms of the Mediterranean Area − EMMA, Vannucci and Gasperini, 2003, 2004; Vannucci et al., 2010) with update of available data (unpublished report). For the Italian area the merged dataset of preferred FMs (when more than one focal solution is available for the same event, see for example Vannucci and Gasperini 2004 for selecting criteria) overall collects ∼4000 FMs data (Fig. 3).