1 Introduction

Past earthquake activity in and around Switzerland has been documented in an uninterrupted series of annual reports from 1879 until 1963 (Jahresberichte des Schweizerischen Erdbebendienstes). Three additional annual reports have been published for the years 1972–1974. These reports together with historical records of earthquakes dating back to the thirteenth century were summarized by Pavoni (1977) and provided the basis for the first seismic hazard map of Switzerland (Sägesser and Mayer-Rosa 1978). With the advent of routine data processing by computer, the wealth of data acquired by the nationwide seismograph network has been regularly documented in bulletins with detailed lists of all recorded events (Monthly Bulletin of the Swiss Seismological Service). Since 1996, annual reports summarizing the seismic activity in Switzerland and surrounding regions have been published in the present form (Baer et al. 1997, 1999, 2001, 2003, 2005, 2007; Deichmann et al. 1998, 2000a, 2002, 2004, 2006, 2008, 2009, 2010, 2011, 2012; Diehl et al. 2013). In the course of reassessing the seismic hazard in Switzerland, a uniform earthquake catalogue covering both the historical and instrumental periods was compiled in 2002 (Fäh et al. 2003). The official seismic hazard map of Switzerland based on this catalogue was released in 2004 (Giardini et al. 2004; Wiemer et al. 2009). In 2009, the Earthquake Catalogue of Switzerland was revised (ECOS-09) and is now available on-line (http://www.seismo.ethz.ch/prod/catalog/index). In addition, numerous studies covering different aspects of the recent seismicity of Switzerland have been published in the scientific literature (for overviews and additional references see, e.g. Deichmann 1990; Pavoni and Roth 1990; Rüttener 1995; Rüttener et al. 1996; Pavoni et al. 1997; Deichmann et al. 2000b; Kastrup et al. 2004; 2007; Husen et al. 2007; Marschall et al. 2013, Singer et al. 2014).

2 Data acquisition and analysis

2.1 Seismic stations in operation during 2013

The Swiss Seismological Service (Schweizerischer Erdbebendienst, SED) operates two separate nationwide seismic networks, a high-gain predominantly broad-band seismometer network (Table 4 in "Appendix") and a low-gain accelerograph network (Table 5 in “Appendix”). The former is designed to continuously monitor ongoing earthquake activity down to magnitudes well below the human perception threshold, whereas the latter is principally aimed at engineering concerns and thus focuses on recording ‘strong motions’ in urban areas. In addition, the SED operates a number of temporary stations for various projects (Table 6 in “Appendix”). SED stations with on-line data acquisition that were operational at the end of 2013 are shown in Fig. 1.

Fig. 1
figure 1

Seismograph stations in Switzerland with on-line data acquisition operational at the end of 2013. The stations defined as high-gain (HG) are mostly equipped with broad-band or 5-s sensors, whereas the strong-motion stations (SM) are accelerometers (see also Table 4, 5, 6 in “Appendix”)

Full size image

The ongoing densification of the Strong Motion accelerograph network with real-time continuous very-broadband accelerometers (Clinton et al. 2011; Cauzzi and Clinton 2013) continued. In 2013, eight new stations were installed (SBAW, SRHE, SLOP, SBAJ2, SRHH, SBERN, SBAV, SBAM2), which concludes the 30-station 4-year first phase of the two-phase project (Michel et al. 2014). Six of these stations have been installed in the city of Basel through a research project with the canton. Due to construction works, the broadband station at EMV (Emosson Dam) was dismantled in 2012. In 2013, a strong motion station SEMOS was installed near the original site of EMV.

The earthquakes induced by the geothermal project in Basel in 2006 have raised concerns about potential seismicity induced by other geothermal projects (even those that do not involve the enhancement of permeability through massive water injections). It is now common practice for responsible operators to include local monitoring capabilities for such projects. An array to monitor a geothermal experiment during the summer of 2013 near St. Gallen that was installed in 2012 was monitored throughout 2013, with some densification following the induced sequence (Fig. 2). This network was operated in cooperation with the St. Galler Stadtwerke and the Swiss Federal Office of Energy. In this network station SGT00 includes a borehole short period sensor at 205 m depth and a surface accelerometer located beside the drill site. Five additional surface broadband stations (SGT01, SGT02, SGT03, SGT04 and SGT05) were operated within a 10 km radius around the borehole landing point (Fig. 2). The four-station short period array around the southeast tip of Lake Geneva, associated to a project near Noville to drill for natural gas, was dismantled in the first half of 2013.

Fig. 2
figure 2

Station array for monitoring seismicity related to the St. Gallen geothermal project in the region of St. Gallen. Red triangles basic monitoring network with on-line data acquisition (see Table 6 in “Appendix”); green triangles mobile stations (SGT06-09: stations with on-line data acquisition, SGT10-12: stations without on-line data acquisition). The yellow square shows the location of the drill site

Full size image

In 2013 the construction of a seismic network that monitors the longterm seismicity down to magnitude ML 1.0 across north-eastern Switzerland was completed. This project, which began in 2011 under a contract with the National Cooperative for the Disposal of Radioactive Waste (Nagra), will improve the understanding of seismotectonic processes in the vicinity of proposed sites for deep nuclear waste repositories. In order to reach this goal, the existing network in northern Switzerland and southern Germany has being significantly densified with seven surface stations and three short-period borehole stations at depths of 100–200 m. These new stations will be in operation for a minimum of 10 years. In 2013, surface stations were installed at EMING, METMA and WALHA (all in Germany) and at DAGMA and ROTHE. Short period borehole sensors were installed at BOBI (154 m) and HAMIK (140 m), both with a surface accelerometer (Fig. 1).

To improve the reliability of locations for events at the periphery of or outside of Switzerland, the SED continues to be engaged in an ongoing cross-frontier cooperative effort to exchange seismic data in real-time. The SED continuously records and archives signals from stations in Austria operated by the Zentralanstalt für Meteorologie und Geodynamik in Vienna (ZAMG, 2 stations); in Italy operated by the Istituto Nazionale di Geofisica e Vulcanologia in Rome (INGV, 11 stations), the Istituto di Geofisica, Universita di Genova (6 stations), the Zivilschutz der Autonomen Provinz Bozen-Südtirol (6 stations) and the Istituto Nazionale di Oceanografia e di Geofisica Sperimentale (OGS) in Trieste (2 stations); in Germany operated by the Landeserdbebendienst Baden-Württemberg in Freiburg (LED, 3 stations) and the Bundesanstalt für Geowissenschaften und Rohstoffe in Hannover (BGR, 2 stations); and in France by the Réseau Sismologique et Géodésique Français (RESIF, 8 stations—3 strong motion). A total of 40 foreign stations were monitored at the SED in 2013, and the number continues to increase as new high-quality stations come on-line in the border region.

2.2 Hypocentre location, magnitude, focal mechanisms and monitoring software

Since 2005, hypocentre locations of the majority local earthquakes have been determined using the software package NonLinLoc (Lomax et al. 2000). The P-wave velocity model used was derived from a 3D tomographic inversion of local earthquake data with constraints from controlled source seismics (Husen et al. 2003), and the S-velocities are calculated from the P-velocity using a Vp/Vs ratio of 1.71. Since October 2012, the Seismic Network uses the SeisComP3 monitoring software for automatic event detection and characterization as well as manual review. More details on the SeisComP3 implementation are in the 2012 Annual Report.

Local magnitudes (ML) are calculated from the maximum amplitude of the horizontal components of the digital broad-band seismograms filtered to simulate the response of a Wood-Anderson seismograph. The attenuation with epicentral distance is accounted for by an empirically determined relation (Kradolfer and Mayer-Rosa 1988). The final magnitude corresponds to the median value of all on-scale station magnitudes, with stations within 15 km being neglected if possible. For the stronger events, the traditional determination of focal mechanisms from the azimuthal distribution of first-motion polarities (fault-plane solutions) is complemented by moment tensors based on full-waveform inversion. This procedure, based on a time domain inversion scheme developed by Dreger (2003), also provides a moment magnitude, MW, the best fitting double couple, and an optimal depth estimate based on the given location. An additional procedure has been implemented that routinely and automatically provides estimates of MW, including earthquakes of lower magnitudes. MW values are computed using a spectral fitting technique following the method of Edwards et al. (2010).

3 Seismic activity during 2013

3.1 Overview

During 2013, the Swiss Seismological Service detected and located 699 earthquakes in the region shown in Fig. 3. Based on criteria such as the time of occurrence, the location, and signal character or on direct communication, 208 additional seismic events were identified as quarry blasts. Magnitude values of the events recorded in 2013 range from ML −0.4 to 4.1 (Fig. 4). The events with ML ≥ 2.5 and the criteria used to assign the quality rating for the given locations as well as the corresponding estimated location accuracy are listed in Tables 1 and 2. Table 1 also includes the available MW values derived from the spectral fitting method of Edwards et al. (2010) and from method of Dreger (2003) when available. Fault-plane solutions based on first-motion polarities are shown in Figs. 5 and 6 (see also Fig. 3) and their parameters are listed in Table 3.

Fig. 3
figure 3

Epicentres and focal mechanisms of earthquakes recorded by the Swiss Seismological Service during 2013. Epicentres of events mentioned in the text are Balzers (Bal), Courrendlin (Cou), Hérémence (Her), Schönau (Sch), St. Gallen (StG), Wohlen (Woh). FL marks the territory of Liechtenstein

Full size image
Fig. 4
figure 4

Earthquake activity during 2013: a magnitude of each event and cumulative number of events (arrows mark the ML 3.5 St. Gallen and ML 4.1 Balzers event); b histogram of magnitudes

Full size image
Table 1 Earthquakes with ML ≥ 2.5
Full size table
Table 2 Criteria and location uncertainty corresponding to the quality rating (Q) of the hypocentral parameters in the event list in Table 1
Full size table
Fig. 5
figure 5

Fault-plane solutions based on first-motion polarities for four events in 2013 (Hérémence, Schönau, Courrendlin, Wohlen; see Table 3). All stereograms are lower hemisphere, equal-area projections. Solid circles correspond to compressive first motion (up); empty circles correspond to dilatational first motion (down). The take-off angles were computed with the NonLinLoc software (Lomax et al. 2000), using the 3D velocity model of Husen et al. (2003). Grey lines show sets of acceptable solutions derived by the HASH algorithm (Hardebeck and Shearer 2002); black bold lines indicate the (preferred) average focal mechanisms of all accepted solutions

Full size image
Fig. 6
figure 6

Fault-plane solutions based on first-motion polarities for three events in 2013 (upper row St. Gallen, Baltzers ML 4.1 mainshock, and Baltzers ML 3.7 aftershock; see Table 3). Symbols and explanation as in Fig. 5. For two of these events (lower row), the double-couple part of the full waveform moment tensor solution is shown as a beach-ball symbol (St. Gallen, Baltzers mainshock—items marked MT in Table 3). Black bold lines indicate the preferred first-motion solution of the same event

Full size image
Table 3 Focal mechanism parameters based on first-motion polarities and full moment tensor inversions (marked as MT) of 7 earthquakes in 2013 (see Figs. 5 and 6)
Full size table

Figure 7 shows the epicentres of the 911 earthquakes with ML ≥ 2.5, which have been recorded in Switzerland and surrounding regions over the period 1975–2013. These events represent about 8 % of the total number of events detected during that time period in the same area. The chosen magnitude threshold of ML 2.5 ensures that the data set is complete for the given period (Nanjo et al. 2010) and that the number of unidentified quarry blasts and of mislocated epicentres is negligible.

Fig. 7
figure 7

Epicentres of earthquakes with magnitudes ML ≥ 2.5, during the period 1975–2013. Grey circles denote earthquakes in the period 1975–2012, bold red circles indicate earthquakes in 2013

Full size image

3.2 Significant earthquakes of 2013

3.2.1 Hérémence

Although the magnitude reached only ML 2.7, the event at Hérémence, which occurred on April 6th at 23:15 UTC, was felt by many people in the Rhone Valley and south of it. About 150 felt reports were collected and the maximum observed intensity was IV. The focal depth of 7 km is well constrained by several stations at about 8 km epicentral distance and consistent with a source in the upper Penninic Nappes. The focal mechanism shown in Fig. 5 (see also Fig. 3) is well constrained by first motion polarities of P-phases and indicates an extensional regime in agreement with other focal mechanisms south of the Rhone Valley (e.g. Kastrup et al. 2004).

3.2.2 Schönau

The small event of ML 2.6 that occurred on June 8th near Schönau in Germany (Fig. 3) is another example of a lower-crustal earthquake in 2013. Its focal depth of 23 km is well constrained. The station BFO of the Black Forest Observatory in Schiltach, Germany, was included to improve the azimuthal coverage of the location and the focal mechanism. Because station BFO is not included in the routinely used model of Husen et al. (2003), the earthquake was located in the regional 3D model of Diehl et al. (2009). Although the earthquake is located outside of Switzerland, the distribution of first motion polarities constrains the focal mechanism reasonably well (Fig. 5). It is mainly a strike-slip mechanism with an almost E–W oriented T-axis (Table 3) and therefore typical for lower-crustal earthquakes in this region (e.g. Kastrup et al. 2004; Singer et al. 2014).

3.2.3 Courrendlin

The ML 3.2 earthquake of July 3rd with epicentre at Courrendlin (Fig. 3) occurred at a depth of 25 km in the lower crust beneath the Swiss Jura. It was felt by only few people in the region. Although strike and dip of the focal mechanism have large uncertainties, the distribution of first-motion polarities indicates a normal fault mechanism with a strike-slip component (Fig. 5) and orientations of P and T-axes consistent with other mechanisms in the region (e.g. Kastrup et al. 2004; Singer et al. 2014).

3.2.4 Induced seismicity in St. Gallen

Initiated by the utility company of St. Gallen, the St. Gallen geothermal projected was targeting existing fracture zones in the Mesozoic limestone layers of the Molasse sedimentary basin as a reservoir for geothermal energy. The targeted fracture zone, which was assumed to be seismically inactive, was imaged by a 3D seismic survey prior to drilling. Drilling operations were initiated in early 2013 and the target depth of 3,660 m (bls) was reached in early July 2013. Starting in July 2012 a dense monitoring network was installed by the SED (for detailed description of the SGT-network, see Sect. 2.1; Figs. 1 and 2) with the goal to provide fast information on potentially induced seismicity to the operator and the public. In parallel to the national detection pipeline, a special processing pipeline was setup in the SeisComP3 monitoring system, including all on-line SGT stations plus selected national high-gain stations in the region. The national as well as SGT processing pipelines were used for automatic triggering, locating, and alarming of potentially induced seismicity. STA/LTA triggers at six stations were required to generate an automatic location in the SeisComP3 system. To further lower the detection threshold, an additional semi-automated screening of signals was preformed. In Figs. 4 and 8 we included only events originally detected by one of the automatic SeisComP3 pipelines.

Fig. 8
figure 8

Cumulative number of events and magnitude of the induced earthquakes in St. Gallen as detected by STA/LTA trigger algorithms using on-line stations of the St. Gallen network (Fig. 2) complemented by stations of the national high-gain network in the region. The routinely derived network ML is overestimated in cases where magnitudes are determined from local SGT stations alone

Full size image

Seismicity initiated shortly after the first injection test was carried out on July 14th (Fig. 8). The following increase in the seismicity coincides with acid stimulations on July 17th. Starting around noon on July 19th, methane entered the borehole. During the following ‘well control’ sequence, the operators decided to pump cold water into the well in order to reduce the pressure buildup. At 7 p.m. local time on July 19th seismicity started to increase with a magnitude ML 2.1 event at 2:40 a.m. local time on July 20th. At 5:30 a.m. local time, the largest event of the sequence occurred, with a magnitude of ML 3.5 (MW 3.3). The ML 3.5 earthquake was followed by several ‘aftershocks’ and the seismicity finally ceased with the end of well-control phase on July 25th (Fig. 8). The increase in seismic activity starting in mid September coincides with well cleaning activities. Once production tests started on October 15, seismicity essentially stopped (Fig. 8). In total 105 events with magnitudes between ML −0.4 and 3.5 have been detected with the automatic system (Fig. 8). It should be noticed, however, that the routinely determined local magnitudes of small events in Fig. 8 are overestimated.

Macroseismic intensities of the ML 3.5 event reached degree IV on the European Macroseismic Scale EMS-98 (Fig. 9a). Compared to the induced event of Basel in 2006 the intensities reported for the St. Gallen event are typically one degree lower (Intensity IV vs. V) and the felt radius is smaller (e.g. Baer et al. 2007), even though it had a similar magnitude. The focal depth of the ML 3.5 event is about 4 km using Pg and Sg phases at distances <80 km in combination with the velocity model of Husen et al. (2003). The resolution of the routinely determined focal depth does not allow deciding whether the source is within the Mesozoic sediments or in the basement. First results of a more detailed analysis that is still underway indicate that the majority of events are located in the Mesozoic sediments. A moment tensor (MT) solution is available for the ML 3.5 event (Fig. 6), which indicates almost pure strike-slip motion. The MT is confirmed by the solution derived from the analysis of first-motion polarities (Fig. 6; see also Fig. 3). Nodal planes of both solutions agree very well. Preliminary results of a relative relocation study show that the left-lateral, NE–SW striking plane was the active plane.

Fig. 9
figure 9

Macroseismic intensities (EMS-98) of (a) the ML 3.5 St. Gallen event of July 20 and (b) the ML 4.1 Balzers mainshock of December 12. Stars indicate the epicentres of the events. Geographic labels: A Austria, D Germany, FL Liechtenstein, AR Arbon, AZ Appenzell, BR Bad Ragaz, BZ Bischofszell, CR Chur, SA Sargans, SG St. Gallen

Full size image

3.2.5 Wohlen

Another lower crustal earthquake of ML 2.7 occurred on November 3rd near Wohlen (Fig. 3). It was felt by a few individuals in the canton Aargau. The well-constrained depth derived by the routine location procedure is 30 km and therefore only about 2 km above the Moho at this location (Wagner et al. 2012). Although gaps in the azimuthal distribution of first-motions polarities lead to larger uncertainties of the computed nodal planes (Fig. 5), the solution is consistent with a strike-slip mechanism with an orientation typical for lower crustal earthquakes in this region (e.g. Kastrup et al. 2004; Singer et al. 2014).

3.2.6 Balzers

On December 12th an earthquake with an ML of 4.1 (MW 3.7) was located in the Rhine valley, close to the village of Balzers in southern Liechtenstein (Fig. 3). The ML 4.1 mainshock occurred at 00:59 a.m. UTC and was widely felt in the Alpine Rhine valley from Chur towards Lake Constance, throughout Liechtenstein and in adjacent areas of Switzerland and Austria (Fig. 9b). The event was strong enough to awaken a considerable portion of the population, and let suspended objects swing, china clatter, doors rattle, etc. More than 750 felt reports have been collected, but no confirmed damage to buildings or objects has been reported. The analyzed felt reports suggest an intensity of IV on the EMS-98 macroseimic intensity scale. Intensities of grade IV reported in the Linth valley in the canton Glarus (Fig. 9b) are presumably due to site effects.

Peak ground acceleration (PGA) reached maximum values of about 3 % g at the closest station to the epicentre, while the maximum recorded peak ground velocity (PGV) was about 0.4 cm s−1. The real-time strong-motion stations closest to the epicentre were three stations in the area of Buchs, SG, characterized by different site conditions. SBUB is located on hard rock on the Buchserberg, SBUA2 is located on the Altendorf alluvial fan bordering the Rhine basin, while SBUH is located in the center of the deep Rhine basin in Buchs. Figure 10a shows the acceleration seismograms of the ML 4.1 mainshock recorded at the three stations. The amplitude of the signal is larger for the stations on sediments (SBUH, SBUA2) compared to the hard rock site (SBUB). Site effects associated to basin response were also observed at the strong-motion station SCUG, installed in the Rhine basin in Chur. The response of the sedimentary basin is particularly apparent after integration of the acceleration waveforms to velocity, as shown in Fig. 10b. While the amplification of velocity values with respect to a rock site at comparable distance is moderate (LIENZ, Fig. 10b), the time histories at the two sites exhibit remarkable differences. At SCUG, the onset of relatively long-period waves of 0.5 s associated to the basin response (Michel et al. 2014) can be clearly appreciated in the horizontal components of ground-motions starting at about 10–11 s after origin. The duration of significant shaking at SCUG is significantly larger than at the rock site LIENZ, where only a few dominant cycles of ground motion can be read from the velocity traces (Fig. 10b).

Fig. 10
figure 10

Ground motions records of the ML 4.1 Balzers mainshock of December 12th (a) Restituted acceleration seismograms recorded at the national strong-motion stations SBUB, SBUH, and SBUA2 (NS-component, 11–13 km epicentral distance). Signals are filtered with a 0.33–100 Hz, 4th order, Butterworth band-pass filter. Seismograms of station SBUH and SBUA2 were offset by −0.2 and −0.4 ms−2, respectively. Note the larger amplitude at stations on sediments (SBUH, SBUA2) compared to the hard rock site (SBUB). b Velocity seismograms obtained after integration of the acceleration recording at station LIENZ (black; at 26 km distance) and SCUG (red at 23 km distance). Restitution to ground acceleration and band-pass filtering (0.33–100 Hz, 4th order, Butterworth) was applied before integration. North–South and Up–Down channels were offset by −0.001 and −0.002 ms−1, respectively. Note the relatively long-period waves starting at about 10–11 s after origin time at station SCUG, which are associated to the basin response

Full size image

The ML 4.1 mainshock was followed by a sequence of aftershocks with local magnitudes of locatable events ranging between ML −0.2 and 3.7. The largest ML 3.7 aftershock occurred on December 27th at 07:08 a.m. UTC and the aftershock seismicity finally ceased after mid March 2014. At the time of the ML 4.1 mainshock, the closest station of the SDSNet, PLONS, was not operational. The closest station was SBUB (Fig. 1) at a distance of about 11 km. Using all 143 manually picked first arriving Pg, Sg, and Pn phases results in a focal depth of about 4 ± 1 km. Using only Pg and Sg phase at distances <150 km results in a focal depth of 6 ± 1 km. The focal depth of 4 km seems mainly constrained by P phases at epicentral distances >150 km. Given the uncertainties in the computation of Pg and Pn ray paths and travel times for these distances (e.g. Diehl et al. 2013), the derived uncertainty of ±1 km might underestimate the true error of the absolute focal depth and the source of 6 km derived from the subset of picks is as plausible as the 4 km source.

To better constrain the focal depth of the aftershocks, a temporary seismic station (QSA01) was installed on the castle hill in Balzers in a distance of about 1 km from the mainshock on December 13th. At the same time the permanent station PLONS was again brought online. With the two additional near-by stations, the uncertainty of the focal depth of the ML 3.7 aftershock is reduced. The solution derived from all 120 first arriving P and S phases results in a focal depth of 6 ± 0.5 km. The imperfect knowledge of the velocity structure, especially at the closest stations (i.e. QSA01), might have a strong impact on the absolute focal depth. The top of the crystalline basement in this area is at a depth of about 6 km and it is overlain by a thick and complex sedimentary sequence (Pfiffner and Hitz 1997). Overestimating the velocities of the sedimentary sequence at QSA01 would result in an overestimated focal depth. Assuming an average P-wave velocity of 5.0–5.3 km/s for the medium above the hypocentre (e.g. Ye et al. 1995) and a Vp/Vs ratio in the range of 1.70–1.8, the S–P time of 0.9 s measured at station QSA01 for the ML 3.7 aftershock is consistent with a depth between 5.6 and 6.8 km. Thus the focal depth of 6 ± 0.5 km derived in the 3D model of Husen et al. (2003) is in good agreement with the observed S–P time at QSA01. A depth of 4 km, as derived for the ML 4.1 mainshock, would require an average Vp < 5 km/s, and/or Vp/Vs > 1.8. Figure 11 shows a comparison of waveforms of the ML 4.1 mainshock with the ML 3.7 aftershock at three stations. The high degree of similarity between the seismograms, especially the S–P and PmP–Pg times, suggests that both events have similar source locations and focal mechanisms. It seems therefore plausible that both events occurred in a depth of about 6 km and we prefer the depth of 6 km derived from the subset of picks as the depth of the mainshock (Table 1).

Fig. 11
figure 11

Displacement seismograms of the ML 4.1 Balzers mainshock (red), the ML 3.7 Balzers aftershock (black), and the ML 2.9 Bad Ragaz event of 2005 (grey) recorded on vertical components at the national high-gain stations WILA, SLE, BALST (see Fig. 1). The signals are filtered with a 1–30 Hz, 2nd order, Butterworth band-pass filter. The time window in the upper panel includes P and S waves, the time window in the lower panel is focused on the P wave. Numbers next to the seismograms indicate the epicentral distance. The high degree in similarity between the ML 4.1 mainshock and the ML 3.7 aftershock, especially the S–P and PmP-Pg times, suggests that both events have similar source locations and focal mechanisms. The general similarity with the Bad Ragaz event suggests a similar source region for both sequences

Full size image

The remaining uncertainties in the absolute focal depth (at least ±0.5 km) and in the geological interpretation of the seismic data of Pfiffner and Hitz (1997), however, do not allow deciding whether the earthquakes occurred in the crystalline basement or in the overlying sediments. Between 2000 and 2005 a sequence of earthquakes occurred close to Bad Ragaz, just about 3 km south of the 2013 Balzers sequence. Focal depths reported for these events vary between 7 and 8 km and the impulsive signal character of P and S waves as well as the clear records at large distances have been interpreted as qualitative indications for a source in the crystalline basement (Baer et al. 2001; Deichmann et al. 2006; Marschall et al. 2013). Figure 11 shows a comparison of the two strongest Balzers events with the 2005 ML 2.9 Bad Ragaz event. The similarity of the general signal character as well as PmP–Pg times suggest a similar source region for both sequences. Both sequences seem to locate in the uppermost part of the basement, but with the given uncertainties a source in the Mesozoic sediments, as discussed for the St. Gallen sequence before, cannot be entirely excluded.

The MT solution derived for the ML 4.1 mainshock indicates a strike-slip mechanism with practically N–S and E–W striking nodal planes (Figs. 3, 6; Table 3). The solution derived from the analysis of first-motion polarities generally agrees with the MT, but strike and dip of the nodal planes show uncertainties due to gaps in the azimuthal distribution (Fig. 6). The clear upward motion observed at station SLUB (Fig. 1) falls within the extensional part of the MT solution (Fig. 6). This discrepancy is most likely related to a modeling error in the take-off angle calculation. No MT solution was derived for the ML 3.7 aftershock, but the two additional near-by stations QSA01 and PLONS significantly reduce the uncertainty of the first-motion mechanism (Fig. 6). The solution agrees very well with the MT solution of the mainshock and, in combination with the similarity of waveforms (Fig. 11), suggests an almost identical slip direction.

Both mechanisms are very similar to the two mechanisms derived for the Bad Ragaz events in 2000 and 2005 (Baer et al. 2001; Deichmann et al. 2006). Here, the N–S striking plane with left-lateral motion was identified as the active plane by relative relocations of fore- and after-shocks (Marshall et al. 2013). To resolve the active plane of the Balzers sequence we relocated 62 fore- and after-shocks in the period between December 2013 and August 2014 using double-difference techniques in combination with time-domain waveform cross-correlation (Waldhauser and Ellsworth 2000). The relocated hypocentres form an E–W striking lineament that extends roughly 1.4 km horizontally (Fig. 12) and thus identifying the E–W striking nodal plane with dextral strike-slip motion as the active plane. The strike of the active planes of the Bad Ragaz sequence and the Balzers sequence are almost perpendicular to each other and it is therefore implausible that the two sequences are the expression of one major N–S striking fault in the Rhine valley.

Fig. 12
figure 12

Relative relocations of the Balzers sequence using double-difference techniques in combination with waveform cross-correlation. The sequence comprises 62 earthquakes recorded between December 9th 2013 and August 25th 2014. Bars indicate relative location errors; colors indicate origin time of the events. The star indicates the epicentre of the foreshock of December 9th, 2013. Grey triangle shows the location of the temporary seismic station QSA01 installed during the day after the mainshock

Full size image

3.2.7 Seismicity associated with the former deep heat mining project in Basel

The seismic activity induced by the geothermal project in Basel in 2006 and 2007 (e.g. Deichmann and Ernst 2009; Deichmann and Giardini 2009) continued to decrease over the years 2008–2011, with only one event (ML 0.6) being recorded by the national network in 2010. In 2012 activity seemed to have picked up again: five events, with magnitudes between ML = 0.9–1.2, were associated with the stimulated rock volume (Diehl et al. 2013). Master-event relocation confirmed that in 2013 two events (on June 29th and December 28th) occurred at the southern edge of the activated rock volume (Deichmann et al. 2014). The ML 0.8 event of December 2013 occurred within about 100 m of the three events recorded in December 2012, whereas the hypocenter of the ML 1.7 event of June 29th was about 300 m shallower and about 200 closer to the borehole (Deichmann et al. 2014). However, both events can be regarded as belonging to the sequence that has been active since 2007 and is located above the casing shoe at the southern periphery of the main microseismic cloud.

4 Discussion

In 2013, as in previous years, a large portion of the seismic activity was concentrated in the Valais, Graubünden, and the immediately adjacent regions. Routinely calculated focal depths for all but 23 events recorded in 2013 are <16 km. All but two of the deeper hypocentres occurred below the Molasse Basin and the Jura of northern Switzerland and southern Germany, with the main activity clustered between 7°E and 10°E (e.g. Deichmann 1992; Deichmann et al. 2000a; Singer et al. 2014). The other two events in 2013 occurred at about 20 km depth beneath the southern margin of the Alps in northern Italy. Both solutions have considerable azimuthal gaps and locations are poorly constrained. The ML 1.5 event of October 4th, however, locates just 9 km to the southwest of the ML 3.8 Erba earthquake of 2001, which represents an example for a well-constrained lower crustal earthquake at the southern margin of the Alps in northern Italy (e.g. Deichmann et al. 2002).

Overall, the seismic activity in and around Switzerland, in terms of the number of events, appears exceptionally high in 2013. The 699 located earthquakes in 2013 represent the highest number of earthquakes since the year 2004 (about 670 located earthquakes). The number is partly explained by the two earthquake sequences in St. Gallen and Balzers and removing the 105 induced earthquakes of St. Gallen results in a number closer to the 13-year average of 497 earthquakes per year. The exceptional high number of 208 located quarry blasts (more than twice the 13-year average of 90 blasts per year) might reflect changes in the detection threshold due to network enhancement or, alternatively, changes in the earthquake analysis procedures. The total number of 18 events with ML ≥ 2.5, on the other hand, is still slightly below the yearly average of about 23 events over the previous 38 years in this magnitude range.