Electronic Supplement to

Possible Triggered Seismicity Signatures Associated with the Vrancea Intermediate-Depth Strong Earthquakes (Southeast Carpathians, Romania)

 by Horia Mitrofan, Florina Chitea, Mirela-Adriana Anghelache, and Madalina Visan

 Seismic Data Selected for the Analysis

1. Utilized Catalogs and Cut-Off Magnitude Criteria

In selecting the events to be addressed by our analysis, we basically relied on data provided by the ROMPLUS (Oncescu et al., 1999) catalog. It has been prepared, in its initial version, from previous catalogs compiled for the Romanian territory, being subsequently continuously updated (it is currently available online at http://infp.infp.ro/catal.php). A conversion scheme (described in Oncescu et al., 1999) allowed magnitude values that previous catalogs had provided (until 1980) in various scales, to be homogenized and, if necessary, corrected, being ultimately reported in a single, moment magnitude (Mw) scale. In fact Mw also represents - since it is intrinsically related to the seismic moment (Kanamori 1978) - the most appropriate scale for constructing seismic release diagrams based on seismic moment computations.

ROMPLUS has been considered as the basic catalog in several previous studies addressing the overall seismicity of Vrancea area (e.g., Byrdina et al., 2006; Enescu et al., 2008; Mäntyniemi et al., 2003; Sokolov et al., 2009 ). According to its authors, ROMPLUS is complete for Mw ≥ 4.5 between 1936–1977 and for Mw ≥ 3.0 between 1978–1997. Yet by considering only events having occurred in the SE Carpathians at depths larger than 40 km, Byrdina et al. (2006) have established that ROMPLUS became homogeneous for Mw > 3.0 only since 1982, while since 1940 it was homogeneous for Mw > 5.0. In contrast, Enescu et al. (2008) stipulate that for events having occurred in Vrancea Seismic Zone (VSZ), at depths larger than 60 km, the completeness magnitude of ROMPLUS is around Mw =2.8 since 1974.

That apparently awkward issue, of selecting a cut-off magnitude which would be appropriate for the whole time-period (1977-1991) and over the entire depth-range considered by the present study, has been to a large extent overcome by choosing the seismic moment as the parameter to be used in our seismic release calculations: as discussed in the section “Outline of the Data Analysis Approach” of the main text, the shape of the corresponding diagrams would be dominated in that case by large and moderate earthquakes (Mignan et al., 2007); in particular, within the series of events occurred at intermediate-depths (≥60 km) in VSZ, the shocks with Mw > 6 will remain highly conspicuous, no matter which cut-off magnitude is selected – i.e. in the 3< Mw <4 range. In contrast, in the series of shallow (<60 km) Mărăşeşti-Galaţi-Brăila lineament (MGBL) earthquakes (with a maximum recorded Mw value of 4.4 – Table S1) the events larger than Mw ~3.5 will dominate; accordingly, rising the cut-off magnitude by more than a few tenths of a unit above Mw ~3 could significantly alter the profile of the corresponding seismic release diagram. We therefore considered for the MGBL domain an additional criterion, based on comparisons both with the reviewed ISC catalog (available at http://www.isc.ac.uk/iscbulletin/search/catalogue/), and with the PDE catalog of the NEIS/NEIC (available at http://earthquake.usgs.gov/earthquakes/eqarchives/epic/). As a result, it became apparent that over the 1977-1991 time-span, all the MGBL events with Mw ≥ 3.2 had hypocenter coordinates computed by one or another (occasionally even by both) of the indicated agencies; in contrast, not every MGBL earthquake with Mw <3.2 complied with this requirement.

We have consequently adopted for the MGBL events series addressed by our analysis (Table S1) a cut-off magnitude Mc=3.2. For uniformity, the same threshold magnitude was also adopted for constructing (Figures 2 and 3 in the main text) the seismic release diagram for the VSZ intermediate-depth domain. It is on the other hand worth noticing that outside the MGBL sub-zone, not all shallow (<60 km) earthquakes with Mw ≥ 3.2 (Table S2) had hypocenteral coordinates computed either by the ISC, or by the NEIS/NEIC agencies. Yet, still for uniformity, the Mc=3.2 criterion has nevertheless been used in constructing the seismic moment release diagram for the Ramnicu Sarat shallow seismicity sub-zone as well (Figure 3 in the main text).

Comparisons between the three utilized catalogs are displayed in detail in Tables S1 and S2.

The large majority of the shallow earthquakes included in Tables S1 and S2 have hypocentral coordinates computed by the ISC (being correspondingly tagged as such); accordingly, for each such event, the ISC catalog also provides detailed data concerning the seismic network utilized in the computations.

Focal depth errors (indicated, when available, in Tables S1 and S2) suggest that uncertainties in the location of a shallow earthquake hypocenter can be rather large; but at the same time, it definitely results that the considered shallow shocks did not occur below the 60 km depth boundary.

In the ROMPLUS catalog it is systematically indicated, for the concerned shallow events which occurred after 1 Aug. 1988, a magnitude estimate error of ± 0.1. Yet no other information is currently available in terms of magnitude errors. Still for certain shallow events it is possible to compare (Tables S1 and S2) the Mw values provided by the ROMPLUS catalog, with a rather abundant set of mb magnitude estimates provided by the PDE (NEIS/NEIC) catalog: occasionally, between the corresponding pairs of values there can be noticed significant discrepancies (due, mainly, to the different algorithms associated to the computations of Mw and mb respectively). It is not expected however that such differences could alter the general pattern of the seismic release diagrams, since the seismicity increases outlined by the ROMPLUS data include virtually all shallow events recorded by the PDE catalog (which implies, at the same time, that in-between those intensification episodes, no shallow events are recorded by PDE).

2. Comparing Shallow Vrancea Earthquakes Focal Mechanisms Determined by Various Authors

A fault plane solutions catalog addressing also shallow (<60 km) Vrancea earthquakes is that compiled from various sources by Radulian et al. (2002). Other rather comprehensive catalogs which include such shallow events are those of Mostrioukov and Petrov (1994) and Sandu and Zaicenco (2008). In Radulian et al. (2002) it is also indicated the total number of P-wave polarities used for computing each fault plane solution; in addition, both Mostrioukov and Petrov (1994) and Sandu and Zaicenco (2008) specify the number of stations for which polarities are inconsistent with the solution they adopted (Table S3).

No other information was yet available about errors involved by the above-indicated focal mechanism computations. We therefore attempted to check if various fault plane solutions independently estimated, for one and the same event, by different authors, resulted in a coherent stress field pattern. The events subject to such a general compliance between the main stress-axes orientations are listed in Table S3 and illustrated in Figure 1 of the main text.

[Tables]

[TableS1]. [ Major intermediate-depth VSZ earthquakes of the 1977-1991 period and the strongest events (Mw≥ 3.2) occurred in the MGBL shallow domain after each of them ]

[TableS2]. [ Major intermediate-depth VSZ earthquakes of the 1977-1991 period and the strongest events (Mw≥ 3.2) subsequently occurred at shallow depth outside the MGBL domain]

[TableS3]. [Shallow (<60 km) Vrancea earthquakes for which a reasonable compliance exists between the P and T axes orientations provided by various catalogs]

REFERENCES

Byrdina, S., P. Shebalin, C. Narteau, and J. L. Le Mouël (2006). Temporal properties of seismicity and largest earthquakes in SE Carpathians, Nonlinear Processes in Geophysics 13, 629–639.

Enescu , B., Z. Struzik, and K. Kiyono (2008). On the recurrence time of earthquakes: insight from Vrancea (Romania) intermediate-depth events,Geophysical Journal International 172 , 395-404.

Kanamori, H. (1978). Quantification of earthquakes, Nature 271, 411-414.

Mäntyniemi, P., V. I. Mârza, A. Kijko, and P. Retief (2003). A new probabilistic seismic hazard analysis for the Vrancea (Romania) Seismogenic Zone, Natural Hazards 29, 371-385.

Mignan, A., G. C. P. King, and D. Bowman (2007). A mathematical formulation of accelerating moment release based on the stress accumulation model, Journal of Geophysical Research 112, B07308; doi:10.1029/2006JB004671.

Mostrioukov, A. O., and V. A. Petrov (1994). Catalogue of focal mechanisms of earthquakes 1964-1990, Materials of the World Data Center, Moscow.

Oncescu, M., V. Marza, M. Rizescu, and M. Popa (1999). The Romanian earthquake catalogue between 984-1997, in Vrancea Earthquakes: Tectonics, Hazard and Risk Mitigation Friedemann Wenzel and Dan Lungu eds., Olaf Novak co-ed., Kluwer, Dordrecht, 43–47.

Radulian, M., E. Popescu, A. Bala, and A. Utale (2002). Catalog of fault plane solutions for the earthquakes occurred on the Romanian territory, Romanian Journal of Physics 47, 663-685.

Sandu, I., and A. Zaicenco (2008). Focal mechanism solutions for Vrancea seismic area, in Harmonization of Seismic Hazard in Vrancea Zone, NATO for Peace and Security. Series C: Environmental Security Anton Zaicenco, Iolanda Craifaleanu and Ivanka Paskaleva eds., Springer, Dordrecht, 17–46.

Sokolov, V. Yu., F. Wenzel, and R. Mohindra (2009). Probabilistic seismic hazard assessment for Romania and sensitivity analysis: A case of joint consideration of intermediate-depth (Vrancea) and shallow (crustal) seismicity, Soil Dynamics and Earthquake Engineering 29, 364-381.