Comprehensive Test Ban Monitoring: Contributions from Regional Moment Tensors to Determine Source Type and Depth

Margaret Hellweg, Douglas Dreger and Barbara Romanowicz


Regional distance moment tensor (MT) analysis can be used, even for relatively small magnitude events (M$>$3.4), to discriminate explosions from naturally occurring earthquakes. For earthquakes, MT analysis can also provide insight into an event's size, depth and type of faulting. Mechanism information is also important for applications like assessing earthquake effects and tsunami warning. In this project, we apply the UC Berkeley full moment tensor code (Minson and Dreger, 2008) to a suite of events recorded regionally at broadband seismic stations operated by the International Monitoring System (IMS).

Figure 2.56: Map showing the events chosen for analysis (stars). Certified IMS stations (as of April 2009) are shown as inverted triangles. The mechanisms shown for the largest events are from the GCMT catalog. The tables give the magnitudes of the selected events.
\epsfig{file=hell09_2_1.eps, width=8cm}\end{center}\end{figure}

Events and Green's Function

For regional moment tensor analysis using the full moment tensor code, it is important to have data from three or more stations with good signals in the period band between 10 s and 100 s or more, depending on the size of the event. Typically, for events with M $<$ 4, we use the period band 10 s $-$ 50 s, for 4 $<$ M $<$ 5 the band 20 s $-$ 50 s, and for 5 $<$ M $<$ 7 the band 20 s $-$ 100 s.

Reviewing the distribution of primary IMS stations certified by April 2009, we chose the region from 30$^{\circ}$E $-$ 70$^{\circ}$E and from 20$^{\circ}$N $-$ 45$^{\circ}$N, for its seismicity and IMS station availability. From the many events in the Reviewed Event Bulletin (REB), we selected two sequences from Western Iran (Figure 2.56) for two reasons: (a) Their epicenters were only about 20 km apart, so propagation paths to regional stations would be similar, and (b) the mechanisms of the largest events, as given in the GCMT catalog (Figure 2.56), were different. Magnitudes as given in the REB and in the PDE catalog (NEIC) are shown in Figure 2.56 for the mainshocks and other selected events. Of the approximately 20 events in each sequence, five were chosen from the REB, spanning a range of depths and magnitudes (Figure 2.56).

Green's functions were calculated for two 1D velocity models. The first is a generic 1D global model, iasp91. The second 1D velocity model is adapted from the Pasyanos, et al. (2004) 3D model for the region. For the chosen events, seismic moment magnitude (M$_w$), depth and source mechanism will be estimated using both types of Green's functions.

Ongoing Work and Perspectives

We received waveform data for the events from the IMS shortly before the ISS09 meeting. Unfortunately, no waveforms were available for stations at distances smaller than 2000 km for any events in 2002. Many of the waveforms for the 2006 events have gaps, and poor signal-to-noise ratios are particularly a problem for the small events. Based on the review of the data, we hope to be able to determine moment tensors for several 2006 events using data from the IMS. We will review recent seismicity in the region to determine whether other events have occurred which could contribute to this study.


The work performed for this project was supported by LLNL CTBT Contract B583480. Travel of MH to the ISS09 in Vienna, Austria, to present the interim results was funded by NSF Grant EAR-0926120.


Minson, S. and D. Dreger, Stable Inversions for Complete Moment Tensors, Geophys. Journ. Int., 174, 585-592, doi:10.1111/j.1365-246X.2008.03797, 2008.

Pasyanos, M.E., W.R. Walter, M.P. Flanagan, P. Goldstein and J. Bhattacharyya, Building and Testing an a priori Geophysical Model for Western Eurasia and North Africa, Pure appl. geophys., 161, 235281, 2004.

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail:
© 2007, The Regents of the University of California