next up previous contents
Home: Berkeley Seismological Laboratory
Next: Waveform Constrained Seismic Velocity Up: Ongoing Research Projects Previous: Finite Fault Inversion of   Contents

Automated Moment Tensor Software for Monitoring the Comprehensive Test Ban Treaty

Margaret Hellweg, Douglas Dreger, Barbara Romanowicz, Jeffry Stevens (SAIC)

Introduction

Seismology makes an important contribution toward monitoring compliance with the Comprehensive Test Ban Treaty (CTBT). One task at the testbed of the Center for Monitorning Research (CMR, Washington DC, USA) and the International Data Center (IDC) of the Comprehensive Test Ban Treaty Organization (CTBTO, Vienna, Austria) is the detection, location and characterization of seismic events in order to distinguish between possible nuclear tests and earthquakes or other natural sources of seismic signals. While this is not particularly difficult for large events, whether natural or man-made, small events present a greater challenge. Although their epicenters and magnitudes can be determined fairly precisely, seismic moment tensor analysis can help in two ways. It not only gives information about the size and mechanism of a source in terms of its seismic moment and the moment tensor components. It provides, in addition, an estimate of the source's depth, which cannot always be reliably determined using normal location techniques. Thus, a large non double-couple component ($ > 50\%$) may be an indication for a nuclear explosion, as compared to the typically more than 70-80% double couple for an earthquake (Dreger and Woods, 2002). The source depth determined from moment tensor analysis may also help to weed out tectonic events from among the more than 100000 events of magnitude 4 and greater that occur annually. Only events at shallow depths need be scrutinized by the monitoring process of the Comprehensive Test Ban Treaty (CTBT).

This project's goal is to implement the process for automatic determination of seismic moment tensors routinely used in real-time at the University of California at Berkeley (UCB, Romanowicz et al., 1993; Dreger and Romanowicz, 1994; Pasyanos et al., 1996) on the testbed at CMR. Although the moment tensor process will not be running in real-time on the testbed, in its final implementation it will run automatically, triggered from the Reviewed Event Bulletin (REB). Thus, it will be an additional, potentially powerful event screening procedure (Pechmann et al., 1995; Dreger and Woods, 2002), providing estimates of: (1) the moment magnitude, $M_{w}$, a very accurate measure of event size; (2) the source depth, which will help distinguish natural events with typical depths greater than than 1 km from nuclear explosions and (3) radiation characteristics, such as deviations from the typically double-couple radiation of earthquakes.

Progress and Results

The automated procedure developed at UCB and implemented at CMR uses two methods for determining the moment tensor. One is a time domain, waveform fitting procedure that utilizes the complete, long-period recordings (CW, Dreger and Romanowicz, 1994; Pasyanos et al., 1996; Fukuyama et al., 1998; Fukuyama and Dreger, 2000). The other tensor method fits the surface waves in the frequency domain (SW). It is adapted from the two-step method of Romanowicz (1982).

Figure 29.1: Moment tensors solutions (dark, colored) determined using the CW method for events in Greece (Mainshock (evt1): Sep. 07, 1999, $M_{w}$ 6.0; evt2 $M_{w}$ 5.6 evt3 $M_{w}$ 4.8). Inverted triangles denote IMS primary stations, while triangles are auxiliary stations. The light gray solutions for evt1 and evt2 are taken from the Harvard CMT catalog
\begin{figure}\begin{center}
\epsfig{file=hell02_f01.eps, width=6cm}\end{center}\end{figure}

During the past year, we have completed the installation of the moment tensor codes on the testbed at CMR. The software package now automatically extracts event information and waveform data from the database there, performs basic quality control and preprocesses the waveforms before running the two inversions to produce independent solutions. As testing has proceeded, we have improved the Greens functions produced for the CW method by applying a flattening algorithm to the radially symmetric velocity structures (Müller, 1973, Müller, 1977). We have also adapted the set of periods used for the SW inversion from those used for the regional application in California to for application world-wide on intermediate-sized events.

We have applied the procedure to events shallower than 200 km with $m_{b} > 5.4$ in a test dataset, the 90 day interval between from July 19, to October 17, 1999. For the event in Greece on September 7, we have investigated the use of data from auxiliary stations of the IMS network in addition to the primary stations. Figure 29.1 shows results for the mainshock ($M_{w}$ 6.0) as well as two aftershocks (evt2 $M_{w}$ 5.6 and evt3 $M_{w}$ 4.8). Clearly, the method is effective in this region, even for the small aftershock.

For events in the test dataset we have run inversions using two different velocity models. The maps in Figure 29.2 show the IMS stations used for the inversions, as well as the moment tensor solutions determined by the complete waveform inversion and the surface wave method, respectively. In both Figure 29.2 A and B, the solutions derived using two different velocity models are compared with the moment tensors given in the Harvard CMT and USGS catalogs.

While the match between catalog source mechanisms and those calculated using the two automated moment tensor methods is good for some events, for others the process is not so successful. One typical problem is that for this interval, data is not always available from many of the primary stations of the seismic network of the International Monitoring Systems (IMS), the data source for the automatic process. For the CW method, for example, the moment tensors derived for events east and northeast of Australia differ from those given by both the Harvard and USGS catalogs. However, for each of these events, data were only available from one primary station less than 5000 km from the epicenter, STKA. The solutions calculated by the CW method are consistent with the waveforms from this station. The dearth of data is apparent for the SW method in in Figure 29.2 B which shows solutions for only 13 of the 19 events shown in Figure 29.2 A.

Perspectives

Currently, we are directing our efforts toward three fronts. First, we will attempt to improve the automated procedure by incorporating data from additional stations. Since 1999, the primary stations of the IMS network have been improved, both in their equipment and in their reliability. In addition, many of the auxilliary stations of the IMS network satisfy the need for the broadband, high dynamic range data which is necessary for the methods to work well. We will factor in data from these stations to improve the solutions. Secondly, we are working to develop and apply quantitative comparisons of the moment tensor solutions from various sources, CW or SW methods, as well as Harvard CMT and USGS. Finally, we are developing a regionalized calibration for the Far East. As part of an advanced concept demonstration, the CMR has collected event information and seismograms, as well as information about the Earth's structure in the region around Lop Nor. We will use this data to generate Greens functions and path information, so that we can calculate moment tensors for events, man-made or natural, occurring in this area.

Figure 29.2: (A) Map showing stations (triangles) and inversion results for the complete waveform method. Taken in order from the event hypocenter (dot), focal mechanisms are from the Harvard CMT catalog, USGS catalog, CW method using iasp91 velocity model and CW method using PREM. (B) Map showing stations (triangles) and inversion results for the surface wave method. Taken in order from the event hypocenter (dot), focal mechanisms are from the Harvard CMT catalog, USGS catalog, SW method using m1066b velocity model and SW method using PREM.
\begin{figure*}\begin{center}
\epsfig{file=hell02_f02.eps, width=12cm}\end{center}
\end{figure*}

Acknowledgements

This project is funded under the Defense Threat Reduction Agency contract DTRA01-00-C-0038.

References

Dreger, D. and B. Romanowicz, Source characteristics of events in the San Francisco Bay Region, USGS Open-file report, 94-176, 301-309, 1994.

Dreger D. and B. Woods, Regional Distance Seismic Moment Tensors of Nuclear Explosions, Tectonophysics, in press, 2002.

Fukuyama, E., M. Ishida, D. Dreger, and H. Kawai, Automated seismic moment tensor determination by using on-line broadband seismic waveforms, Jishin, 51, 149-156, 1998.

Fukuyama, E. and D. Dreger, Performance test of an automated moment tensor determination system, Earth Planets Space, 52, 383-392, 2000.

Müller, G., Theoretical Body Wave Seismograms for Media with Spherical symmetry; Discussion and Comparison of Approximate Methods, J. Geophysics, 39, 229-246, 1973.

Müller, G., Earth-Flattening Approximation for Body waves Derived from Geometric Ray Theory; Improvements, Corrections and Range of Applicability, J. Geophysics, 42, 429-436, 1977.

Pasyanos, M., D. Dreger, and B. Romanowicz, Toward real-time estimation of regional moment tensors, Bull. Seism. Soc. Am., 86, 1255-1269, 1996.

Pechmann, J.C., W.R. Walter, S.J. Nava, and W.J. Arabasz, The February 3, 1995, $M_{L}$ 5.1 seismic event in the Trona mining district of southwestern Wyoming, Seis. Res. Lett., 66, 25-34, 1995.

Romanowicz, B., Moment tensor inversion of long period Rayleigh waves: a new approach, J. Geophys. Res., 87, 5395-5407, 1982.

Romanowicz, B., M. Pasyanos, D. Dreger, and R. Uhrhammer, Monitoring of strain release in central and northern California using broadband data, Geophys. Res. Lett., 20, 1643-1646, 1993.



next up previous contents
Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94 720-4760
Questions or comments? Send e-mail: www@seismo.berkeley.edu
© 2002, The Regents of the University of California.