Virtual Seismologist (VS)

The Virtual Seismologist (VS, Cua, 2005; Cua and Heaton, 2007; Cua et al., 2009) was conceived as a Bayesian approach to earthquake early warning (EEW) originally developed to estimate earthquake magnitude, location, and the distribution of peak ground shaking using observed picks and the earliest available ground motion amplitudes, based on predefined prior information (candidate priors include network topology or station health status, regional hazard maps, earthquake forecasts, and the Gutenberg-Richter magnitude-frequency relationship) and envelope attenuation relationships (Cua, 2005).  VS was envisioned to be a deterministic intelligent, automated algorithm with similar results to human seismologists’ quick and fairly accurate “back-of-the-envelope” interpretations of real-time (and at times, incomplete) earthquake information from both strong motion and high gain instruments, using a mix of experience, background information, and wave-physics principles. 

Development

The formulation of the VS Bayesian methodology, including the development of the underlying relationships describing the dependence of various channels of ground motion envelopes on magnitude and distance, and how these pieces come together in EEW source estimation, was the result of the PhD research (1998-2004) of Dr. Georgia Cua with Prof. Thomas Heaton at Caltech.

The first real-time VS prototype implementation was developed by Georgia Cua and Michael Fischer at ETH Zurich from 2006-2012. For this and all subsequent implementations, prior information is not included. This prototype used point source location estimates generated by the Earthworm Binder module (Dietz, 2002) as inputs to the VS magnitude estimation. This architecture underwent continuous real-time testing in California (2008-2018) and Switzerland (2010-2014). In California, VS was one of the three EEW algorithms that made up the original version of the CISN ShakeAlert EEW system.

VS was retired from ShakeAlert in 2018. In 2012/13, with funding from the EU projects NERA and REAKT, VS was integrated into SeisComP by the Seismic Network group at the SED (Behr et al., 2016) and gempa GmbH. Both real-time VS implementations (Binder- and SeisComP-based) focus on real-time processing of available pick and envelope data. Although the codes were effectively rewritten, the basic architecture used in the original Earthworm-based implementation is used in SeisComP. In the first SeisComP (Seattle) implementation, VS-specific pre-processing (scenvelope) and post-processing (scvsmaglog) modules were included, alongside the VS magnitude module (scvsmag). Later, scenvelope and scvsmaglog were replaced by generic EEW pre/post-processing modules. In SeisComP, the VS magnitudes, MVS, are typically computed on the preferred point source origins determined by scautoloc or scanloc. These standard locators can be optimised to be as fast as possible, i.e. with first solutions using a minimum number of picks, with all stations being used for triggering where needed configured with high triggering SNR ratios for noisy stations. VS magnitudes are available as soon as a single station has more than 1s of data after the P-wave, so there is usually no additional delay once the first origin is available. MVS is continuously updated every second using all available data at all stations included in the currently preferred origin.

Access and license

The VS EEW modules are free and open source, and are part of the SeisComP distribution from Seattle v2013.200. They were originally distributed under the SED Public License for SeisComP Contributions. VS is part the ETHZ-SED SeisComP EEW (ESE) system since 2021 under the GNU Affero General Public License (Free Software Foundation, version 3 or later).

The Earthworm implementation is no longer available or maintained.

VS is currently providing real-time alerts in Switzerland (Massin et al., 2021; Behr et al., 2012), Central America (Massin et al., 2022; Porras et al., 2021), and Taiwan. Feasibility studies using offline playbacks were carried out in Iceland, Greece, Romania and Türkiye.

Behr, Y., Cua, G., Clinton, J., Heaton, T. (2012). Evaluation of Real-Time Performance of the Virtual Seismologist Earthquake Early Warning Algorithm in Switzerland and California. Abstract 1481084 presented at 2012 Fall Meeting, AGU, San Francisco, Calif., 3-7 Dec.

Behr, Y. D., Cauzzi, C., Clinton, J. F., Jonsdottir, K., Comoglu, M., Erlendsson, P., et al. (2015). Exploring the Readiness for Earthquake Early Warning at Seismic Networks Across Europe. Seismological Research Letters, 86(2B), 738–739. http://doi.org/10.1785/0220150017

Behr, Y., J. F. Clinton, C. Cauzzi, E. Hauksson, K. Jónsdóttir, C. G. Marius, A. Pinar, J. Salichon, and E. Sokos (2016). The Virtual Seismologist in SeisComP: A New Implementation Strategy for Earthquake Early Warning Algorithms, Seismological Research Letters, March/March 2016, v. 87, p. 363-373, doi:10.1785/0220150235

Behr, Y., J. Clinton, P. Kästli, C. Cauzzi, R. Racine, M‐A. Meier (2015). Anatomy of an Earthquake Early Warning (EEW) Alert: Predicting Time Delays for an End‐to‐End EEW System, Seismological Research Letters, May/June 2015, v. 86, p. 830-840, doi:10.1785/0220140179

Cua, G. (2005). Creating the Virtual Seismologist: developments in ground motion characterization and seismic early warning. PhD thesis, California Institute of Technology, Pasadena, California.

Cua, G., and T. Heaton (2007). The Virtual Seismologist (VS) method: a Bayesian approach to earthquake early warning, in Seismic early warning, editors: P. Gasparini, G. Manfredi, J. Zschau, Springer Heidelberg, 85-132.

Cua, G., M. Fischer, T. Heaton, S. Wiemer (2009). Real-time performance of the Virtual Seismologist earthquake early warning algorithm in southern California, Seismological Research Letters, September/October 2009; 80: 740 - 747.

Dietz, L. (2002). Notes on configuring BINDER_EW: Earthworm’s phase associator, http://folkworm.ceri.memphis.edu/ew-doc/ovr/binder_setup.html (last accessed June 2013)

Massin, F., J.F. Clinton, and M. Böse (2021). Status of Earthquake Early Warning in Switzerland. Lausanne, Switzerland, Front. Earth Sci. 9, 707654. doi:10.3389/feart.2021.707654.

Massin, F., M. Böse, B. E. Burgoa, J. F. Clinton, G. Marroquin, C. Munoz, B. Orihuela, M. Protti, W. Strauch, and R. Yani (2022). Towards national public earthquake early warning systems across central America, AGU Fall Meeting, Chicago, IL, 12–16 December 2022, abstract NH34B-01.

Porras, J., F. Massin, M. Arroyo-Solórzano, I. Arroyo, L. Linkimer, M. Böse, and J. Clinton (2021). Preliminary Results of an Earthquake Early Warning System in Costa Rica. Front. Earth Sci., 23 September 2021. Sec. Solid Earth Geophysics 9. https://doi.org/10.3389/feart.2021.700843.