Teleseismic estimates of radiated seismic energy: The E/M₀ discriminant for tsunami earthquakes

Abstract

Abstract. We adapt the formalism of Boatwright and Choy for the computation of radiated seismic energy from broadband records at teleseismic distances to the real-time situation when neither the depth nor the focal geometry of the source is known accurately. The analysis of a large data set of more than 500 records from 52 large, recent earthquakes shows that this procedure yields values of the estimated energy, E oe, in good agreement with values computed from available source parameters, for example as published by the National Earthquake Information Center (NEIC), the average logarithmic residual being only 0.26 units. We analyze the energy-to-moment ratio by defining ©=1oglo(Eoe/Mo).For regular earthquakes, this parameter agrees well with values expected from theoretical models and from the worldwide NEIC catalogue. There is a one-to-one correspondence between values of {3 that are deficient by one full unit or more, and the so-called "tsunami earthquakes", previously identified in the literature as having exceedingly slow sources, and believed due to the presence of sedimentary structures in the fault zone. Our formalism can be applied to single-station measurements, and its coupling to automated real-time measurements of the seismic moment using the mantle magnitude Mm should significantly improve real-time tsunami warning.

Global evaluation of large earthquake energy from 1997 through mid-2010

“…[5] This study is built upon the methods described by Boatwright and Choy [1986] and Newman and Okal [1998], but utilizes advancements in computational code for more rapid and automated processing, and it is benefited by the vast improvement in the readily available broadband seismic data from globally distributed stations.…”

“…To maintain consistency with the previous catalog of Newman and Okal [1998], we limit the frequency range to between 14 mHz and 2 Hz (period range between 0.5 and 70 s). At frequencies above 2 Hz much of the signal falls below background noise levels, particularly for smaller events [e.g., Choy and Boatwright, 1995;Newman and Okal, 1998;Vassilou and Kanamori, 1982]. The lower frequency limit of 14 mHz is chosen because it corresponds to the time window of evaluation (70 s) of Newman and Okal [1998], and is maintained for catalog comparisons.…”

“…The square of the velocity seismogram u and frequency-dependent attenuation correction t* are integrated over the angular frequency w. Represented in seconds, t* is the reciprocal of the characteristic w representing the frequency by which an exponential decay of energy is 1/e its original value [Der et al, 1982]. The values of t* used in this study were introduced by Choy and Boatwright [1995] and put in equation form by Newman and Okal [1998].…”

“…However, by calculating the per station energy flux in the frequency domain, it is possible to directly correct for frequencydependent anelastic attenuation, obtaining a new theoretical station energy flux corrected for attenuation "* [Boatwright and Choy, 1986;Newman and Okal, 1998]:…”

“…Because events with the same M 0 can have considerable differences in their radiated energy, damage evaluations based solely on M 0 may inaccurately characterize the true damage caused by an earthquake. This is particularly the case for both high stress earthquakes that have disproportionately energetic shaking given their M 0 [Choy and Kirby, 2004], and tsunami earthquakes (TsE hereafter), which can be energetically deficient by more than a factor of ten relative to global averages of E/M 0 , due to their slow rupture [Newman and Okal, 1998]. To better characterize the global distribution of earthquake energy release, we develop a catalog of recent earthquake energies and their energy-to-moment ratio using available broadband seismic recordings at teleseismic distances for events of M 0 ≥ 10 19 Nm (M W ≥ 6.7), and include earthquakes up to 20 times larger in moment release than in previous systematic studies.…”