« Website Guide »  Home » About Us » Stacey S. Martin » 2003 NGRI

About Us | Quake Alerts | Search

Hough, S.E., R. Bilham, K., Mueller, N., Ambraseys, and S. Martin, “Great Intraplate Earthquakes: Insights from Macroseismic Data”, Indo-US Workshop on Seismicity & Geodynamics, N.G.R.I., pp. 19-20, October 2003.

Abstract: To understand intraplate earthquake processes and genesis it is critical to first understand the earthquakes themselves. Macroseismic data provide the most direct constraint of magnitudes of important historic earthquakes in both India and the central/eastern United States. by virtue of their substantially improved spatial sampling, such data can also complement instrumental recordings of modern earthquakes. In this paper we present and discuss analysis of several key intraplate earthquakes in the U.S. and India; in the former region, the 1811-1812 New Madrid sequence, and, in the latter, the 1819 Allah Bund and 2001 Bhuj earthquakes. The shaking effects of the 1819 and 2001 events are found to be very similar, suggesting a similar magnitude for both events. Detailed analysis of teleseismic recordings of the Bhuj earthquakes yields Mw 7.6 for this event (Anatolik and Dreger, 2003). Detailed studies of both the 1819 and 2001 events, the Bhuj earthquake in particular, may help scientists answer vexing questions concerning the New Madrid sequence. The shaking effects from the Bhuj and the largest New Madrid earthquakes are also observed to be similar.

could help refine these estimates. (This would however, require a re-visitation of original accounts, to ensure systematic interpretation of macroseismic effects). The temporal clustering of the 1811-1812 New Madrid events as well as the Allah Bund/Bhuj pair highlights the importance of "earthquake interactions" in intraplate regions. We show that the New Madrid sequence, including present-day seismicity (which appears to represent a long-lived aftershock sequence) is consistent with the tenets of Coulomb stress transfer theory. In low strain-rate regions where mature plate boundary faults have not developed, the effects of static stress transfer may be particularly important in controlling the timing of major earthquakes.

Careful consideration of historic accounts also yields compelling evidence for the number of remotely triggered earthquakes in 1812: such events are attributed to the dynamic stress changes associated with S/surface waves from larges mainshocks. The observed triggered earthquakes in 1812 include several events large enough to be potentially damaging. We further propose that one of the (alleged) New Madrid mainshocks, on 23 January 1812, may itself be a remotely triggered earthquake, with a location near the location to the 1968 Illinois earthquake, which occurred on a blind thrust fault at 20-25 km depth. Intensity data for the 1812 event are consistent with expectations for a similarly deep event. Such triggered events presumably do not represent a wholly new source of hazard but rather a potential source of dependent hazard. That is, the common assumption is that the triggering will cause only a "clock advance" rather than causing earthquakes that would have otherwise occurred. However, here again, in a low strain-rate region, a given dynamic stress change can potentially represent a much large clock advance than the same change would cause in a high strain-rate region.

Recent observations from seismically active regions reveal that the most remotely triggered earthquakes are small events, in many cases too small to be observed without dense local monitoring networks. to understand the statistics and physical processes associated with remotely triggered earthquakes, good monitoring networks and catalogs are therefore essential. As mentioned above, some remotely triggered earthquakes are large, and these events can be important for another reason as well. In regions with low seismicity and a short historic period, overlooked remotely triggered earthquakes may be important events. It is thus possible that significant events are currently missing from the historic catalogs. Such events,

largest known historic earthquakes in three states (Kentucky, Illinois and Mississippi) may have been remotely triggered earthquakes. The southern Illinois, Wabash Valley source zone in particular emerges as a potentially more active zone than has been generally recognized.

Overall, our static stress change modeling and the remotely triggered earthquake results suggest that earthquake interactions - at a range of time and length scales - might be especially important in low strain-rate regions. This further suggests that seismic hazard assessment might need to take in to account the possibility of dependent events, that is, events that are much closer in time and/or space than would be generated in a Poissonian process.
 

Page Updated: 21 Feb 2008 | Website Disclaimer

© Amateur Seismic Centre, Pune,