Missing Great Earthquakes
Abstract
The occurrence of three earthquakes with Mw greater than 8.8, and six earthquakes larger than Mw8.5, since 2004 has raised interest in the long-term rate of great earthquakes. Past studies have focused on rates since 1900, which roughly marks the start of the instrumental era. Yet substantial information is available for earthquakes prior to 1900. A re-examination of the catalog of global historical earthquakes reveals a paucity of Mw ≥ 8.5 events during the 18th and 19th centuries compared to the rate during the instrumental era (Hough, 2013, JGR), suggesting that the magnitudes of some documented historical earthquakes have been underestimated, with approximately half of all Mw≥8.5 earthquakes missing or underestimated in the 19th century. Very large (Mw≥8.5) magnitudes have traditionally been estimated for historical earthquakes only from tsunami observations given a tautological assumption that all such earthquakes generate significant tsunamis. Magnitudes would therefore tend to be underestimated for deep megathrust earthquakes that generated relatively small tsunamis, deep earthquakes within continental collision zones, earthquakes that produced tsunamis that were not documented, outer rise events, and strike-slip earthquakes such as the 11 April 2012 Sumatra event. We further show that, where magnitudes of historical earthquakes are estimated from earthquake intensities using the Bakun and Wentworth (1997, BSSA) method, magnitudes of great earthquakes can be significantly underestimated. Candidate 'missing' great 19th century earthquakes include the 1843 Lesser Antilles earthquake, which recent studies suggest was significantly larger than initial estimates (Feuillet et al., 2012, JGR; Hough, 2013), and an 1841 Kamchatka event, for which Mw9 was estimated by Gusev and Shumilina (2004, Izv. Phys. Solid Ear.). We consider cumulative moment release rates during the 19th century compared to that during the 20th and 21st centuries, using both the Hough (2013) compilation and the Global Earthquake Model (GEM) catalog released in June, 2013. The GEM catalog includes three 19th century earthquakes of M8.5 and three M8.4s, and no 19th century earthquakes larger than 8.5. Cumulative moment release rates are notoriously difficult to estimate, but using the Hough (2013) compilation the 19th century moment release rate appears to be roughly half of the rate during the instrumental era; using the GEM catalog the 19th century rate appears to be roughly ¼ the instrumental rate. Thus, either 1) the global moment release rate varies by a factor of two or more on century time scales, or 2) the best available historical catalogs significantly underestimate great earthquake magnitudes and overall moment release rates. One can also consider whether magnitudes of great earthquakes were systematically underestimated during the first half of the 20th century, prior to the advent of long-period seismometry. We consider whether the 19th century moment release rate can be made consistent with the rate during the instrumental era using individual event magnitudes within the uncertainties estimated by past published studies. Lastly we consider the expected variability in global moment release rate, assuming a linear b-value up to Mmax9.5 and a Poissonian rate.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFM.S32A..05H
- Keywords:
-
- 7215 SEISMOLOGY Earthquake source observations;
- 7212 SEISMOLOGY Earthquake ground motions and engineering seismology;
- 7240 SEISMOLOGY Subduction zones