Exploring the Deep Sea and Beyond, Volume 2, themed issue Scholl et al. Great (≥Mw8.0) megathrust earthquakes and the subduction of excess sediment and bathymetrically smooth seafl oor David W. Scholl1,2, Stephen H. Kirby1, Roland von Huene1, Holly Ryan1, Ray E. Wells3, and Eric L. Geist3 1U.S. Geological Survey, Emeritus, Menlo Park, California 94025, USA 2Department of Geology and Geophysics, Emeritus, University of Alaska Fairbanks, Fairbanks, Alaska 99775, USA 3U.S. Geological Survey, Menlo Park, California 94025, USA ABSTRACT characteristic lengthy rupturing of high- to signifi cantly modify or arrest rupture con- magnitude IPT earthquakes. In these areas tinuation (Kodaira et al., 2000; Mochizuki et al., Using older and in part fl awed data, Ruff subduction of a weak sedimentary sequence 2008; Bilek, 2010; Singh et al., 2011; Wang and (1989) suggested that thick sediment enter- further enables rupture continuation. Bilek, 2011; Trehu et al., 2012; El Hariri et al., ing the subduction zone (SZ) smooths and 2013; Wang and Bilek, 2014), as commonly do strengthens the trench-parallel distribution INTRODUCTION subducting ridges (Franke et al., 2008; Sparkes of interplate coupling. This circumstance et al., 2010; von Huene et al., 2012; Kopp, 2013) was conjectured to favor rupture continua- Ruff (1989) observed that the entrance of an and also some upper plate structures (Bejar- tion and the generation of high-magnitude “excess quantity” of sediment into a lengthy Pizarro et al., 2013). (≥Mw8.0) interplate thrust (IPT) earth- subduction (≥230–300 km) sector of subduction With respect to the posited rupture-promot- quakes. Using larger and more accurate zone (SZ) favors the nucleation there of inter- ing effect of subducted sediment, Ruff (1989) compilations of sediment thickness and plate thrust (IPT) or megathrust earthquakes of lamented that, “the statistical correlation instrumental (1899 to January 2013) and magnitude Mw8.2 or greater (Fig. 1). He con- between excess sediments and great earthquake pre-instrumental era (1700–1898) IPTs (n = jectured that subducted sediment worked to occurrence is less than compelling.” His infer- 176 and 12, respectively), we tested if a com- both strengthen interplate coupling and smooth ence that a relation did exist was drawn from a pelling relation existed between where IPT the lateral or trench-parallel distribution of population of 19 instrumentally recorded earth- earthquakes ≥Mw7.5 occurred and where coupling strength, a circumstance that promotes quakes and a then-available but incomplete and thick (≥1.0 km) versus thin (≤1.0 km) sedi- rupture continuation and the consequent genera- largely inaccurate tabulation of the thickness of mentary sections entered the SZ. tion of high-magnitude or great (≥Mw8.0) IPT sediment entering SZs. After 1989 and through Based on the new compilations, a statisti- earthquakes. January 2013, an additional 16 great (≥Mw8.0), cally supported statement (see Summary and An “excess” quantity was considered a thick- fi ve giant (≥Mw8.5), and two super giant Conclusions) can be made that high-magni- ness adequate to nourish the building of an (≥Mw9.0) IPT earthquakes broke at 14 different tude earthquakes are most prone to nucleate accretionary frontal prism. This thickness is trench sectors (Tables 1–6). These more recent at well-sedimented SZs. For example, despite commonly estimated at ≥1 km (von Huene earthquakes include the re-seized 1946 Unimak, the 7500 km shorter global length of thick- and Scholl, 1991; Clift and Vannucchi, 2004; Alaska, or Scotch Cap megathrust (Lopez and sediment trenches, they account for ~53% Scholl and von Huene, 2007). Subducting sedi- Okal, 2006). Also since 1989, accurate tables of instrumental era IPTs ≥Mw8.0, ~75% ment enters the subduction channel (see Fig. 2, and maps of the global distribution of trench- ≥Mw8.5, and 100% ≥Mw9.1. No megathrusts panel C) that physically separates the upper and sediment thickness have become available (von >Mw9.0 ruptured at thin-sediment trenches, lower plate (Cloos and Shreve, 1988a, 1988b; Huene and Scholl, 1991; Scholl and von Huene, whereas three occurred at thick-sediment Moore et al., 2007; Collot et al., 2011). Mega- 2007; Heuret et al., 2012). trenches (1960 Chile Mw9.5, 1964 Alaska thrust rupturing occurs along the top or bottom To test the Ruff conjecture concerning the Mw9.2, and 2004 Sumatra Mw9.2). or within the subduction channel. rupture-promoting effect of sediment thickness However, large Mw8.0–9.0 IPTs com- In contrast, the subduction of bathymetrically alone (we did not evaluate a strengthening effect, monly (n = 23) nucleated at thin-sediment rough seafl oor would be expected to produce a notion that has been challenged by Wang and trenches. These earthquakes are associated an uneven or heterogeneous distribution of Bilek [2014]), we compiled improved and larger with the subduction of low-relief ocean fl oor coupling strength. This situation would condi- data sets of trench-axis sediment thickness and and where the debris of subduction erosion tion short-duration rupturing typical of lower vetted instrumental (1899 through January 2013) thickens the plate-separating subduction magnitude IPT earthquakes (see Ruff, 1989; era IPT earthquakes of magnitude ≥Mw7.5 (see channel. The combination of low bathymet- Fig. 3). It is now generally recognized that, Table 5) and pre-instrumental-era (1700–1898) ric relief and subduction erosion is inferred in fact, the subducted relief of seamounts and earthquakes of estimated magnitude ≥Mw8.0 to also produce a smooth trench-parallel seamount groups, although capable of localiz- (see Table 6). Descriptively simplifi ed catalogs distribution of coupling posited to favor the ing rupture initiation (Bilek et al., 2003), tend of instrumental-era IPT earthquakes of magni- Geosphere; April 2015; v. 11; no. 2; p. 236–265; doi:10.1130/GES01079.1; 16 fi gures; 8 tables. Received 25 May 2014 ♦ Revision received 28 October 2014 ♦ Accepted 4 February 2015 ♦ Published online 11 March 2015 236 For permissionGeosphere, to copy, contact April [email protected] 2015 © 2015 Geological Society of America Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/236/3333818/236.pdf by guest on 30 September 2021 on 30 September 2021 by guest Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/236/3333818/236.pdf Geosphere, April 2015 237 2015 April Geosphere, Megathrust earthquakes andsedimentsubduction earthquakes Megathrust Figure 1. Numbered trench sectors, blue for thin and red for thick sediment, identify where a Mw7.5 or greater instrumental interplate thrust (IPT) has been recorded. Sector numbers match those listed on Tables 1–4 for thin- and thick-sediment trench sectors. Sectors on these tables are compiled alphabetically, and numbered accord- ingly, by geographic area. Unnumbered thick-sediment sectors with horizontal striping identify where a ≥Mw7.5 IPT has not nucleated but where a future great (≥Mw8.0) megathrust is deemed likely. Not shown are the thickly sedimented (~7 km) Makran Trench of the northern Arabian Sea, and the equally sediment-fl ooded greater Hellenic and Gibraltar Trenches of the Mediterranean region (see also Kopp et al., 2000; Smith et al., 2013). Of these subduction zones, only the Makran has ruptured in an instrumentally recorded megathrust earthquakes ≥Mw7.5 (1945, Mw8.1; Tables 2 and 3). The great 1755 Lisbon earthquake of the Gibraltar region, which may have been an IPT, is estimated at >Mw8.5 (Gutscher et al., 2002; Thiebot and Gutscher, 2006). Scholl et al. Thick Trench Section (>1.0 km) A Aleutian Trench AccretionaryAccretionary frontalfrontal prismprism PacificPaci Plate fic P late ~2 km 10 km R/V Ewing, 1994 Thin Trench Section (<1.0 km) B 4 NE Japan (Tohoku) Trench 4 Frontal Cenozoic) Slope Apron ( Prism 6 ~0.6 km 6 8 ForearcForearc BasementBasement (Late(Late Cretaceous)Cretaceous) 8 PacificP Plate acific Plate 10 10 12 10 km 12 VonVon HueneHuene etet al.,al., 19941994 km km Trench Sediment Entering Subduction Channel Subduction Channel 2 2 C oic) enoz onon (Cenozoic)(C Apprr SlopeSlope A 4 Ecuador Trench 4 km ForearcForearc BasementBasement (Mesozoic)(Mesozoic) km SubductingSubd S NazcaNaz Plate ucting 6 ca Plate ediment 6 SubdSu bductionuct Cha ion Ch 10 km annelnnel 8 CollotCollot etet al.,al., 20022002 8 Figure 2. Seismic reflection images of: (A) a thick-sediment trench (see also Fig. 12), (B) a thin-sediment trench, and (C) a thick-sediment trench section entering the subduction channel (see also Fig. 12). tude ≥Mw7.5 (n = 176) and pre-instrumental Because our purpose was to test for a defi ni- Based on the instrumental data listed in megathrusts of magnitude ≥Mw8.0 (n = 12) are tive association of sediment thickness versus Tables 1, 2, and 3 and the plots of Figures 4–13, listed on Tables 1–4. In these tables, IPT earth- occurrence of high-magnitude megathrust earth- a statistically supported statement (see Sum- quakes are linked to an occurrence area along a quakes—either statistically strong, weak, or not mary and Conclusions) can be made that large sector of trench distinguished by a fl uctuating there—we only used the two thickness bins IPT earthquakes are most prone to nucleate at but average sediment fi ll that can be character- noted above and did not divide sectors further well-sedimented SZs. For example, despite the ized as either thick or thin. Along thick sectors into categories of very thin (>0.5 km) or very 7500 km shorter global length of thick-sediment the average fi ll is >1.0 km, whereas that for thin thick trench sequences (>2.0 km). Nonethe- trenches (thin = ~21,500 versus thick = ~14,000 sectors is <1.0 km. Trench sectors are shown on less, as described by Heuret et al. (2012), useful km), at thick-sediment trenches occurred ~53% Figure 1.