CASCADING, COMPOUND, MULTI-FAULT RUPTURES: A NEW CLASS OF EARTHQUAKE

This presentation concerns the manner in which faults can interact over timescales of seconds in earthquakes. Compilations of historical earthquake rupture maps have led to a general consensus that fault segment boundaries of a few kilometres are sufficient to halt any earthquake rupture. However this inference has recently been challenged by new observations of compound, multi-segment earthquakes, which can near-instantaneously bridge fault gaps tens of kilometers wide. Here, I use the 27th February 1997 Harnai, Pakistan (M_w 7.1) earthquake – probably the earliest well-documented example of this new class of rupture – to illustrate. In this instance, dynamic stresses generated by seismic waves from the first fault rupture are the probable cause of slip initiation on the second fault. Other recent notable examples include the 14th November 2016 Kaikoura, New Zealand (M_w 7.8), 16th April 2016 Kumamoto, Japan (M_w 7.0), 7th December 2012 Sanriku-Oki, Japan (M_w 7.3), 11th April 2012 Indian Ocean (M_w 8.6), 2nd January 2011 Araucania, Chile (M_w 7.1), 29th September 2009 Samoa-Tonga (M_w 8.1), and 17^th June 2000 Reykjanes, Iceland (M_s 6.6) earthquakes. This rapidly growing list – which now includes subduction zone, fore-arc, ocean transform, continental, and intraplate examples – implies that cascading multi-fault earthquakes are a common and global phenomenon. Their recent discovery merely reflects advances in earthquake rupture imaging, in particular the now-pervasive application of Interferometric Synthetic Aperture Radar (InSAR) and the advent of the seismic back projection technique. These events expose a serious flaw in state-of-the-art earthquake rupture forecast models that disregard cascading, multiple-fault ruptures of this type.