• journal_title：Bulletin of the Seismological Society of America
• Contributor：Han Yue ; Thorne Lay
• Publisher：Seismological Society of America
• Date：2013-05-01
• Format：text/html
• Language：en
• Identifier：10.1785/0120120119
• journal_abbrev：Bulletin of the Seismological Society of America
• issn：0037-1106
• volume：103
• issue：2B
• firstpage：1242
• section：Articles

The space–time history of fault slip during the 11 March 2011 Tohoku earthquake (M_w 9.0) is determined using high sample rate three‐component Global Positioning System (GPS) recordings from regional stations across Japan, teleseismic broadband P waves, global R₁ source time functions determined by empirical Green’s function deconvolutions of short‐arc Rayleigh waves, and ocean‐bottom deformation observations. Least‐squares inversions are performed for models with prescribed rupture‐front expansion velocity. Joint inversion yields improved resolution of slip compared with inversions using any single data type for both checkerboard rupture simulations and the actual data. Joint inversion stabilizes inversions with respect to some key parameters, such as the rupture expansion velocity and subfault total rupture durations, due to lower dependence on these parameters of some datasets (mainly the high sample rate [1  sample/s] three‐component GPS recordings [hr‐GPS] data). The preferred joint inversion model has a seismic moment estimate of 4.2×10²²  N·m (M_w 9.0), with a primary large‐slip patch with maximum slip of ∼50–60  m located up‐dip of the hypocenter on the shallow megathrust and distributed slip of 20–30 m near the hypocenter. A down‐dip low‐slip extension to the south is also resolved, with centroid source time later than 110 s.

Online Material: Figures of inverted slip distributions and associated source time functions and centroid times, and waveform fits.