Open AccessTsunami waveform inversion incorporating permanent seafloor deformation and its application to tsunami forecasting
Author(s) -
Tsushima Hiroaki,
Hino Ryota,
Tanioka Yuichiro,
Imamura Fumihiko,
Fujimoto Hiromi
Publication year - 2012
Publication title -
journal of geophysical research: solid earth
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2011jb008877
Subject(s) - seafloor spreading , geology , tsunami earthquake , seismology , inversion (geology) , waveform , amplitude , geodesy , deformation (meteorology) , trench , arrival time , tide gauge , tectonics , geophysics , sea level , oceanography , physics , chemistry , organic chemistry , layer (electronics) , quantum mechanics , voltage , transport engineering , engineering
We propose a method of tsunami waveform inversion to accurately estimate a tsunami source by incorporating the effect of permanent seafloor deformation recorded by ocean‐bottom pressure gauges (OBPGs) within the source region. We developed a general expression of water‐depth fluctuation recorded at an OBPG following seafloor deformation of arbitrary spatiotemporal distribution. By assuming that coseismic rupture propagates with infinite velocity, the general expression can be reduced to an equation relating observed OBPG waveforms to initial sea‐surface displacement at the source by using a Green's function consisting of two terms: the Green's function used in regular tsunami inversion and a correction term to account for water‐depth change in response to permanent seafloor deformation. By using the two‐term Green's functions, the effect of seafloor deformation can be taken into account in tsunami source estimation. We applied the revised inversion method to observations of coseismic seafloor deformation and tsunami during the 2003 Tokachi‐oki earthquake ( M w 8.3) at two OBPG stations near the Kuril Trench. The tsunami source model we estimated is consistent with models previously derived using various other geophysical data sets. Furthermore, the coastal tsunami waveforms we modeled match the observed tsunami well. Forecasts of tsunami arrival times and first peak amplitudes by our method can be obtained 20 min after an earthquake, and can be provided to the coastal communities nearest to the source with a lead time of ∼10 min.