Effectiveness of Real-Time Near-Field Tsunami Inundation Forecasts for Tsunami Evacuation in Kushiro City, Hokkaido, Japan - Post-Tsunami Hazard: Reconstruction and Restoration - page 161

11.3.2 Earthquake Source Parameters of the JMA Earthquake

Early Warning (EEW) System

The JMA uses a real-time seismic data processing system to determine an
earthquake's hypocenter and magnitude. The calculated hypocenter parameters of
longitude, latitude, depth, and magnitude are used as input for the JMA tsunami
forecasting system. Although earthquake source parameters can be obtained by the
EEW system within a short time period (3 min), the system might underestimate the
magnitude of great earthquakes, Mw more than 8.5, as was revealed during the 2011
Tohoku earthquake event. The JMA's initial magnitude estimate for the earthquake
was M 7.9, and the fi nal estimate from the system was M 8.1 (Ohta et al. 2012 ),
which are much smaller than the Mw 9.0 estimated by many studies (e.g., Ide et al.
2011 ; Yoshida et al. 2011 ; Ozawa et al. 2011 ; Fujii et al. 2011 ; Gusman et al. 2012 ;
Satake et al. 2013 ). Therefore, during an earthquake event, we prefer to wait for more
data to become available to obtain more reliable earthquake source parameters.

11.3.3 Earthquake Source Parameters from W Phase Inversion

Previous studies show that W phase data can give a reliable earthquake magnitude
estimate (Kanamori and Rivera 2008 ; Duputel et al. 2011 ; Gusman and Tanioka
2013 ; Benavente and Cummins 2013 ). Our previous study showed that reliable cen-
troid moment tensor solutions of the 2011 Tohoku earthquake can be estimated by
use of 5 or 10 min of W phase data recorded at Japanese F-net stations (Gusman and
Tanioka 2013 ; Gusman et al. 2014 ). The centroid moment tensor solution from the
W phase inversion, with a scaling relation of moment magnitude and rupture area,
can be used to estimate a simple fault model. Benavente and Cummins ( 2013 )
explored the ability of W phase waveform inversions to recover fi rst-order coseis-
mic slip distribution for great earthquakes.

11.3.4 Earthquake Fault Model from GPS

Time Series Inversion

To estimate a fault model in real-time from Real Time Kinematic (RTK) GPS time
series data, Ohta et al. ( 2012 ) developed the Real-time Automatic detection method
for Permanent Displacement (RAPiD) algorithm. The coseismic displacement
fi elds caused by the 2011 Tohoku earthquake were estimated by the RAPiD algo-
rithm using 1-Hz GPS time series data recorded at 527 GEONET stations. The fi nal
fault model (Mw 8.7) and coseismic displacement fi eld due to the earthquake was
estimated within 4 min and 35 s after the earthquake occurred (Ohta et al. 2012 ).