S43C-01
Detecting Tsunami Genesis and Scales Directly from Coastal GPS Stations
Different from the conventional approach to tsunami warnings that rely on earthquake magnitude estimates, we have found that coastal GPS stations are able to detect continental slope displacements of faulting due to big earthquakes, and that the detected seafloor displacements are able to determine tsunami source energy and scales instantaneously. This method has successfully replicated three historical tsunamis caused by the 2004 Sumatra earthquake, the 2005 Nias earthquake, and the 1964 Alaska earthquake, respectively, and has been compared favorably with the conventional seismic solutions that usually take hours or days to get through inverting seismographs. Because many coastal GPS stations are already in operation for measuring ground motions in real time as often as once every few seconds, this study suggests a practical way of identifying tsunamigenic earthquakes for early warnings and reducing false alarms.
S43C-02
Tsunami Observations on Hydrophones and Island Seismic Stations
The tsunami generated by the great Indonesian earthquake of 26 December 2004 was recorded across a myriad of technologies, many of which had not been designed, nor expected, to record tsunami signals. We reported on the tsunami signals from this event observed at hydrophones, intended for nuclear test monitoring, and broadband seismometers that are part of the global seismic network (GSN). Our observations led us to examine more recently reported tsunamis and other historic tsunamis. The great Sumatra-Andaman earthquake (Mw 9.1) produced high-frequency (greater than 5 mHz) dispersed tsunami signals, in addition to the destructive wave, recorded by hydrophone stations offshore from Diego Garcia and Cape Leeuwin, Australia, and by many seismic stations in the Indian Ocean and on the coast of Antarctica. Dispersed energy was observed to 60 mHz. The details within the dispersed signal provided source information to which tide gauge data are insensitive. The source of high-frequency signals could be determined using event- to-station distances estimated from the dispersion. Fine structure in the tsunami signal indicated a possible secondary high-frequency source. The dispersion observations and modeling also identified individual reflector sources over basin-wide distances. Two other recent tsunamis were observed in the Indian Ocean. The 28 March 2005 earthquake (Mw 8.6) produced high-frequency tsunami waves (to 20 mHz) observed at the Diego Garcia hydrophone station and the AIS seismic station. In addition, the lower frequency, non-dispersed tsunami waves were seen at four other seismic stations. The Mw 7.7 earthquake on 17 July 2006 south of Java also generated high frequency tsunami waves (to 10 mHz). Clear, dispersed signals were observed on hydrophone stations and seismic stations at the Cocos-Keeling Islands, and Casey, Antarctica. The first arriving energy is consistent with a source located at the earthquake epicenter. However, the strongest signals at Cocos-Keeling and Diego Garcia islands arrive after the initial onset, indicating a secondary source area or strong reflections. Island seismic station data from the Pacific were examined for tsunami signals for 14 earthquakes in the circum- Pacific region between 1994 and 2003 with Mw 7.1 to 8.4 that were known to be tsunamigenic. The seismic signals were compared to tide gauge recordings when available. Tsunami signals from 0.5 mHz up to 2-8 mHz were observed at the expected arrival time on the horizontal components of low-elevation seismic stations near coastlines for eight of the 14 earthquakes. Lower amplitude signals were observed for the other six. Dispersed tsunami signals between 1 and 8 mHz were observed for two earthquakes (Mw 8.2-8.4) and were consistent with predictions. The observations suggest that seismic data could be used to complement tide gauges and ocean bottom pressure recorders for tsunami monitoring and research.
S43C-03 INVITED
The Potential Role of Satellites in Detecting and Tracking Tsunamis
Until the morning of December 26 2004 it was not entirely clear if a radar altimeter in polar-orbit around the Earth could detect a low amplitude tsunami wave travelling at high speed across the ocean surface. The answer came when by chance the satellite Jason flew over the Indian Ocean on that fateful morning and recorded a double- peaked wave with peak-to-trough amplitude of 80 cm. A constellation of fit-for-purpose microsatellites, equipped with radar altimeters, had already been proposed as a means of delivering continuous global observations on sea state to ships, platforms and forecasting centres. It will be demonstrated through a computer animation how such a system, working in conjunction with the global network of ground stations for detecting undersea earthquakes, could be instantly switched to pick up and track the progress of a tsunami. Furthermore, all studies carried out to date on the role of ocean currents and eddies in determining the global transport of heat, and its effect on future climate, have concluded that finer spatial resolution is required than that achieved today. It will be shown how this requirement could also be met by a constellation of altimeters. Various schemes for operating such a system will be discussed.
S43C-04
UHF radar signature of a tsunami approaching coastal areas: modeling, experiments and application to tsunami warning
Perhaps for the first time, satellite altimeters provided transects of the 12/26/04 tsunami elevation across the Indian Ocean, while the event unfolded. Here we use well-established Ultra High Frequency (UHF) radar technology, to develop a method that could provide warning of an incoming tsunami to coastal populations. When a tsunami reaches the continental shelf, the mostly depth-uniform current it induces greatly increases in speed (maybe up to 10-20 cm/s) and may induce significant Doppler shifts in ocean surface waves, particularly for those of smaller wavelength (high frequency). Given proper processing, such shifts could be identified by shore-based UHF radars as a tsunami signature and trigger a warning. Due to the tsunami slowing down with decreasing water depth, from the shelf break to shore, warning times of 5-15 minutes could be conceivable, depending on the shelf width, which could be sufficient to proceed with vertical evacuation in exposed areas. Here, we use a Higher Order Spectral (HOS) Method to model fully nonlinear sea states caused by wind, down to typical UHF wavelength of order 10 cm, as well as Doppler shifts and wave shoaling/refraction caused by slowly varying depth uniform currents. Such currents can be obtained for selected case studies, e.g., from tsunami propagation modeling, using a standard long wave model. UHF radar backscattering is modeled by a Boundary Element Method, solving Maxwell's equations, developed and validated in earlier work. We present initial results of this modeling study, in terms of spatio-temporal UHF radar signatures and their sentsitivity to governing physical parameters. We also present large-scale laboratory results for radar backscattering spectra measured during wave-current interaction experiments performed in the FIRST basin (La Seyne, France). We intend to apply our detection methodology to Southern Thailand, based on earlier modeling work we performed for the 12/26/04 tsunami propagation.
S43C-05
Can Ionospheric Sounding Help Tsunami Warning Systems ?
A series of ionospheric anomalies following the Sumatra tsunami has been recently reported in the scientific literature (e.g., Liu et al. 2006; DasGupta et al. 2006; Occhipinti et al. 2006). Similar anomalies were also observed after the tsunamigenic earthquake in Peru in 2001 (Artru et al., 2005). All these anomalies show the signature in the ionosphere of tsunami-generated internal gravity waves (IGW) propagating in the neutral atmosphere over oceanic regions. The strong amplification mechanism of atmospheric IGW allows to detect these anomalies when the tsunami is offshore where the see level displacement is still small. In addition, the dense coverage of ionospheric sounding instruments over the oceans increases over time and more instruments will be able to provide ionospheric measurements: i.d., Doppler sounding, over-the-horizon radar (OTH) and space-based GPS data (e.g., COSMIC). Most of the ionospheric anomalies are also deterministic and reproducible by numerical modeling (Occhipinti et al., 2006), this latter will supply an useful help in the estimation of expected anomalies. The sensitivity of altimeters, OTH radar, ground-based and space-based GPS measurements is analyzed in this work by the way of the modeling. The results are used to discuss the role of ionospheric sounding in the future oceanic monitoring and tsunami warning system. [Artru et al., 2005] Geophys. J. Int., 160, 2005 [DasGupta et al., 2006] Earth Planet. Space, 35, 929-959. [Liu et al., 2006] J. Geophys. Res., 111, A05303. [Occhipinti et al., 2006] Geophys. Res. Lett., 33, L20104, 2006
S43C-06 INVITED
Speeding up seismic tsunami warning using W phase
W phase is a long period phase arriving before S wave. It can be interpreted as superposition of the fundamental, 1st, 2nd, and 3rd overtones of Rayleigh waves and has a group velocity of 8 km/s at 1000 s and 8.6 km/s at 100 sec. At a distance of 50° the W-phase energy is mainly contained within a time window of 700 s after the P arrival. Since the amplitude of long period waves better represents the tsunami potential of an earthquake, the use of W phase has a merit in assessing the tsunami potential at an earliest possible time. To use W phase we need to solve two problems. First, many STS-1 records at short distances are clipped at or before the surface wave arrival. Thus, with the traditional frequency domain deconvolution, the wraparound effect from the end of the record makes the beginning part unusable, even if W phase is on scale before the surface wave arrival. Second, no systematic analysis has been made to use W phase for source studies. We solved the first problem by using a time-domain recursive deconvolution method described by Zhu with some modification. We investigated the second problem by inverting the displacement seismograms windowed over a short duration after the P arrival. The duration is given by 15*Δ (in degree) s. Thus, at a distance of 50°, we use the record only up to 23 min after the origin time which is the distinct advantage of using W phase for tsunami warning purposes. If many stations are available at shorter distances, the time can be considerably shortened. The bandwidth of W phase is approximately from 0.001 to 0.01 Hz, and we band-pass filter the data from 0.0005 to 0.005 Hz in most cases. Having extracted W phase from the vertical component records, we concatenate them in time in the order of distance to obtain a single W-phase time series. This time series constitutes the data column vector. We synthesize similar concatenated W-phase time series by mode summation for each of the moment tensor elements. These time series constitute the column vectors of the matrix which maps the unknown moment tensor to the data. We performed a linear inversion using a point source for several large earthquakes including the 2004 Sumatra-Andaman earthquake, the 2005 Nias earthquake, the 2006 Kuril Is. earthquake, and the 2007 Peruvian earthquake. The results are satisfactory, and give promise of the use of W phase for rapid assessment of tsunami potential. A practical procedure for operational use of W phase for tsunami warning purposes will be discussed.
S43C-07
A Rapid Energy-Duration Discriminant for Tsunami Earthquakes
"Tsunami earthquakes" are a class of rare, but devastating slow earthquakes that create tsunami waves much larger than expected from their initial magnitudes. Because these events rupture more slowly than standard "strong" earthquakes, their resultant shaking is much weaker, while they rupture much longer than similarly sized events. By utilizing global seismic data from recent large earthquakes recorded at teleseismic distances (30-80°), we determined the cumulative time history of earthquake energy release and unique rupture durations, TR. Using data for all earthquakes with MW >7.2, we identified a unique and robust discriminant for "tsunami earthquakes" based on their long duration and weak shaking. A simple logarithmic discriminant of the form: log E=a+b × log\ TD clearly separates normal strong shaking earthquakes from slow tsunami earthquakes. For broad-band energy estimates, EBB, b~2, and a=25 to 26.8, while for short-period energy (between 0.5 and 2 Hz), EHF, b~1.6, and a=25 to 27.5. In both cases, a nearly 102 difference exists between the weakest normal earthquake and the strongest-shaking tsunami earthquake. While broad-band energy results are more physically meaningful, the high-frequency results are preferred because they do not suffer significantly from later phases, yield a clearer discriminant, and can be performed with more widely available short-period instruments. The methodology described here clearly identified the 1992 MW 7.7 Nicaragua, the 1994 MW 7.8 and the 2006 MW 7.8 Java earthquakes as slow tsunami earthquakes, by their markedly small energy for duration. Because the method uses a real-time estimated-correction for focal solutions (described in Newman and Okal, [1998]), it can be applied in real-time allowing for a rapid tsunami earthquake assessment within ~10 minutes of an earthquake rupture, depending on size and station availability.