The 15 August 2007 Peru Earthquake and Tsunami: Influence of the Source Characteristics on the Tsunami Heights

0 Laboratoire de Geophysique , CEA, BP 640, 98713 Papeete, Tahiti, French Polynesia 1 CEA, DAM, DIF, F-91297 Arpajon, France The tsunami caused by the 2007 Peru earthquake (Mw 8.0) provoked less damage than by the seismic shaking itself (numerous casualties due to the earthquake in the vicinity of Pisco). However, it propagated across the Pacific Ocean and small waves were observed on one tide gauge in Taiohae Bay (Nuku Hiva, Marquesas, French Polynesia). We invert seismological data to recover the rupture pattern in two steps. The first step uses surface waves to find a solution for the moment tensor, and the second step uses body waves to compute the slip distribution in the source area. We find the slip distribution to consist of two main slip patches in the source area. The inversion of surface waves yields a scalar moment of 8.9 1020 Nm, and bodywave inversion gives 1.4 1021 Nm. The inversion of tsunami data recorded on a single deep ocean sensor also can be used to compute a fault slip pattern (yielding a scalar moment of 1.1 1021 Nm). We then use these different sources to model the tsunami propagation across the Pacific Ocean, especially towards Nuku Hiva. While the source model taken from the body-wave inversion yields computed tsunami waves systematically too low with respect to observations (on the central Pacific Ocean DART buoy as on the Polynesian tide gauge), the source model established from the surface-wave inversion is more efficient to fit the observations, confirming that the tsunami is sensitive to the low frequency component of the source. Finally we also discuss the modeling of the late tsunami arrivals in Taiohae Bay using several friction coefficients for the sea bottom. 1. Introduction and Objectives On August 15, 2007, at 23:40 UTC, a Mw 8.0 earthquake occurred off the Peruvian coast on a shallow portion of the subduction zone where the Nazca plate subducts beneath the South American plate at a rapid convergence rate of about 6.5 to 7.0 cm/yr (KENDRICK et al., 2003). In addition to numerous casualties and severe damage in the region of Pisco, the sea-floor displacement associated to the quake triggered a destructive tsunami, as normal given the earthquake magnitude and such a shallow hypocentral depth. The runup reported along the coastline measured up to almost 10 m locally, and on average 3 to 5 m along at least 100 km of the Peruvian coast (SLADEN et al., 2007c; FRITZ et al., 2008; http://www.eeri.org/lfe/pdf/peru_coast_tsunami.pdf). This event occurred in an area frequently struck by major thrusting events, immediately north to the 12 November, 1996 Mw 7.7 earthquake. Further southward, the 1868 historical seismic gap has been partly ruptured by the Mw 8.4 23 June 2001 earthquake, however the identified gap may correspond to a magnitude 9 event as in 1868 (DORBATH et al., 1990). This seismic gap, together with the nearby gap corresponding to the North Chile 1877 earthquake (DELOUIS et al., 1997; CHLIEH et al., 2004), constitute large areas expected to produce major earthquakes (COMTE and PARDO, 1991). Such being the case, the 2001 Peru earthquake seems to have ruptured a significant part of the 1868 gap, while the 1877 remains unbroken (Fig. 1). While it is well known which seismological parameters are relevant to issue a warning (location, seismic moment, and focal depth), the difficulty in determining the accurate level of impact and possible flooding for specific bays and harbors is real, particularly because the actual azimuth and rupture direction are difficult to define in the first minutes after the quake, and yet they define the area where the maximum impact is expected. Recent events in 2006 have also stressed the need for cautious appraisal of tsunami impact in remote sites, as was the case for Crescent City during the November 15, 2006 tsunami initiated in the Kuril Islands. The Marquesas Islands (French Polynesia) are among these remote far-field sites where the impact can be strong, but highly dependent upon the fault azimuth and rupture direction (HE BERT et al., 2001a, b; OKAL and HE BERT, 2007). The dependence on relation with the seismic moment is the first-order aspect to take into account (TALANDIER and OKAL, 1989), however corrections including azimuthal influence and propagation distance are now considered (REYMOND et al., 2007) to help define more accurately the expected level of tsunami impact. In addition, the slip heterogeneities during the rupture may also influence the impact in the far field, as was already observed in the Indian Ocean (HE BERT et al., 2007). The 15 August 2007 event presents the opportunity to discuss the influence of the seismic rupture characteristics on the tsunami heights, especially in the far field. Seismic inversions have been performed to rapidly compute the moment and to obtain the slip distribution. Another insight to the earthquake source is also given by a simple inversion of the tsunami waves recorded on a single DART signal located off Chile. Then we compare tsunami heights computed in the far field (French Polynesia) using the various source models considered, and we emphasize the influence of the source characteristics on these water heights. Our results finally allow us to discuss the relevant parameters of the source that facilitate improvement of the tsunami hazard assessment, and that could be considered in the frame of real-time modeling. 2. Seismological Overview of the 15 August 2007 Earthquake Given the magnitude and depth quickly determined on August 15, 2007, a tsunami information bulletin was first issued by the Pacific Tsunami Warning Center at 23:54 UTC, while the estimated magnitude was then only Mw 7.5. Subsequently an expanding regional warning (Mw 7.9) was issued at 00:19 UTC, based on the variable-period mantle magnitude Mm (OKAL and TALANDIER, 1989; REYMOND et al., 1991; SCHINDELE et al., 1995) (further information available on http://www.prh.noaa.gov/ptwc/messages/pacific/2007/). Meanwhile in Tahiti (French Polynesia) the TREMORS system, similarly based on Mm, provided at 00:35 UTC a rapid estimation of the seismic moment corresponding to Mw 8.1. The Polynesian Tsunami Warning System (CPPT: Centre Polynesien de Prevention des Tsunamis) then issued a yellow warning bulletin for the Marquesas Islands at 00:57 UTC. While the PTWC finally cancelled the warning at 02:10 UTC, the warning in Polynesia was revised to the orange level at 03:50 UTC for the Marquesas, since these islands are well known to amplify tsunami waves coming from Chilean and Peruvian subduction earthquakes. Finally the alert was cancelled at 11:30 UTC, two hours after the arrival of the waves. During the watch level, only coastal sites and light boats have been evacuated. No damage was observed. 2.1 PDFM Inversion A method of quick inversion of seismic moment tensor using very small data sets (named PDFM: Preliminary Determination of Focal Mechanism) started in 1997 with the goa (...truncated)