We present a detailed timeline of the warning procedures as they unfolded at the Laboratoire de Géophysique in Papeete, Tahiti, during the nights of 26–27 February 2010 (Maule, Chile tsunami) and 10–11 March 2011 (Tohoku tsunami). In particular, we discuss how the flow of information available to the warning center (including seismic evaluations obtained both locally and from other warning centers, as well as maregraph and DART buoy data) built up and eventually led to red alerts, which the local authorities used in both cases to impose an evacuation of low-lying areas on 68 islands. While the alerts were successful in Polynesia, a difficulty arose in 2011 when the alert had to be reinstated immediately as the all clear was being declared, since the maximum amplitude was carried by the fourth wave packet. We also present a complete dataset of 119 values of run-up and inundation surveyed in the aftermath of the two tsunamis, principally in the Marquesas Islands where their effects were maximal, and on Tahiti and Moorea for the 2011 event. The highest run-up (4.45 m) was observed in 2011 in the Bay of Taipivai on Nuku Hiva, where seven houses were flooded. We find no clear correlation between run-up values at the same locations in 2010 and 2011, suggesting that local responses are controlled by details specific to each tsunami. In 2010, in the village of Puamau on Hiva Oa (Marquesas), a delayed harbor response, probably due to resonance of the bay upon arrival of short-period components dispersed outside the shallow-water approximation, flung a launch onto a wharf, 7 h after the first arrivals, and 2.5 h after issuance of the all clear. |
