Solving the inverse problem, the tsunami source can be estimated from the measured run-up data. The estimated source condition is analyzed to determine whether or not the earthquake mechanism inferred from fault dislocation models is consistent. This type of analysis for the 1992 Nicaragua tsunami led Kanamori and Kikuchi (1993) to propose the mechanism of “slow-slip tsunami earthquakes”—deceptively mild quakes that generate anomalously large tsunamis. Understanding such a phenomenon is critical for adequate tsunami risk assessment. The measured run-up data can also be used as a benchmark to validate the hydrodynamic models. For example, the measured run-up data for the 1993 Okushiri tsunami were used for the model validation exercise at the community workshop (Yeh et al., 1996). This benchmark problem is adopted in the recent model validation guideline by the NTHMP (OAR PMEL-135, Synolakis et al., 2007).
Tsunami surveys in the past have revealed many tsunami characteristics. For example, locally high anomalous run-up resulting from the 1992 Flores and the 1998 Papua New Guinea tsunamis indicated the possible occurrence of earthquake-induced submarine landslides (Yeh et al., 1993; Synolakis et al., 2002). The field survey in Babi Island—a small cone-shaped island where 263 people were killed in the normally safe lee side of the island by the 1992 Flores tsunami—led the subsequent numerical simulations (Liu et al., 1995) and large-scale laboratory experiments (Briggs et al., 1995). The comprehensive study revealed the unexpected tsunami behavior (Yeh et al., 1994). When it hit the island, the tsunami split in two. The split tsunami wrapped around the island and joined to create a new, larger wave that crashed into the lee side of the island. This phenomenon that is unique to tsunami is also adopted as one of the benchmark problems in OAR PMEL-135.
Tsunami surveys are also needed for other important observations: flow effects on manmade structures and natural geomorphologic features, social impacts, and identifications of all salient features for the use of future tsunami loss reduction. Tsunami field surveys also provide us with evidence that tsunamis are capable of transporting sediments, rocks, and boulders (Bourgeois et al., 1999). Such information and data are important not only for future prevention of scouring and structure damages, but also for the assessment of geological evidence of prehistoric tsunami events.
Systematic and organized field surveys specifically aimed at the social impacts were initiated for the first time in response to the 2004 Indian Ocean tsunami event (Suter et al., 2009). Appropriate social science post-event research audits hold the potential to document important lessons to be learned. Such social science research of this type would cover a range of topics; it would be conducted at different points in time after an event; and it would be performed by researchers with varied and specific training, expertise, and experience. The range of topics benefiting from this post-event investigation includes but not limited to how well the warning system functioned as a system across the varied players involved in the system, e.g. the TWCs, state and local government, and the public; the adequacy of TWC and state and local government messages to each other and the public in terms of how those messages influenced protective action-taking; and much more. An adequate social science research agenda would include both quick-response reconnaissance research to capture perishable data and longer-term research conducted months or longer after an event.