Combining measurements and models for real-time tsunami forecast

Since the 2004 Indian Ocean tsunami, the most destructive tsunami in recorded history, worldwide awareness of tsunami hazard has peaked and global expansion of tsunami forecasting tools has made dramatic progress in both instruments and technology. Two of the most seminal advances in tsunami forecast since the Indian Ocean tsunami are: 1) the deployment of an extensive network of sensors to acquire deep ocean tsunami measurements, and 2) the introduction of real time numerical simulation as an vital tool in tsunami forecasting (Bernard and Robinson, 2009). To date, the number of deep-ocean tsunameters has grown from 9 in 2004 to 48 (Figure 1), forming a global tsunami monitoring network in the Pacific, the Indian Ocean and the Atlantic that is currently co-managed by multi-nations: Australia, Chile, Indonesia, Thailand and the United States. Since 2005, these tsunameters have detected 15 tsunamis generated by major earthquakes in the Pacific and Indian Ocean, building growing confidence of accuracy and reliability in far-field tsunami forecast (Titov, 2009). Japan has also installed six cabled Ocean Bottom Tsunami Meters (OBTMs), using the same sensor as the buoy-based tsunameters, to provide timely near-field tsunami forecasting for Japanese coasts since 1980s (Tsushima et al., 2009). The global implementation of these deep-ocean tsunami detectors has significantly accelerated the development and implementation of more accurate tsunami forecasting systems. Previous systems, relying on seismometers or coastal tide gages, had resulted 15 of 20 tsunami false alarms since 2009) respectively, are monitoring worldwide tsunami activity for rapid forecast to minimize tsunami impact for at risk coastal communities. While the Japan and Australia