Long-Term Ocean Observing Coupled with Community Engagement Improves Tsunami Early Warning

FIGURE 1. Schematic of ocean-based geophysical instrumentation and data communications installation. A wave glider and a Deep-Ocean Assessment and Reporting of Tsunamis (DART) tsunameter communicate with a satellite. An autonomous underwater vehicle (AUV) collects data from the water column for later transmission via the satellite. Other instrumentation includes a recoverable geodetic transponder, a trawl-resistant and current-protected seismometer, and a self-calibrating pressure gauge. The 2004 magnitude (M) 9.1 Sumatra-Andaman Islands earthquake in the Indian Ocean triggered the deadliest tsunami ever, killing more than 230,000 people. In response, the United Nations Educational, Scientific, and Cultural Organization (UNESCO) established three additional Intergovernmental Coordination Groups (ICGs) for the Tsunami and Other Coastal Hazards Early Warning System: for the Caribbean and Adjacent Regions (ICG/CARIBE-EWS), for the Indian Ocean, and for the Northeastern Atlantic, Mediterranean, and Connected Seas. Along with the ICG for the Pacific Ocean, which was established in 1965, one of the goals of the new ICGs was to improve earthquake and tsunami monitoring and early warning. This need was further demonstrated by the 2011 Great East Japan (Tōhoku-oki) earthquake and tsunami, which killed more than 20,000 people, and other destructive tsunamis that occurred in the Solomon Islands, Samoa, Tonga, Chile, Indonesia, and Peru. In response to the call to action by the UN Decade of Ocean Science for Sustainable Development (2021– 2030), as well as the desired safe ocean outcome (von Hillebrandt-Andrade et al., 2021), the Intergovernmental Oceanographic Commission (IOC) of UNESCO approved the Ocean Decade Tsunami Programme in June 2021. One of its goals is to develop the capability to issue actionable alerts for tsunamis from all sources with minimum uncertainty within 10 minutes (Angove et al., 2019). While laudable, this goal presents complexities. Currently, warning depends on quick detection as well as the location and initial magnitude estimates of an earthquake that may generate a tsunami. Other factors that affect tsunamis, such as the faulting mechanism (how the faults slide past each other) and areal extent of the earthquake, currently take at least 20–30 minutes to forecast and are still subject to large uncertainties. Hence, agencies charged with tsunami early warning need to broadcast public alerts within minutes after an earthquake occurs but may struggle to meet this 10-minute goal without further technological advances, some of which are outlined in this article. To reduce loss of life through adequate tsunami warning requires global ocean-based seismic, sea level, and geodetic initiatives to detect high-impact earthquakes and tsunamis, combined with sufficient communication and education so that people know how to respond when they receive alerts and warnings. The United Nations

of its goals is to develop the capability to issue actionable alerts for tsunamis from all sources with minimum uncertainty within 10 minutes (Angove et al., 2019). While laudable, this goal presents complexities. Currently, warning depends on quick detection as well as the location and initial magnitude estimates of an earthquake that may generate a tsunami. Other factors that affect tsunamis, such as the faulting mechanism (how the faults slide past each other) and areal extent of the earthquake, currently take at least 20-30 minutes to forecast and are still subject to large uncertainties. Hence, agencies charged with tsunami early warning need to broadcast public alerts within minutes after an earthquake occurs but may struggle to meet this 10-minute goal without further technological advances, some of which are outlined in this article.
To reduce loss of life through adequate tsunami warning requires global ocean-based seismic, sea level, and geodetic initiatives to detect high-impact earthquakes and tsunamis, combined with sufficient communication and education so that people know how to respond when they receive alerts and warnings. The United Nations International Strategy for Disaster Reduction defines an early warning system as "a set of capacities needed to generate and disseminate timely and meaningful warning information to enable individuals, communities, and organizations threatened by a hazard to prepare and to act appropriately and in sufficient time to reduce the possibility of harm or loss" (UNISDR, 2012). In short, a successful early warning system requires technology coupled with human factors (Kelman and Glantz, 2014).
In this article, we explore case studies from Japan and Canada, where scientists are leading the way in incorporating ocean observing capabilities in their early warning systems. We also explore advancements and challenges in the Caribbean, an area with a complex tectonic environment that would benefit greatly from increased global ocean observing capabilities. We also explore physical and social science interventions necessary to reduce loss of life.

TECHNOLOGICAL CAPABILITIES
Seafloor seismometers measure Earth motions in three dimensions across an extensive frequency band, from tides, earthquake-caused resonances, and seismic waves to sounds created by whales and ships (e.g., Kohler et al., 2020;Kuna and Nábelek, 2021; Figure 1). Seafloor bottom pressure recorders enable detection of a tsunami wave and its speed, direction, and wavelength, providing information to help forecast coastal tsunami height and duration (e.g., Rabinovich and Eblé, 2015). For instance, Deep-ocean Assessment and Reporting of Tsunamis autonomous underwater vehicles, and buoys closer to or at the sea surface have helped reduce delays in delivery of data to TWCs (Figure 1). Wave gliders, which look like autonomous surfboards with collapsible propellers, use wave and solar energy for electrical power and propulsion.
Within minutes, onboard communications systems access orbiting satellites in order to send data from seafloor sensors to onshore collection points.
In the past two decades, Japan, Canada, and the United States have also installed seismic and bottom pressure recorders onto regional ocean bottom fiber-optic cable arrays located in the Pacific Ocean. For example, in waters off southwestern Japan in the Nankai Trough, the Japan Agency for Marine-Earth Science and Technology (JAMSTEC) constructed DONET, the Dense Oceanfloor Network system for Earthquakes and Tsunamis (Aoi et al., 2020).
DONET connects various sensors to a node (a junction that connects sensors to a submarine cable) to provide data for evaluating the coupling and slip behavior along the Nankai Trough, a fault area presumed to be primed for a future earthquake and potential tsunami (Figures 2a and 3a The probability of an M8 earthquake occurring in this region in the next 30 years is estimated to be more than 80% (Geological Survey of Japan, 2021).  nificantly reduced due to damage to onshore networks and regional seismic and sea level networks by Hurricanes Maria and Irma, and then Hurricane Iota in 2020 ( Figure 5).
Other regional and international partners lost numer-  (2004).

TSUNAMI EARLY WARNING
Collecting and distributing near-real-time data from longterm seafloor deployments will increase the accuracy of earthquake parameters, such as a quake's location, the  implementing Tsunami Ready (https://www.tsunamiready.org), the IOC-UNESCO international performance-based community recognition program on tsunami preparedness; and organizing the annual regional Tsunami Exercise (CARIBE WAVE; https://www.tsunamizone.org/caribewave/).  Howe et al., 2019). The increased number of globally distributed stations will provide valuable earthquake data that will lead to more reliable and accurate tsunami warning systems, as well as provide information that will greatly enhance details of three-dimensional global structure from Earth's outer surface all the way to its inner core.

ACKNOWLEDGMENTS
Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US government. We thank Stephanie Ross, Jeff McGuire, Shane Detweiler, Mike Diggles, and two anonymous reviewers for their contributions and insights to this work. We thank Jennifer Matthews (UCSD) for her work on