The U.S. East Coast Meteotsunami of June 13, 2013

NOAA's two Tsunami Warning Centers (TWCs) provide advance notification to coastal communities concerning tsunami hazards. While the focus is primarily on seismic sources, the U.S. East Coast event of June 13, 2013 indicates the importance of understanding and forecasting atmospherically-driven tsunamis, or meteotsunamis, as well. Here we describe an approach which explains the generation of this event by atmospheric processes, and suggests that the causative forces can be monitored and used to forecast meteotsunami occurrence. The U.S. East Coast tsunami of June 13, 2013 was well recorded at tide gauges from North Carolina to Massachusetts as well as at Bermuda and Puerto Rico. It also triggered DART 44402, just east of the Atlantic shelf break at 39.4N. As there was no seismic energy release associated with the tsunami and an eastward propagating major weather system crossed the Atlantic coast just before the tsunami, the focus turned to atmospheric forcing. Tsunami forecast models used at the two U.S. TWCs were modified to introduce moving atmospheric pressure distributions as sources. In a simple case, a north-south oriented line air pressure jump of width 50 km and pressure of 4 mb at sea level was moved eastward at 20 m/s. The speed matched both the storm speed at the coast and the long wave speed for 40 m deep water, thus allowing for resonant coupling of atmosphere to ocean in the shelf region (Proudman Resonance). Considering the simplicity of the source, a reasonable comparison between the modeled and observed tsunami was obtained with regards to arrival time and height. The proposed source also offers an explanation of the later wave arrivals at US tide gauges. These typically lagged the arrival at Bermuda - a location much further east. This pattern can be explained within the context of Proudman resonance if the waves arriving at coastal stations originated at the shelf break as reflected waves. Model animations of wave dynamics corroborate this phenomenon. The contribution of edge waves generated as the system moves over the coast is also examined. Remaining questions include the importance of shelf parameters in setting the wave fetch and the 'Q' of Proudman resonance along the Atlantic coastline. In other words, are some stretches of shelf more conducive to tsunami formation than others? Wind stress was disregarded in the initial modeling work leaving its possible importance as another unanswered question. Operational questions include how to detect likely meteotsunami conditions with real-time meteorological measurements, and what form alerts should take. The minimum necessary temporal resolution of the pressure sensors along with their density and siting needs to be determined. Because details of the source, such as direction and speed of propagation, will likely subject unique sections of coastline to tsunami attack, the detailed analysis of data from sensor arrays to be used in forecasting will be important.