Kévin Juhel , Quentin Bletery , Andrea Licciardi, Martin Vallée, Céline Hourcade, Théodore Michel

Prompt ElastoGravity Signals (PEGS) are light-speed gravity-induced signals recorded by seismometers before the arrival of seismic waves. They have raised interest for early warning applications but their weak amplitudes, close to the background seismic noise even for large earthquakes, have questioned PEGS actual potential for operational use. A deep-learning model has recently demonstrated its ability to mitigate this noise limitation and to provide in near real-time the earthquake moment magnitude (Mw). However, this approach has proven to be efficient only for very large earthquakes (Mw > 8.3) of known focal mechanism. Here we show unprecedented performance in full earthquake characterization from PEGS using the dense broadband seismic network deployed in Alaska and Western Canada. Our deep-learning model is able to provide accurate magnitude and focal mechanism estimates of Mw > 7.8 earthquakes, 2 minutes after origin time (hence the tsunamigenic potential). For very large earthquakes whose rupture is still ongoing after 2 minutes, the model tracks the instantaneous magnitude from that time until the rupture completion. Our results represent a major step towards the routine use of PEGS in operational warning systems, and demonstrate its potential for tsunami warning in the Alaska region, and other densely-instrumented areas.