Millimeter‐Scale Tsunami Detected by a Wide and Dense Observation Array in the Deep Ocean: Fault Modeling of an Mw 6.0 Interplate Earthquake off Sanriku, NE Japan.

A new dense and widely distributed tsunami observation network installed off northeast Japan detected millimeter‐scale tsunamis from an Mw 6.0 shallow interplate earthquake on 20 August 2016. Based on the fault model deduced from this data set, we obtained a stress drop of 1.5 MPa for this event, similar to those associated with typical interplate earthquakes. The rupture area was unlikely to overlap with regions where slow earthquakes occur, such as low‐frequency‐tremors and very‐low‐frequency‐earthquakes. The results demonstrated that this new network has dramatically increased the detectability of millimeter‐scale tsunamis. Some near‐source stations were contaminated by large pressure offset signals irrelevant to tsunami, and we must therefore be careful when analyzing these data. Nonetheless, the new array enables estimations of the stress drops of moderate offshore earthquakes and can be used to elucidate the spatial variation of mechanical properties along the plate interface with much higher resolution than previously possible. Plain Language Summary: Tsunamis are generated when an earthquake occurs beneath the seafloor. Far fewer tsunami observations have been recorded from moderate earthquakes than large to giant earthquakes because tsunamis created by moderate earthquakes have been too small to be observed. On 20 August 2016, a moderate earthquake occurred off Sanriku, in northeastern Japan, and a tsunami with a height of less than 1 cm was recorded by a new seafloor tsunami observation network. This network has many tsunami sensors distributed much closer to each other and over a much wider area than any other previous network in the world. Using these data, this study estimated the source location and size, and the slip amount of the 2016 earthquake with higher accuracy, which was impossible to achieve from past observations because they were too far away from the earthquake and the signals were too small. Using this source information, we could estimate the stress drop associated with the earthquake, which is important because the stress drop information deepens our understanding of how and why earthquakes happen. Key Points: Millimeter‐scale tsunamis from an Mw 6.0 earthquake were captured by the S‐net, a new nationwide pressure gauge array off Sanriku, JapanTsunami signals were identified from the pressure data adjacent to the source, which were contaminated by signals irrelevant to tsunamisWe inferred the stress drop of the earthquake from the tsunami data more reliably than could be done from seismogram analysis [ABSTRACT FROM AUTHOR]