We mapped temporal seismic velocity variations during the Kumamoto earthquake due to fault rupture and a volcanic eruption.

Many significant geological features in central Japan reflect the history of tectonic events such as volcanic eruptions, fault ruptures, and the collision of the Izu-Bonin arc. Because most previous studies focused on deep, large-scale structures at relatively low resolution or on local structures using high-resolution data (e.g., active source seismic data), tectonic structures over a wide area throughout central Japan have not been characterized in detail. In this study, we estimated 3D S wave velocity structures with high spatial resolution by extracting surface waves from 1-year-long ambient seismic data recorded by the Hi-net high-sensitivity seismograph network. We computed cross spectra of ambient noise data between station pairs and then extracted phase velocity dispersion curves using a frequency domain method. We developed an algorithm to calculate phase velocity and obtained numerous dispersion data. We then estimated the 3D S wave velocity structure by applying a direct surface-wave inversion method. Our results clearly mapped heterogeneous features such as tectonic lines, sedimentary plains, and volcanic systems. We observed some low-velocity anomalies below volcanoes. Active fault zones identified as low-velocity zones were consistent with features on seismic reflection profiles. Several geological blocks that are related to the Izu collision were observed around the Izu Peninsula. We observed previously unreported geological features of the island of Honshu. Our high-resolution S wave velocity model can be used for not only interpretation of geological structures but also hazard assessment of earthquake.Plain Language Summary Techniques based on the speed of seismic waves are used to observe features of the Earth's interior in Japan, but most studies using them have focused on deep structures (tens or hundreds of kilometers deep) over a wide area or on local structures with high-resolution data. The tectonic structures over a wide area throughout central Japan have therefore not been well characterized. Studies of detailed, shallow structures typically use artificial seismic waves; however, these are difficult to do over wide areas due to the cost, safety considerations. In this study, we used only passive sources, generated by ambient (background) vibrations of the Earth, to extract details of the shallow crust and construct a three-dimensional geologic model based on seismic wave speeds in central Japan. Our results clearly mapped geological features such as tectonic lines, sedimentary plains, and volcanoes. An active fault zone was identified as a low-speed zone that agreed with features identified in previous studies using active sources. Several geological blocks were identified around the Izu Peninsula, and the features observed on our results reflected processes of the tectonic evolution. Our results can also be used to predict how earthquake waves propagate and thereby contribute to disaster evaluation.

Seismic ambient noise with frequencies > 1 Hz includes noise related to human activities. A reduction in seismic noise during the COVID-19 pandemic has been observed worldwide, as restrictions were imposed to control outbreaks of the SARS-CoV-2 virus. In this context, we studied the effect of changes in anthropogenic activities during COVID-19 on the seismic noise levels in the Tokyo metropolitan area, Japan, considering time of day, day of the week, and seasonal changes. The results showed the largest reduction in noise levels during the first state of emergency under most conditions. After the first state of emergency was lifted, the daytime noise reverted to previous levels immediately on weekdays and gradually on Sundays. This was likely because economic activities instantly resumed, while non-essential outings on Sundays were still mostly avoided. Furthermore, the daytime noise level on Sundays was strongly reduced regardless of changes on weekdays after the second state of emergency, which restricted activities mainly at night. Sunday noise levels gradually increased from the middle of the second state of emergency, suggesting a gradual reduction in public concern about COVID-19 following a decrease in the number of infections. Our findings demonstrate that seismic noise can be used to monitor social activities.