New developments in crustal deformation research using pseudo-strain gauges with GNSS data; Test for the 2016 Kumamoto earthquake (M7.3), Japan

In China, Yu et al. (2021) investigated the relationship between network connectivity and earthquakes using correlation coefficients between stations of multiple borehole strainmeter data and they reported that anomalies tended to increase 20 days before the earthquake. In Japan, 67 strain gauges have been installed, but the number is limited and there is a regional bias. Therefore, we decided to investigate the possibility of using the GNSS continuous observation system (GEONET) deployed by the Geospatial Information Authority of Japan (GSI) at approximately 1,300 locations throughout Japan to construct a pseudo-strain gauge consisting of four GEONET observation points to index the amount of crustal deformation. In this study, we will devise a pseudo-strainmeter using GNSS observation data and develop a method to accurately detect crustal deformation that is a precursor to M7-class earthquakes. Specifically, we analyze GNSS data from past M7-class damaging earthquakes, and investigate the effectiveness of the method for detecting earthquake precursor variations in strain with high accuracy from a case study analysis perspective. Assuming that pseudo-strain gauges exist at the diagonal intersections, we constructed an algorithm to detect anomalies in in-situ strain variation by determining the time variation of the correlation coefficient between strain and two orthogonal components at the diagonal intersections, and targeted the detection capability of earthquake related variation to the 2016 Kumamoto earthquake (M7.3). The area of analysis was the entire Kyushu region, and the spatio-temporal changes in the correlation coefficients of strain and network connectivity were investigated in detail. The results showed that the strain correlation coefficients of the N-S and E-W components at each pseudostrain station before and after the earthquake were closer to 1 the closer the stations were to the epicenter, and that the network coupling degree The increase in network coupling was confirmed at pseudo-strain stations within about 100 km of the epicenter. Seven days before the main shock, an increase in network coupling was confirmed in a linear region extending from the Hinagu-Futagawa fault zone, the epicenter of the foreshock and main shock, to Aso and Oita. Details will be reported at the time of the presentation.