5:15 PM - 6:45 PM
[SGD02-P13] Effects of three-dimensional heterogeneity on post-seismic deformation of inland earthquakes: a case study of the 2016 Kumamoto earthquake.
Keywords:Crustal movement, The 2016 Kumamoto earthquake, Post-seismic deformation, Viscoelastic relaxation
The occurrence of inland earthquakes remains largely enigmatic, with emerging evidence suggesting a correlation with heterogeneous crustal structures. This correlation posits that stress concentrations in the upper crust are a consequence of deformations in the lower crust (Iio, 2009). Crucially, the protracted spatiotemporal dynamics of stress and strain distributions are purportedly influenced by the heterogeneous viscosity in the Earth (e.g., Yamasaki and Seno, 2005).
This research was centered on the April 16, 2016 Kumamoto earthquake in Kumamoto Prefecture, Japan. Our objective was to model viscoelastic relaxation in a three-dimensional heterogeneous body, extending the observational period beyond the immediate post-seismic phase (Moore et al., 2017; Pollitz et al., 2017), which employed postseismic data to deduce the subsurface viscosity across Kyushu. Utilizing data from GEONET observation stations in Kyushu, processed by the Geospatial Information Authority of Japan, we first executed detrending and common mode error elimination from the time series data. Our analysis focused on estimating three-dimensional displacements attributable to viscoelastic relaxation, covering the period up to April 16, 2023, while excluding data proximate to the earthquake's occurrence. These estimates were subsequently juxtaposed with simulations of Maxwell viscoelastic relaxation via a semi-analytical boundary element method (Barbot and Fialko, 2010). Our experimental design encompassed a range of scenarios, each tailored to account for the complex three-dimensional heterogeneities that extend beyond the basic stratification of the upper crust, lower crust, and mantle.
In analyzing the data, we specifically focused on the residuals between the observational data and the numerical simulations for each scenario. This approach revealed that the inclusion of a significant low-viscosity zone, reminiscent of the Beppu-Shimabara graben, notably improved the accuracy of displacement predictions in the northwest region of Kyushu compared to models that did not incorporate such a structure. However, within the central region of Kyushu, particularly within the Beppu-Shimabara graben, the analysis highlighted persistent discrepancies between the observational data and the numerical outputs, especially in terms of horizontal displacement. To account for this inconsistency, we hypothesized a localized increase in divergent movement within the central Kyushu area of the Beppu-Shimabara graben following the Kumamoto earthquake.
This research was centered on the April 16, 2016 Kumamoto earthquake in Kumamoto Prefecture, Japan. Our objective was to model viscoelastic relaxation in a three-dimensional heterogeneous body, extending the observational period beyond the immediate post-seismic phase (Moore et al., 2017; Pollitz et al., 2017), which employed postseismic data to deduce the subsurface viscosity across Kyushu. Utilizing data from GEONET observation stations in Kyushu, processed by the Geospatial Information Authority of Japan, we first executed detrending and common mode error elimination from the time series data. Our analysis focused on estimating three-dimensional displacements attributable to viscoelastic relaxation, covering the period up to April 16, 2023, while excluding data proximate to the earthquake's occurrence. These estimates were subsequently juxtaposed with simulations of Maxwell viscoelastic relaxation via a semi-analytical boundary element method (Barbot and Fialko, 2010). Our experimental design encompassed a range of scenarios, each tailored to account for the complex three-dimensional heterogeneities that extend beyond the basic stratification of the upper crust, lower crust, and mantle.
In analyzing the data, we specifically focused on the residuals between the observational data and the numerical simulations for each scenario. This approach revealed that the inclusion of a significant low-viscosity zone, reminiscent of the Beppu-Shimabara graben, notably improved the accuracy of displacement predictions in the northwest region of Kyushu compared to models that did not incorporate such a structure. However, within the central region of Kyushu, particularly within the Beppu-Shimabara graben, the analysis highlighted persistent discrepancies between the observational data and the numerical outputs, especially in terms of horizontal displacement. To account for this inconsistency, we hypothesized a localized increase in divergent movement within the central Kyushu area of the Beppu-Shimabara graben following the Kumamoto earthquake.