5:15 PM - 6:45 PM
[STT38-P04] Characterizing Spatial Distribution of Seismic Intensity to Constrain Hypocenters of Historical Earthquakes
Keywords:Seismic intensity, Kanto region, hypocenter estimation, historical earthquakes, basis function expansion
The spatial distribution of seismic intensity varies depending on the earthquake location, radiation pattern, and underground structure. In general, seismic intensity is expected to exhibit a concentric distribution with the strongest intensity around the epicenter. However, non-concentric seismic intensity has often been observed for earthquakes beneath the Kanto region, Japan (Nakanishi & Horie, 1980 JPE; Nakamura et al., 2007 Historical Earthquakes). Recent research indicates that such non-concentric seismic intensity is primarily caused by three-dimensional heterogeneity in seismic attenuation structure (Nakamura et al., 2023 Zisin). This implies that the spatial pattern of the seismic intensity provides valuable information for investigating earthquake locations in the Kanto region.
Based on the above knowledge, we have initiated the development of a procedure to identify the source locations of historical earthquakes by comparing the seismic intensity recorded in historical documents with that of present-day earthquakes. The only available information on historical earthquakes is their seismic intensity, which is often estimated from the building collapse rate and degree of building damage. In our first report (Ishise et al., 2023 SSJ), we compiled the spatial pattern of seismic intensity for the present-day earthquakes in the Kanto region and confirmed that they vary depending on the source locations.
In this study, we examine an objective evaluation of the similarity in seismic intensity patterns among present-day earthquakes in the Kanto region as the next step. The geometry and density of station networks usually differ between the historical records and the present-day earthquakes. Therefore, we focused on the characteristics of the spatial patterns instead of seismic intensity itself. Here, we targeted ~550 present earthquakes that occurred in the Kanto region from 2001 to 2021. In the examination, we first reconstructed the continuous field of observed seismic intensity for each earthquake by using the DCB expansion. Then, we extracted the DCB coefficients as characters of the seismic intensity pattern. Finally, we conducted the t-SNE clustering (var der Maaten and Hinton, 2008 JMLR) to assess the similarity of the seismic intensity patterns by clustering the DCB coefficients.
The result of clustering analysis demonstrated that earthquakes in close proximately are successfully grouped together based on their seismic intensity patterns. These results suggest that our examined approach can objectively identify source locations based on seismic intensity patterns.
Based on the above knowledge, we have initiated the development of a procedure to identify the source locations of historical earthquakes by comparing the seismic intensity recorded in historical documents with that of present-day earthquakes. The only available information on historical earthquakes is their seismic intensity, which is often estimated from the building collapse rate and degree of building damage. In our first report (Ishise et al., 2023 SSJ), we compiled the spatial pattern of seismic intensity for the present-day earthquakes in the Kanto region and confirmed that they vary depending on the source locations.
In this study, we examine an objective evaluation of the similarity in seismic intensity patterns among present-day earthquakes in the Kanto region as the next step. The geometry and density of station networks usually differ between the historical records and the present-day earthquakes. Therefore, we focused on the characteristics of the spatial patterns instead of seismic intensity itself. Here, we targeted ~550 present earthquakes that occurred in the Kanto region from 2001 to 2021. In the examination, we first reconstructed the continuous field of observed seismic intensity for each earthquake by using the DCB expansion. Then, we extracted the DCB coefficients as characters of the seismic intensity pattern. Finally, we conducted the t-SNE clustering (var der Maaten and Hinton, 2008 JMLR) to assess the similarity of the seismic intensity patterns by clustering the DCB coefficients.
The result of clustering analysis demonstrated that earthquakes in close proximately are successfully grouped together based on their seismic intensity patterns. These results suggest that our examined approach can objectively identify source locations based on seismic intensity patterns.