5:15 PM - 7:15 PM
[SSS11-P17] Effects of heterogeneity of subsurface soil on spacial variation of ground motion
Keywords:subsurface soil, heterogeneity, spatial variation, Chiba array
Introduction
It has been pointed out that heterogeneity of the subsurface soil causes spatial variation in ground motion at adjacent points on the surface and affects the response of buildings. On the subsurface soil heterogeneity, though there is one-dimensional information such as boring survey data, there is little information on spatial heterogeneity. In this study, we investigate the spatial heterogeneity of the subsurface soil from acoustic tomography data and analyze the relationship between the heterogeneity of the subsurface soil and the spatial variation of ground motions.
Heterogeneity of Chiba array subsurface soil
In this study, an acoustic tomography survey was conducted on the site where seismometers of Chiba array of the Institute of Industrial Science, the University of Tokyo was located. The arrangement of seismometers at Chiba Array and the survey line of the acoustic tomography are shown in Fig. 1. The survey line and the depth of the acoustic tomography survey are both 50 m. Fig. 2 shows the estimated Vp distribution in the survey section. Since the groundwater level at the site was around GL-15 m, Vp distribution could be estimated below GL-15 m. There is a significant difference in the Vp distribution with the boundary at approximately GL-30m.
Table 1 shows the 1-D soil property model of the site. Layers 1 to 4 are based on the PS logging at the C0 observation site of Chiba Array (Katayama et al., 1990). Based on the acoustic tomography survey, the soil properties of layer 3 were determined from Vp distribution of GL-15m to GL-30m, as well as the soil properties of layer 4 from Vp distribution of GL-30m or deeper. In addition, layers 5 and 6 were added as layers below GL-100 m, based on the JSHIS subsurface structure model.
The coefficient of variation and correlation function for the third and fourth layers were estimated from Vp distribution in Fig. 2. The estimated results of correlation distances are shown in Table 1. For both layers 3 and 4, the correlation distance in the horizontal direction is greater than the vertical direction.
Spatial variation of ground motions
Based on Vs and heterogeneity parameters shown in Table 1, the spatial variation of ground motion at adjacent points was evaluated by the method of Tokumitsu et al. (2024). For Layers 3 and 4, Correlation distances and coefficients of variation of Vp predicted from acoustic tomography was applied. Coefficients of variation for layers 1 and 2 were estimated from the variation in N-values and correlation distances were set to be smaller than those for layers 3 and 4. Coefficients of variation for layers 5 and 6 were set smaller than those for the upper layers, and correlation distances were set larger than those for the upper layers. The spatial variation of ground motions was evaluated using the power spectrum ratio ε2 of the scattering component to the coherent component of ground motions between two points. The estimated results of ln(ε2+1) for a separation distance between two points of 5-30 m are shown in Fig. 3 as solid lines. In addition, ln(ε2+1) at each separation distance estimated from the observation records of the Chiba array are shown in Fig. 3 as dotted lines. The ln(ε2+1) of the observation records is significantly larger. This may be due to the fact that the evaluation of ln(ε2+1) by Tokumitsu et al. (2024) does not include the effect of multiple reflections at the layer boundary, and that ln(ε2+1) is evaluated based on a finite ground depth.
Acknowledgements: We received cooperation from the Facilities and Environment Department, Chiba University in conducting the acoustic tomography survey. We used observation records from Chiba Array released by the Assoc. for Earthquake Disaster Prevention. GMT was used for some of the figures. This study was conducted with the support of a grant of Nuclear Industry Infrastructure Strengthening Project by the Agency for Natural Resources and Energy in FY2024.
It has been pointed out that heterogeneity of the subsurface soil causes spatial variation in ground motion at adjacent points on the surface and affects the response of buildings. On the subsurface soil heterogeneity, though there is one-dimensional information such as boring survey data, there is little information on spatial heterogeneity. In this study, we investigate the spatial heterogeneity of the subsurface soil from acoustic tomography data and analyze the relationship between the heterogeneity of the subsurface soil and the spatial variation of ground motions.
Heterogeneity of Chiba array subsurface soil
In this study, an acoustic tomography survey was conducted on the site where seismometers of Chiba array of the Institute of Industrial Science, the University of Tokyo was located. The arrangement of seismometers at Chiba Array and the survey line of the acoustic tomography are shown in Fig. 1. The survey line and the depth of the acoustic tomography survey are both 50 m. Fig. 2 shows the estimated Vp distribution in the survey section. Since the groundwater level at the site was around GL-15 m, Vp distribution could be estimated below GL-15 m. There is a significant difference in the Vp distribution with the boundary at approximately GL-30m.
Table 1 shows the 1-D soil property model of the site. Layers 1 to 4 are based on the PS logging at the C0 observation site of Chiba Array (Katayama et al., 1990). Based on the acoustic tomography survey, the soil properties of layer 3 were determined from Vp distribution of GL-15m to GL-30m, as well as the soil properties of layer 4 from Vp distribution of GL-30m or deeper. In addition, layers 5 and 6 were added as layers below GL-100 m, based on the JSHIS subsurface structure model.
The coefficient of variation and correlation function for the third and fourth layers were estimated from Vp distribution in Fig. 2. The estimated results of correlation distances are shown in Table 1. For both layers 3 and 4, the correlation distance in the horizontal direction is greater than the vertical direction.
Spatial variation of ground motions
Based on Vs and heterogeneity parameters shown in Table 1, the spatial variation of ground motion at adjacent points was evaluated by the method of Tokumitsu et al. (2024). For Layers 3 and 4, Correlation distances and coefficients of variation of Vp predicted from acoustic tomography was applied. Coefficients of variation for layers 1 and 2 were estimated from the variation in N-values and correlation distances were set to be smaller than those for layers 3 and 4. Coefficients of variation for layers 5 and 6 were set smaller than those for the upper layers, and correlation distances were set larger than those for the upper layers. The spatial variation of ground motions was evaluated using the power spectrum ratio ε2 of the scattering component to the coherent component of ground motions between two points. The estimated results of ln(ε2+1) for a separation distance between two points of 5-30 m are shown in Fig. 3 as solid lines. In addition, ln(ε2+1) at each separation distance estimated from the observation records of the Chiba array are shown in Fig. 3 as dotted lines. The ln(ε2+1) of the observation records is significantly larger. This may be due to the fact that the evaluation of ln(ε2+1) by Tokumitsu et al. (2024) does not include the effect of multiple reflections at the layer boundary, and that ln(ε2+1) is evaluated based on a finite ground depth.
Acknowledgements: We received cooperation from the Facilities and Environment Department, Chiba University in conducting the acoustic tomography survey. We used observation records from Chiba Array released by the Assoc. for Earthquake Disaster Prevention. GMT was used for some of the figures. This study was conducted with the support of a grant of Nuclear Industry Infrastructure Strengthening Project by the Agency for Natural Resources and Energy in FY2024.