2008年岩手・宮城内陸地震の震源域を対象とした計測震度トモグラフィ解析による減衰構造不均質分布の推定

Abstract

Listen

Translate article

In the 2008 Iwate-Miyagi Nairiku Earthquake occurred at volcanic area, large amplitude strong ground motions were recorded at observatories near fault region. Because subsurface structure is very complicated in the volcanic area, it is pointed out that considering heterogeneous attenuation structure is important to investigate characteristics of earthquake ground motions. In this study, we conducted tomographic analysis using seismic intensity data of aftershock records for the purpose of estimating heterogeneous attenuation structure. We also tried to reproduce seismic intensity of aftershocks which occurred near and on the asperity of the main shock. By comparing estimated and observed seismic intensity, we indicated how the heterogeneous attenuation structure influenced amplitudes of earthquake ground motions. The tomographic analysis using seismic intensity was conducted in three phases of procedures. First, we obtained an attenuation relationship of seismic intensity using strong motion records of inter-plate and intra-plate earthquakes, and evaluated a site factor of each station. Next, we obtained the attenuation relationship for crustal earthquakes using records of aftershocks assuming the same site factor. We then calculated residuals of each observed and estimated seismic intensity data. Finally, supposing residuals to be attributed to the heterogeneous attenuation structure, a block inversion was conducted dividing the focal area of the main shock to strip form with horizontal sizes of approximately 3km. An attenuation level of each block was expressed as increase/decrease ratio of seismic intensity (we called “G” value). In this block inversion weighting by the ray length which passes through the block, 326 records of Aratozawa Dam, Kurikoma Dam and KiK-net stations IWTH25, IWTH26, MYGH02, IWTH28 and AKTH04 which located near and on the fault plane of main shock were used. To confirm a resolution of G value estimated at each block, we conducted a numerical experiment using checkerboard. Although large errors were found where the number of ray was sparse, errors were less than 10% in most domains on the fault plane of the main shock. We conducted the same inversion using observed records, and found that estimated heterogeneous attenuation structure was very complicated. As indicated in the previous studies, high attenuation zone was distributed around Mt. Kurikoma. Around IWTH25 which located just above side of the main shock hypocenter, low attenuation zone was distributed. Around Aratozawa Dam and Kurikoma Dam, attenuation structure was very complicated too. We tried to reproduce seismic intensity at IWTH25, Aratozawa Dam and Kurikoma Dam for 3 aftershocks which occurred near and on the asperity of the main shock. When uniform attenuation structure was used, the observed seismic intensity of IWTH25 and Aratozawa Dam were larger than estimated. Observed seismic intensities at Kurikoma Dam were slightly smaller than estimated. In the case of considering heterogeneous attenuation, reproducibility of observed records was greatly improved. In case that azimuths of hypocenters are different, amplitudes of the earthquake ground motions are different even if magnitude and hypocentral distance are the same. We concluded that the propagation path greatly influences amplitudes of earthquake ground motions. It is thought that this complicated attenuation structure was one of factors that large amplitude strong ground motions were recorded at particularly IWTH25 and Aratozawa Dam in the main shock.