A simple but high-resolution ground exploration method for building design

本研究的目的是建立一种简单而高精度的利用微震探测地表浅层横波速度结构的方法。为此，首先对观测方法和观测设备进行了改进，使观测变得简单、机动。然后使用这些设备，我们开发了一个系统，可以非常简单地预测站点放大特性。在第一年，我们将笔记本电脑和PCMCIA卡型AD转换器相结合，开发了一个数据采集系统，使在线查看数据成为可能。利用该系统在K-NET和KiK-net观测点进行了第二年和第三年的微震测量，并将这些数据与以往的观测数据进行了分析。并与光谱反演得到的S波场放大特征和S波速度结构反演得到的理论H/V特征进行了比较。结果表明，谱反演方法获得的微震H/V比峰值频率与地震运动场地特征峰值频率平均存在约1:1的关系。但变异较大，对数标准差为2.12倍或1/2.12倍。在绘制地震运动的峰值频率与微震H/V比值的关系时，发现当微震H/V峰值水平超过4倍时，我们可以很好地预测峰值频率，误差在±30%以内。然而，当微震H/V比峰值电平与地震运动峰值电平比较时，微震峰值电平明显小于地震运动峰值电平，两者之间没有很好的相关性。结果表明，当土-地面阻抗对比较大时，微震的H/V比值可以可靠地用于检测峰值频率，但其峰值水平不能作为地震现场放大的指标。少

The objective of this research is to establish a simple yet highly precise exploration method for shallow S-wave velocity structures of ground by using microtremors. For this purpose, the observation method and observation equipment were first improved, which would enable us simple and mobile observation. Then using that equipment we develop a system that can predict quite simply site amplification characteristics. In the first year by combining a notebook PC and PCMCIA card type AD converter, we develop a data acquisition system that makes it possible to see data online. In the second and third year microtremor measurements at observation points of K-NET and KiK-net were carried out using this system, and we analyzed these data together with the observed data obtained in the past. Comparisons with the S-wave site amplification characteristics obtained by the spectral inversion analysis and theoretical H/V characteristics by the S wave velocity structures, which were inverted from the … More site amplification characteristics, were carried out. As a result, it turned out that about 1:1 relation between the peak frequency of microtremor H/V ratios and the peak frequency of the site characteristics of the earthquake motions acquired by the spectral inversion method exists on the average. However, the variation was large and logarithmic standard deviation was found to be 2.12 times or 1/2.12 times. When we plot relationship between the earthquake motions' peak frequencies versus those of microtremor H/V ratios, it is found that we can well predict peak frequency within ±30% of error if the H/V peak level of microtremors exceeds 4 times. However, when the peak levels of microtremor H/V ratios and those of earthquake motions were compared, the levels of microtremor peaks were remarkably smaller than those of earthquake motions, and there is no good correlation between them. Thus we found that the H/V ratios of microtremors can be reliably used for detecting peak frequency if the impedance contrast of soil ground is large, but their peak levels could not be used as indicator of the earthquake site amplification. Less