利用不同方法估算流动台站的场地响应

本文以2003年7月21日云南大姚6.2级地震序列为例, 分别利用地面运动反演法(Moya方法)、 水平与垂直向之比法(Nakamura方法)和参考台站法研究大姚地区场地响应特征。主要结论有: ① Moya方法计算的场地响应结果表明, 大姚地区场地的放大作用是比较明显的, 平均在3~20倍, 最高达到40倍; 卓越频率位于3~6 Hz之间, 但在高频处(约10 Hz以上)场地响应衰减得比较明显; 不同岩性台基的台站, 其场地响应结果存在一定的差别, 台基为风化砾岩的场地响应结果比较平缓, 其场地响应值比台基为土层的其他台站小一些。② 入射角在40°以上和以下范围, Nakamura方法得到的场地响应形态有差异, 但在各自范围内, 场地响应与入射角的依赖关系不明显。③ 利用同一台站不同时间记录到的地震资料, 由三种方法各自得到的场地响应都具有较好的稳定性; 由于垂直分量仍具有放大作用(在10 Hz以下), Moya的结果大于Nakaruma的结果, 但两者形态上一致性较好; 参考台站法由于受参考台站的影响, 计算得到的场地响应形态和其他两种方法差异较大。④ Moya方法是依赖于震源模型的方法, 它能给出绝对的场地响应值, 其它两种方法的结果是一种相对场地响应, 受参考对象的影响较大。

Estimation of Site Responses for Temporary Stations Using Various Methods

Site response characteristics in Dayao area were analyzed using various methods such as ground-motion-inversion method (Moya′s method) and horizontal-to-vertical-spectral ratio (Nakamura′s method) and reference-site method, using the records of earthquake sequences of the 21 July 2003 Dayao (M6.2), Yunnan province. Our results show that ① Site response results by Moya′s method indicate that the site amplifications in the Dayao area are apparent, between 3 and 20 times, up to 40 times at the most. The dominant frequency is between 3 and 6 Hz, but the site responses decay very fast at the high frequency (≥10 Hz). There are some differences among the site responses for the stations with different rock base. Site responses appear more flat for the stations with efflorescent conglomerate base and their values are smaller than other stations with soil base. ② Site responses by Nakamura′s method with the angle of incidence bigger or smaller than 40°appear to be different, but with no obvious dependence on the angle of incidence in its own ranges. ③ Site responses by these three methods with the seismic data recorded by the same station in different times appear to have good stability. Site responses by Moya′s method are bigger than that of Nakamura′s method at the frequency <10 Hz due to the existing of amplification in the vertical component, but their forms are coherent. ④ Moya′s method is a source-dependent approach to estimate absolute site effects and the other two methods are reference-dependent approaches to obtain relative site effects and are deeply affected by the reference site.