基于地震记录的震动图校正方法研究

Research on Correction Method for ShakeMap Based on Seismic Data

针对地震动强度速报工作,利用地震动衰减关系和地震台站记录到的数据研究地震动强度衰减中地震事件内部(intra-earthquake)的不确定性.通过台站实际观测值与理论计算值之间的残差分析,提出了利用台站数据对缺少台站地区理论估计值的对数偏差校正方法.研究分析表明,线性偏差校正方法对系统偏差校正的效果不显著,与使用汪素云2000年回归的我国东部峰值加速度(PGA)衰减关系的结果相当;针对特定震级的地震不约束R0(近场饱和因子)拟合的衰减关系,由于近场台站分布的不均匀性,使得R0过大,衰减关系对近场地震动参数缺乏控制;对数偏差校正方法的结果与固定R0拟合地震动衰减关系的结果大致相当,适合于地震动强度速报中缺少台站数据的地区地震动参数理论估计值的系统偏差校正.最后以2010年3月4日中国台湾屏东6.4级地震为例,比较了点圆和点椭圆衰减关系模型的对数残差及震动图的分布形态,说明衰减关系模型对震中地区地震动参数估计值的分布影响较大,并利用对数偏差校正方法快速生成震动图展示了该校正方法的可靠性.

Uncertainty of intra-earthquake for the quick reported seismic events was investigated with ground motion attenuation relationships and seismic recorded data. The method of logarithmic deviation correction for theoretical estimates was proposed by using the residuals analysis between actual observation data and the calculated value. Research and analysis shows that the method of linear bias correction has no significant effect on systematic bias, which is as much as Wang2000 attenuation relations of Peak Ground Acceleration （PGA） in eastern China Because of the uneven distribution of stations near the source, R0 was too large, when an attenuation relation of specific magnitude earthquake was fitted with an unfixed R0. This attenuation relationship could not reveal distribution law of ground motion parameter near the source. However, results of logarithmic bias correction was as well as that of ground motion attenuation relationship fitted by a fixed R0, and which was more suitable to error correction of ground motion parameter in areas lack of station data for ShakeMap. Finally, the results of March 4,2010 Pingdong ME 6.4 earthquake show that the models of attenuation relationship have great effect on the distribution of ground motion parameters near the source by comparison attenuation relationship of the point-circle model with that of the point-ellipse model. And the results also show that method of logarithmic deviation correction is reliable for shakemap followed great earthquake.