【作者】 宋晋东；

【导师】 李山有；

【作者基本信息】 中国地震局工程力学研究所 ， 防灾减灾工程及防护工程， 2007， 硕士

【摘要】 地震预警是近年来国际上备受关注的减少人员伤亡和经济损失的手段之一,本文围绕地震预警中的震级快速估算开展了以下工作:1.介绍了地震预警的基本思想、国内外地震预警系统的建设现状以及相关技术研究进展。2.在长短项比法识别震相的基础上,研究了应用线性拟合法由阈值触发点到时回推地震波真实到时的方法。有限地震记录的计算结果表明,该方法具有较高的识别精度。3.在详细介绍地震预警研究中常用的瞬时特征周期方法和平均周期方法估算地震震级的基础上,应用实际地震记录对两方法进行了比较分析,结果表明平均周期方法优于瞬时特征周期方法、带通滤波后的估算结果比未滤波的估算结果更精确。4.在日本学者Odaka的B-Δ方法基础上,提出将平均周期作为一个参数引入到B-Δ方法中,从而发展了一种由幅值与周期联合估算震级的新方法。实际地震记录的计算结果表明,改进后的方法较原有方法提高了估算震级与实际震级间的拟合相关系数,降低了拟合标准偏差,在一定程度上改善了大震级处的饱和现象。更多还原

【Abstract】 As a very important means for preventing and reducing earthquake disaster, earthquake early warning (EEW) has received considerable attention from international society in recent years. The rapid magnitude estimation of earthquake early warning is revolved in this paper as follows:1. Basic idea of the earthquake early warning, domestic and foreign earthquake early warning system construction present situation, as well as technique advances in relative research are introduced.2. On the basis of seismic phase recognition by using STA/LTA method, the method by using linearity fitting from arriving time at threshold to actual arriving time of seismic wave is studied. The application results of real seismic records indicate that this method has higher identification accuracy.3. Two estimation methods, the instantaneous predominant period method and the average period method, commonly used in earthquake early warning research are detailed introduced. In this article, the two methods are compared using actual earthquake records, and the results indicate that the average period method is superior to the instantaneous predominant period method as well as the results after data filtered are more accuracy than the un-filtered estimate results.4. In this paper, based on the B-Δmethod of the Japanese scholar Odaka, we propose a new earthquake magnitude estimation method united amplitude and period, by taking the average period as a parameter. The results of actual earthquake records indicate that, the improved method has less discrepancy and simultaneously the underestimate at great magnitude is reduced.更多还原