利用早期余震预测的Omi-R-J方法对2017年四川九寨沟MS7.0地震的应用研究

在地震序列的早期阶段获得科学可靠的余震预测结果，始终是地震预测研究的前沿课题.针对新近发展的Omi-R-J方法的适用性问题，以及相比于传统的Reseanberg-Jones （R-J）模型在地震序列参数拟合、余震预测效能上的差异等问题，本文以2017年四川九寨沟M_S7.0地震序列为例，利用多时间窗的连续滑动拟合、预测和检验，以及通过构建重采样随机地震序列目录的系统检验等方式，开展了应用研究和比较研究.结果表明：相比于R-J模型，Omi-R-J方法可在余震记录较不完整的地震序列早期阶段获得稳定、可靠的序列参数，p值、c值和b值的标准差相比于R-J模型参数明显减小.N-test方法检验结果表明，在R-J模型等传统方法无法获得预测结果的地震序列早期阶段，Omi-R-J方法预测结果可以较高的比例通过N-test检验，在后续时段的预测效果也明显优于R-J模型.利用随机地震序列目录的测试结果表明，余震记录完整性下降对Omi-R-J方法预测结果的影响相对较小，在全部时段的完整性下降的影响高于在部分时段的完整性下降.上述结果对进一步地将Omi-R-J方法应用在震后早期余震预测中具有一定科学借鉴意义.

Application of the Omi-R-J method for forecast of early aftershocks to the 2017 Jiuzhaigou,Sichuan, MS7.0 earthquake

It is a focused issue to forecast early aftershocks. This paper discusses the applicability of the recently developed Omi-R-J method, and compares it with the traditional Reseanberg-Jones (RJ) model in parameters fitting and the effect of aftershock forecasting. Taking the Jiuzhaigou M_S7.0 earthquake sequence as an example, we use the sliding-type parameters fitting, forecasting, and statistical tests in multiple time windows, and systematic statistical tests of randomly resampled aftershock sequence to conduct the comparison. The results show that stable and reliable parameters of aftershock sequence can be obtained by using the Omi-R-J method in the early stage with significant missing events. The standard deviations of the p-value, c-value, and b-value of the Omi-R-J method are significantly reduced compared with the R-J model. The N-test results show that we can obtain a satisfied forecast result by using the Omi-R-J method in the early stage of the aftershock sequence for which the traditional methods such as the R-J model cannot be performed, and the forecasting effect is also better than the R-J model in the subsequent time period. The results of a systematic test by using the randomly resampled aftershock sequence show that the incomplete record of the aftershock sequence has a relatively small impact on the Omi-R-J method. Among them, the decrease in completeness during the whole period has a greater impact on the forecasting result than that in the partial period. These results have certain significance for further application of the Omi-R-J method to the forecast of early aftershocks.