自动捡拾P波到时综合方法的选取与探讨

快速、准确以及可靠的震相自动识别，不仅可为政府震后决策提供快速可靠的地震信息，还对减轻地震灾害损失和提高公众对政府可信度具有较大价值。以云南强震动台网实际观测记录为基础，选取了2008年至2017年期间震级在M5.0至M7.0间共计20余次地震事件，借鉴国内外P波震相自动拾取的相关研究，用最常用的长短时平均STA/LTA结合AIC准则综合捡拾法和长短时平均STA/LTA结合BIC准则综合捡拾法这两种不同的综合分析方法，将涵盖了云南盈江、腾冲、彝良、洱源和景谷等地震多发区域的记录P波到时捡拾，并对捡拾准确度、可靠度以及相应速率进行对比探讨。统计分析结果表明:在精确捡拾部分中，相比AIC准则，BIC准则的构架与算法更加灵活简单，且其抗干扰信号能力强，能有效避免干扰信号引起的误触发，可在漏捡拾与误捡拾之间寻求最佳平衡，对地震数据实现快速有效的实时处理，更利于云南省内地震预警发展。

Discussion on the synthetic method of automatically picking up of P wave

The fast, accurate and reliable automatic identification of seismic phase not only provides the government with fast and reliable seismic information for post-earthquake decision-making, but also has great value for reducing earthquake disaster losses and improving the public credibility of the government. In this paper, based on the actual ground motions observed in more than 20 earthquake events by strong motion network of Yunnan, during 2008 to 2017 with the magnitudes between ML5.0 to ML7.0. Referring existing methods of automatic pick up of P wave phase, the most common STA/LTA method combined with AIC criterion and combined with BIC criterion are used respectively for comparison. The analyzed data covers the ground motions in Yingjiang, Tengchong, Yiliang, Eryuan, and Jinggu in Yunnan province. The accuracy, reliability and efficiency of the two synthetic pick up methods are compared. Statistical analysis results show that compared with the AIC criterion, the architecture and algorithm of BIC criterion is more flexible, simple and the anti-interference ability is stronger, which can effectively avoid the interference signal caused by false triggering, leakage pick up can be avoided and achieve the optimal balance between the false pick up. The fast real-time processing of seismic data benefits to earthquake early-warning development in Yunnan province.