Guo Qinmeng, Yong Shanshan, Wang Xin'an
The Key Laboratory of Integrated Microsystems, Peking University Shenzhen Graduate School, Shenzhen 518055, China.
Shenzhen Earthquake Monitoring and Prediction Technology Research Center, Peking University Shenzhen Graduate School, Shenzhen 518055, China.
Entropy (Basel). 2021 Mar 30;23(4):411. doi: 10.3390/e23040411.
To verify the relationship between AETA (Acoustic and Electromagnetics to Artificial Intelligence (AI)) electromagnetic anomalies and local earthquakes, we have performed statistical studies on the electromagnetic data observed at AETA station. To ensure the accuracy of statistical results, 20 AETA stations with few data missing and abundant local earthquake events were selected as research objects. A modified PCA method was used to obtain the sequence representing the signal anomaly. Statistical results of superposed epoch analysis have indicated that 80% of AETA stations have significant relationship between electromagnetic anomalies and local earthquakes. These anomalies are more likely to appear before the earthquakes rather than after them. Further, we used Molchan's error diagram to evaluate the electromagnetic signal anomalies at stations with significant relationships. All area skill scores are greater than 0. The above results have indicated that AETA electromagnetic anomalies contain precursory information and have the potential to improve local earthquake forecasting.
为了验证AETA（声学与电磁学至人工智能）电磁异常与当地地震之间的关系，我们对AETA台站观测到的电磁数据进行了统计研究。为确保统计结果的准确性，选取了20个数据缺失少且当地地震事件丰富的AETA台站作为研究对象。采用改进的主成分分析方法获取代表信号异常的序列。叠加时代分析的统计结果表明，80%的AETA台站电磁异常与当地地震之间存在显著关系。这些异常更有可能出现在地震之前而非之后。此外，我们使用莫尔昌误差图来评估具有显著关系的台站的电磁信号异常。所有区域技能得分均大于0。上述结果表明，AETA电磁异常包含前兆信息，具有改善当地地震预测的潜力。
