Assessment of signal processing methods for geomagnetic precursor of the 2012 M6.9 Visayas, Philippines earthquake

• Published in: Acta geophysica 2019-10, Vol.67 (5), p.1297-1306
• Main Authors: Yusof, Khairul Adib, Abdul Hamid, Nurul Shazana, Abdullah, Mardina, Ahadi, Suaidi, Yoshikawa, Akimasa
• Format: Article
• Language: English
• Subjects: Diurnal variations; Earth and Environmental Science; Earth Sciences; Earthquake prediction; Earthquakes; Fluctuations; Frequency ranges; Geomagnetic activity; Geomagnetic anomalies; Geomagnetic data; Geomagnetism; Geophysics/Geodesy; Geotechnical Engineering & Applied Earth Sciences; Magnetic anomalies; Mathematical analysis; Methods; Polarization; Power spectral density; Precursors; Ratio analysis; Research Article - Applied Geophysics; Seismic activity; Signal processing; Stations; Storms; Structural Geology
• ISSN: 1895-6572; 1895-7455
• DOI: 10.1007/s11600-019-00319-w

This study assessed two signal processing methods on geomagnetic data to detect precursory signals appearing before the M6.9 Visayas, Philippines earthquake on 6 February 2012. It aimed to compare the polarization ratio analysis method with the diurnal variation ratio method in terms of reliability and effectiveness. The geomagnetic data were obtained from the MAGDAS magnetometer network for Cebu (CEB) and Legazpi (LGZ) stations which served as the primary and remote stations, respectively. The polarization ratio analysis method was performed on the primary station data to obtain power spectral density in an ultra-low-frequency range before finding the ratio of vertical to total horizontal component. Meanwhile, the diurnal variation ratio method was used to calculate the difference between the daily maximum and minimum values. Then, the ratio of the daily differences of the primary station to the remote station for each individual component was calculated. The disturbance storm time index which describes global geomagnetic activity originating from the Sun was utilized to verify that any observed geomagnetic fluctuations were not caused by solar-terrestrial effect. A precursory anomaly was found using the polarization ratio analysis method which appeared 2 weeks before the earthquake. It is concluded that this method might be effective and reliable in detecting geomagnetic anomalies preceding upcoming earthquakes. In contrast, although the diurnal variation ratio method did show perceivable fluctuations, the running averages were not statistically significant to be considered a precursor. The discrepancy between the analytical results of the two methods may be attributed to the detectability of the earthquake being studied which had a relatively low magnitude. Hence, future studies which utilize more earthquake events need to be conducted to reach a definitive conclusion. Graphic abstract