Ionospheric response to the 2004 Sumatra-Andaman earthquake: evaluation of a novel histogram-based metric

Abstract

In this study, we investigated the effectiveness of a new metric for determining ionospheric perturbations before and on the day of the 26 December 2004 Sumatra-Andaman earthquake (EQ), the longest EQ recorded in modern history. For this purpose, we used the SlantTotal Electron Content (TEC) (STEC) values from the SAMP station, the closest International GNSS System (IGS) station to the epicenter. The trend was estimated using the Least Squares Polynomial Fitting (LSPF) method, and the residual time series were transformed into histogram vectors. Additionally, the Sliding Window Interquartile Range (SW-IQR) method was applied to better characterize local TEC anomalies. A new metric (M-metric) based on these histogram vectors was proposed. EQ-related ionospheric anomalies were compared using both metrics for the same Pseudo Random Noise (PRNs) on the day of the EQ and one day before. With the M-metric, significant perturbations were detected on the EQ day, following the EQ time, at PRN-13 and PRN-23, in line with the literature. According to the Differential Rate of TEC (DROT) method, medium-scale traveling ionospheric disturbances (TID) were observed at PRN-3 and PRN-23 on the EQ day, while large-scale TID (LSTID) were detected at PRN-11, PRN-13, and PRN-20. One day before the EQ, LSTID occurred at PRN-11 and middle-scale TID (MSTID) occurred at PRN-20. However, as noted in the literature, no significant perturbation was observed on either the day of the EQ or the day before, especially at PRN-11. In conclusion, the new metric, in combination with SW-IQR, successfully detected ionospheric perturbations from this EQ, demonstrating improved (c) 2025 COSPAR. Published by Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.