Statistical analysis of TEC anomalies as earthquake precursors using GPS data for the case study of Assam, India

Abstract

In Northeast India, 17 earthquakes with Mw > 4.1 were recorded, with 14 of them preceded by significant anomalies in ionospheric Total Electron Content (TEC). Probability analysis showed that if TEC anomalies exceed 2σ for over 16 days, there is an 80 % likelihood of an earthquake. Multiple linear regression analysis revealed a low R² (∼0.017) between hourly TEC and factors like Dst, Ap-Index, IMF-Bz, and F10.7. However, the models for hourly data were significant (p ∼2.08x10^-14). Hence, we used multiple regressions on daily data to identify hours when independent variables best explained TEC variability. The study aims to develop a model to determine the epicentral distance, precursory time, and magnitude of potential near-field earthquakes in the Tezpur (Eastern Himalaya) region, based on TEC anomaly peaks. However, Mw > 4.1 earthquakes may not always show TEC perturbations due to factors like geography, solar activity, and radon. The findings are thoroughly discussed in the manuscript.