Analysis of Ionospheric TEC Variations and Prediction of TEC during Earthquakes Using Ordinary Kriging Based Surrogate Model

Abstract

Earthquake indicators provide a huge advantage in the preparation for the calamity and its aftermaths. Total electron content (TEC) is an ionospheric measurement that varies before the possible happenstance of an earthquake. In this research paper, ordinary kriging based surrogate model (OKSM) which is used for forecasting ionospheric TEC variation is adapted as an earthquake indicators in low-latitude and mid-latitude regions. Seven International GNSS Service (IGS) stations from the mid-latitude and low-latitude region are chosen for observation. The stations are from different countries such as Indonesia (BAKO), Ecuador (RIOP), Greece (ORID), Cyprus (NICO), Hawaii (HNLC) and Italy (MATE & MAT1). The OKSM prediction program is performed with previous 2 months of TEC data from the observed date and its accompanying solar parameters such as Planetary K and A index (Kp & Ap), Radio Flux at 10.7 cm (F10.7) and disturbance storm time (DST) index acquired from OMNIWEB data servers. The performance of the model is evaluated using statistical metrics such as root mean square error (RMSE), normalized – RMSE (NRMSE), mean absolute error (MAE), Pearson’s correlation coefficient (PCC) and chi-squared goodness of fit test. The evaluation of the prediction model on the same date shows that the performance deviation of OKSM is in the range of 2–5 TECU and also establishes the fact that the prediction capability of the OKSM is accurate. Application of the OKSM constructed with previous days data collected from high earthquake prone areas by omitting effects of solar storms and solar activities will act as a highly economical and simple early warning indicator of an earthquake.