%0 Journal Article %T Kalman filter and Neural Network methods for detecting irregular variations of TEC around the time of powerful Mexico (Mw=8.2) earthquake of September 08, 2017 %J Journal of the Earth and Space Physics %I Institute of Geophysics, University of Tehran %Z 2538-371X %A Akhoondzadeh Hanzaei, Mehdi %D 2018 %\ 12/22/2018 %V 44 %N 4 %P 87-97 %! Kalman filter and Neural Network methods for detecting irregular variations of TEC around the time of powerful Mexico (Mw=8.2) earthquake of September 08, 2017 %K Earthquake Precursor %K Ionosphere %K Geomagnetic activity %K GPS %K Mexico earthquake %K TEC %R 10.22059/jesphys.2018.258251.1007007 %X In 98 km SW of Tres Picos in Mexico (15.022°N, 93.899°W, 47.40 km depth) a powerful earthquake of Mw=8.2 took place at 04:49:19 UTC (LT=UTC-05:00) on September 8, 2017. In this study, using three standard, classical and intelligent methods including median, Kalman filter, and Neural Network, respectively, the GPS Total Electron Content (TEC) measurements of three months were surveyed to detect the potential unusual variations around the time and location of Mexico earthquake. Every three implemented methods indicated a striking irregular variation of TEC at the earthquake time. However, on the earthquake day, the geomagnetic indices Dst and Ap have exceeded the allowed ranges and even reached maximum values during the studied time period. Besides, the solar index of F10.7 showed high activity around the earthquake day. Therefore, it is difficult to acknowledge the seismicity nature of the detected TEC unusual variations on earthquake day. Therefore, in this case, we encounter a mixed and complex behavior of ionosphere. %U https://jesphys.ut.ac.ir/article_67738_ca53ea2073ad26fad3988d1e06142b16.pdf