Statistical analysis of the ionosphere using Bayes' theorem and Naive Bayes classifier

The ionosphere is an important layer of the atmosphere that lies between 60 and 1000 km altitude and has a density of 1012 electrons per cubic meter, ionized to the plasma state by radiation from the sun. The most determining parameter of ionospheric plasma is the electron density, which shows variability and correlation with solar, geomagnetic and seismic activity and solar flares, sunspot number, solar wind, geomagnetic storms. An important measurable quantity of electron density is the Total Electron Content (TEI), which provides an efficient way to investigate the structure of the ionosphere and upper atmosphere. TEC is defined as the line integral of electron density along a beam path or the total number of electrons along a beam path. In the last decade, the Global Positioning System (GPS), which has a worldwide receiver network, provides an easy way to predict TEC (GPS‐TEC). The spatial-temporal variability of the ionosphere is also affected by the spatial-temporal trends and disturbances in the geomagnetic field, gravitational waves and seismic activities coupled to the upper atmosphere and ionosphere. Some of these variations produce wave-like oscillations that propagate in the ionosphere at a certain frequency, duration, and speed. In this study, Bayes Theorem and Naive Bayes Classifier are used to detect the disturbances in the ionosphere, disturbances due to seismic, solar, geomagnetic activities and deviations from the quite state of the ionosphere. Naive Bayes classifier and Bayes Theorem is applied to the IONOLAB-TEC data obtained from GPS stations located in Turkey during 1999 solar eclipse and Marmara Earthquake.