Investigation on the prediction of Large Scale Traveling Ionospheric Disturbances over Europe

Abstract

During geomagnetic storm events a large amount of energy is transferred from the solar wind and interplanetary magnetic field into the Earth's magnetosphere-ionosphere-thermosphere (MIT) system. Such storm periods are closely associated with ionosphere and thermosphere perturbations, including the occurrence of irregularities at highlatitudes, changes in the thermosphere circulation, deviations of electron densities in the ionosphere from the quiet conditions and Large Scale Traveling Ionospheric Disturbances (LSTIDs), which are the signature of atmospheric gravity waves that are generated at high-latitudes due to Joule heating. Such disturbances propagate equatorward with wavelengths greater than 1000 km, periods between 30 min and 3 h and horizontal velocities ranging from 400 to 1000 m/s. Throughout the years, many investigations have been conducted using Total Electron Content (TEC) measurements derived from Global Navigation Satellite Systems (GNSS) to monitor and track LSTIDs. Although different statistical studies have been performed to characterize such disturbances, a reliable operational model for predicting LSTIDs has not been developed yet. In this scenario, we propose a methodology based on machine learning for the prediction of LSTIDs. In our study, we consider data from geomagnetic storms events ranging from 2001 up to 2017 and we use solar wind information obtained from the Lagrange point L1 as one of the input information to predict LSTIDs activities. Our results show that such approach has a good potential for LSTIDs prediction activities at mid-latitudes.