Prediction of Total Electron Content (TEC) variations using space weather data with AI

Prediction of Total Electron Content (TEC) variations using space weather data with AI

Tharun Racharla, Saurabh Sakharkar, Sakshi Rajurkar, Aman Mishra

Information Technology Department Tulsiramji Gaikwad Patil College of engineering and Technology Mohgaon, Maharashtra, India

Abstract—The effect of Space weather is usually linked to disturbances in the ionosphere (gradients in the Total electron content (TEC)). This has significant effect especially for GPS users causing degradation in range measurements, loss of lock by the receiver of the GPS signal. The ionospheric electron density is sensitive to the space weather parameters like sunspot numbers, F10.7, interplanetary magnetic field etc. and shows variation with respect to them along with other parameters like longitude, latitude, hour of day, etc. The variation in the TEC cannot be estimated or calculated, but can be predicted based on previous space weather data. In this study, we propose a novel ionospheric TEC forecasting model developed using Random Forest Regression algorithm. We have used observation data from the International reference Ionosphere (IRI) to model and forecast ionospheric TEC. The prediction model will be hosted on flask serer-web application.

Key words: Total electron content, Random Forest Regression, Global Positioning System, Artificial Intelligence, Machine learning, web application component.