European Space Weather Week 2025 - cloned July 3

Real-Time Monitoring and AI-Based Forecasting of Ionospheric TEC Over Northern Africa and the Eastern Mediterranean

Not scheduled

20m

Poster APL1 - Space Weather Services and Alerts for End-Users: Bridging Forecasting, Infrastructure, and Communication

Speaker

Hassan Nooreldeen (Egyptian Space Agency (EgSA))

Description

We present a real-time hybrid deep-learning system for nowcasting and forecasting ionospheric Vertical Total Electron Content (VTEC) across Northern Africa and the Eastern Mediterranean, specifically over longitudes 15°–45° E and latitudes 15°–40° N. Our ground-based network comprises four Trimble-pivot GNSS reference stations—New Cairo (EGSA), Marsa Matrouh (MTRH), Arish (ARSH), and New Valley (VRS2)—that continuously process live GNSS observations to derive VTEC. Historical data spanning January 2022 to May 2025 train a convolutional neural network (CNN) for spatial feature extraction and a recurrent neural network (RNN) for temporal forecasting. The model ingests geomagnetic indices and solar parameters alongside fresh 10-minute-interval observations to generate up to 24 hours of ahead predictions (144 steps), updating its forecast every 10 minutes. Deployed at the Egyptian Space Weather Center (ESWC) of the Egyptian Space Agency, this system delivers high-resolution, operational ionospheric products—vital for satellite operators, aviation navigation, and communication infrastructure within the specified Northern Africa and the Eastern Mediterranean sector.