The ionosphere directly affects GPS accuracy, radio communications, aviation navigation, and power grid stability. Predicting ionospheric disturbances 1–6 hours ahead would give operators actionable warning time. Every relevant data source (solar wind, geomagnetic indices, GPS-derived electron content measurements) exists in different formats, cadences, and coordinate systems with significant missing data. No unified, ML-ready benchmark dataset exists. This project builds it, then proves it works by training a forecasting model on top.
