Enhancing STEREO-HI data with machine learning for efficient CME forecasting

Observing and forecasting Coronal Mass Ejections (CME) is crucial due to the potentially strong geomagnetic storms generated and their impact on satellites and electrical devices. With its near-real-time availability, STEREO-HI beacon data is the perfect candidate for efficient forecasting of CMEs. However, previous work concluded that prediction based on beacon data could not achieve the same accuracy as with high-resolution science data due to data gaps and lower quality. We have introduced a new method to improve the resolution and quality of near-real-time beacon data by using advanced machine-learning techniques while maintaining consistency between consecutive frames. This method also allows us to forecast intermediary and subsequent frames using a data-driven model for CME propagation within HI images. The output generated by our model produces smoother and more detailed time-elongation plots (J-plots) that are used as input for the Ellipse Evolution model based on Heliospheric Imager observations (ELEvoHl). We have compared the data produced by our model with the science data and analysed its impact on CME forecasting and propagation.