Predicting the arrival of coronal mass ejections (CMEs) in near-Earth space is a challenging and important problem, as CMEs are the main driver of severe space weather. The Heliospheric Imager (HI) cameras provide observations of the plasma between the Sun and Earth and can be used to track the evolution of CMEs as they flow away from the Sun, toward Earth. Techniques have been developed to use the HI observations to predict the arrival of CMEs in near-Earth space, potentially allowing the HI observations to be used in a space weather forecasting context. We assess how well these methods work for four CMEs observed by the HI cameras. We found that for these four events the techniques that use HI observations do not perform as well as the standard CME forecasting techniques used by the Space Weather Prediction Center. We try to improve the HI-based predictions by using a citizen science approach to develop a better method of tracking the CMEs evolution. The improved CME tracking does not significantly improve the CME predictions and suggests that some of the assumptions in the HI based CME prediction techniques are poor assumptions.

Barnard, L. A., C. A. de Koning, C. J. Scott, M. J. Owens, J. Wilkinson, and J. A. Davies (2017), Testing the current paradigm for space weather prediction with heliospheric imagers., Space Weather, 15, doi:10.1002/2017SW001609.