Quantitative comparison of methods for predicting the arrival ofcoronal mass ejections at Earth based on multiview imaging

We investigate the performance of six methods for predicting the coronal mass ejection (CME) time of arrival (ToA) and velocity at Earth using a sample of nine Earth-impacting CMEs between March 2010 and June 2011. The CMEs were tracked continuously from the Sun to near Earth in multiviewpoint imaging data from STEREOSun-Earth Connection Coronal and Heliospheric Investigation (SECCHI) and SOHOLarge Angle and Spectroscopic Coronagraph (LASCO). We use the Graduate Cylindrical Shell model to estimate the three-dimensional direction and height of the CMEs in every image out to 200R. We fit the derived three-dimensional (deprojected) height and time (HT) data with six different methods to extrapolate the CME ToA and velocity at Earth. We compare the fitting results with the in situ data from the Windspacecraft. We find that a simple linear fit above a height of 50R gives the ToA with an error ±6h for seven (78%) of the CMEs. For the full sample, we are able to predict the ToA to within ±13h. These results are a half day improvement over past CME arrival time methods that only used SOHO LASCO data. We conclude that heliographic height-time measurements of the CME front made away from the Sun-Earth line and beyond the coronagraphic field of view are sufficient for reasonably accurate predictions of their ToA. Key Points We study six methods for predicting the arrival time of nine CMEs at Earth The CMEs were tracked in multi-viewpoint data from SECCHI and LASCO Our results are a half day improvement over CME methods that only used LASCO