The combination of the Wang‐Sheeley‐Arge (WSA) coronal model, ENLIL heliospherical model version 2.7, and CONED Model version 1.3 (WSA‐ENLIL with CONED Model) was employed to form ensemble forecasts for 15 halo coronal mass ejections (halo CMEs). The input parameter distributions were formed from 100 sets of CME cone parameters derived from the CONED Model. The CONED Model used image processing along with the bootstrap approach to automatically calculate cone parameter distributions from SOHO/LASCO imagery based on techniques described by Pulkkinen et al. (2010). The input parameter distributions were used as input to WSA‐ENLIL to calculate the temporal evolution of the CMEs, which were analyzed to determine the propagation times to the L1 Lagrangian point and the maximum Kp indices due to the impact of the CMEs on the Earth's magnetosphere. The Newell et al. (2007) Kp index formula was employed to calculate the maximum Kp indices based on the predicted solar wind parameters near Earth assuming two magnetic field orientations: a completely southward magnetic field and a uniformly distributed clock‐angle in the Newell et al. (2007) Kp index formula. The forecasts for 5 of the 15 events had accuracy such that the actual propagation time was within the ensemble average plus or minus one standard deviation. Using the completely southward magnetic field assumption, 10 of the 15 events contained the actual maximum Kp index within the range of the ensemble forecast, compared to 9 of the 15 events when using a uniformly distributed clock angle.
Abstract © AGU
Emmons, D. J., Acebal, A. O., Pulkkinen, A., Taktakishvili, A., MacNeice, P., & Odstrcil, D. (2013). Ensemble forecasting of coronal mass ejections using the WSA-ENLIL with CONED Model. Space Weather, 11(3), 95–106. https://doi.org/10.1002/swe.20019