Comparisons of CME Morphological Characteristics Derived from Five 3D Reconstruction Methods

Abstract

We compare different methods to reconstruct three-dimensional (3D) coronal mass ejection (CME) morphology. The explored methods include geometric localisation, mask fitting, forward modelling, polarisation ratio, and local correlation tracking plus triangulation. These five methods are applied to the same CME event that occurred on 7 August 2010. Their corresponding results are presented and compared, especially in their propagation direction and spatial extent in 3D. We find that the mask fitting and geometric localisation methods produce consistent results. Reconstructions including three-view observations are more precise than reconstructions done with only two views. Compared to the forward modelling method, in which an a priori shape of the CME geometry is assumed, the mask fitting has more flexibility. The polarisation ratio method makes use of the Thomson scattering geometry. We find that spatially the 3D CME derived from the mask fitting lies mostly in the overlap region obtained with the polarisation method using data from STEREO. In addition, the mask fitting can help resolve the front/back ambiguity inherent in the polarisation ratio method. However, the local correlation tracking plus triangulation did not show consistent results with the other four methods. This method performed poorly, primarily because the two STEREO spacecraft had a large angular separation. Under these circumstances, it is difficult to identify points taken from independent images that correspond to the same physical feature. Excluding the local correlation tracking method, the latitude of the CME's centre of gravity derived from the other methods deviates within 1^∘, and the longitude differs within 19^∘.