Determining CME parameters by fitting heliospheric observations: Numerical investigation of the accuracy of the methods

Transients in the heliosphere, including coronal mass ejections (CMEs) and corotating interaction regions can be imaged to large heliocentric distances by heliospheric imagers (HIs), such as the HIs onboard STEREO and SMEI onboard Coriolis. These observations can be analyzed using different techniques to derive the CME speed and direction. In this paper, we use a three-dimensional (3-D) magneto-hydrodynamic (MHD) numerical simulation to investigate one of these methods, the fitting method of  and . Because we use a 3-D simulation, we can determine with great accuracy the CME initial speed, its speed at 1 AU and its average transit speed as well as its size and direction of propagation. We are able to compare the results of the fitting method with the values from the simulation for different viewing angles between the CME direction of propagation and the Sun-spacecraft line. We focus on one simulation of a wide (120–140°) CME, whose initial speed is about 800 km s⁻¹. For this case, we find that the best-fit speed is in good agreement with the speed of the CME at 1 AU, and this, independently of the viewing angle. The fitted direction of propagation is not in good agreement with the viewing angle in the simulation, although smaller viewing angles result in smaller fitted directions. This is due to the extremely wide nature of the ejection. A new fitting method, proposed to take into account the CME width, results in better agreement between fitted and actual directions for directions close to the Sun–Earth line. For other directions, it gives results comparable to the fitting method of Sheeley et al. (1999). The CME deceleration has only a small effect on the fitted direction, resulting in fitted values about 1–4° higher than the actual values.