...$R=r/r_{\odot}$

The R(t) data utilized in the following analysis are taken from the LASCO CME Catalog compiled by Seiji Yashiro and Grzegorz Michalek under the guidance of Nat Gopalswamy: http://cdaw.gsfc.nasa.gov/CME_list/

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...)data

The distance R=3 corresponds to the average radial distance of the first measurement in the LASCO-C2 field-of-view (for details see Vrsnak et al. 2004b).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... angle

$\psi$ is measured from the north pole in the anti-clockwise direction.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... available

ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/SOLAR_FLARES/XRAY_FLARES/

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... used

ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/SOLAR_FILAMENTS/

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... CMEs

The five eliminated events had the "initial'' velocities v₃= 1300, 1110, 1100, 1010, and 780 km s^-1, so apparently the probability of an erroneous association decreases with the CME speed.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... temperatures

Note that often very weak flare-like brigthenings are observed in the chromospheric or EUV spectral line images during DSF-CMEs. Sometimes such signatures are also seen in the soft X-ray images, but the associated flux increase is too weak to be recognized in the GOES (whole sun) flux measurements, especially in the periods of enhanced background flux. So, one may argue that the two-class CME concept does not make sense from this point of view too, since strictly speaking, there are no real non-flare CMEs.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...2003)

Note that in the model by Chen (1989), as well as in later modificatins (e.g., Chen & Krall 2003) the supply of the poloidal flux is attributed to the emerging flux process and not to the reconnection. On the other hand, Vrsnak (1990) considered the effect of reconnection, but unfortunately, only qualitatively.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... height

Note that most of CMEs start to accelerate when their height is comparable to the footpoint half-separation (Vrsnak et al. 1991; Chen & Krall 2003).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.