3 The sample

Table 1 presents the parameters of EIT waves, type II bursts and related flares. In the second column a start time of the events is given which demands for some comments. Due to the "rough'' EIT image cadence (one image in about 15 minutes) the wave front is usually visible only at some distance from the flaring active region. It is difficult to give an exact estimate of the ignition site and time of the disturbance in EIT frames. In Table 1, "start time'' means the time interval between two subsequent images within which an EIT wave appears. Further, there is a source of instrumental error involved in the recording of the EIT image time. This introduces an additional uncertainty of several minutes into image timing. Figure 2 gives the relation between the EIT image times and the flare onset reference "associated type III bursts''. The black bars denote the start time window of the EIT waves.

The decay rate of EIT waves is difficult to define. The waves weaken while propagating. Often, the time interval between subsequent images can be too large to obtain subsequent images before the wave fades from observability. The EIT wave speed (third column) is estimated relative to the solar surface by measuring successive positions of the wave front on different images. When three consecutive images contain an EIT wave transient, two speeds are listed in the catalogue corresponding to two pairs of images.

Figure 10: Event on 25 May 1997

Negative type II drift rates (D_f) mean shock propagation toward lower plasma densities. In brackets, the frequency f is given where D_f has been measured. The type II shock speed V is estimated using the onefold Newkirk model V=2H(D_f/f) (Newkirk 1961). H is the coronal density scale height radially slightly depending on the frequency f.