Flares and coronal mass ejections (CMEs) are important for the evolution of the atmospheres of planets and their potential habitability, particularly for planets orbiting M stars at a distance < 0.4 AU. Detections of CMEs on these stars have been sparse, and previous studies have therefore modelled their occurrence frequency by scaling up solar relations. However, because the topology and strength of the magnetic fields on M stars is different from that of the Sun, it is not obvious that this approach works well. We used a large number of high-resolution spectra to study flares, CMEs, and their dynamics of the active M dwarf star AD Leo. The results can then be used as reference for other M dwarfs. We obtained more than 2000 high-resolution spectra (R~35000) of the highly active M dwarf AD Leo, which is viewed nearly pole on. Using these data, we studied the behaviour of the spectral lines Halpha , Hbeta , and He I 5876 in detail and investigated asymmetric features that might be Doppler signatures of CMEs. We detected numerous flares. The largest flare emitted 8.32 x 10 31 erg in Hbeta and 2.12 x 10 32 erg in Halpha . Although the spectral lines in this and other events showed a significant blue asymmetry, the velocities associated with it are far below the escape velocity. Although AD Leo shows a high level of flare activity, the number of CMEs is relatively low. It is thus not appropriate to use the same flare-to-CME relation for M dwarfs as for the Sun.

Using the Kepler K2 data, we studied the flare-activity of young K-and M-stars in the Upper Sco region and found that they have 10000 to 80000 times as many super-flares with E> 5E34 erg than solar like stars. The power-law index for flares is $N/dE~E^-1.2 for K stars, N/dE~E^-1.4 for early M-star and dN/dE~E^-1.3 for late M stars, which is about the same as that of the Sun. We also observed the active M-star AD Leo spectroscopy for 222 hours, detected 22 flares but no coronal mass ejections.