Free Astronomy Magazine September-October 2014

25 EXOPLANETS lar activity and, what's more, planet “e” does no longer require an eccentric orbit, which was instead a common assumption in previous works. The “loss” of two potentially habitable plan- ets has however positive implications. First, it places particular emphasis on the neces- sity of an in-depth knowledge of the sur- face activity of a star in order to under- stand how many planets it actually hosts (if they are looked for with the radial-velocity method); and second, it further strength- ens the suspicion that on red dwarfs there may be active regions considerably differ- ent from those of the solar- type. Finally, the technique of subtracting the contribu- tion given by the stellar activ- ity to the variation of the ra- dial velocity (not always fea- sible) can help detect the sig- nals from previously unseen small-sized planets. T he planetary system of GJ 581 after the reordering by Robertson’s team. In the foreground is planet “c”, in the distance “b” and the smaller “e”. The video be- low emphasizes the discovery of GJ 581d, the plan- et on which re- searchers had high hopes of an astrobiological nature, but that turned out to be non-existent. [ESO/PHR/UPR Arecibo] to the point of becoming negligible, while the signal from planet "g" disappeared completely. The best proof that the procedures follow- ed by Robertson's team are those that best represent reality lays in the fact that the signals attributable to the actually existing planets, GJ 581e, GJ 581b and GJ 581c (the latter with a period of 12.91 days), emerge markedly stronger after correction for stel- n