When the first “hot Jupiter” planets were found they were a big surprise — no-one had expected to find massive Jupiter-sized planets very close to stars, in orbits of only a few days. Most planet-formation theory says that they can’t have formed there, and must have formed much further out, beyond the “snow line” where it is much colder.

Much investigation has gone into discovering what moves the planets inwards to become hot Jupiters. One favourite explanation is the long-term effect of gravitational perturbations to the planet’s orbit, caused by another massive planet or low-mass companion star much further out.

If this is right we should be able to find these outer companions, and one method is to monitor the radial-velocity motion of the host star, looking for the gravitational pull caused by the outer companion. Hence one would expect the stars’ radial velocity to show a short-term cycle with the period of hot Jupiter, plus a much longer term trend.

An important paper just announced by Heather Knutson and colleagues announces the results of monitoring 51 hot-Jupiter systems — including 18 WASP planets — using the HIRES spectrograph on the 10-m Keck telescopes on top of Mauna Kea in Hawaii. They confirm long-term radial-velocity trends previously suspected in 9 systems and report newly found trends in 7 other systems.

Four WASP systems (WASP-8, WASP-10, WASP-22 and WASP-34) are found to have radial-velocity trends indicating a massive outer companion. The plot has the radial-velocity on the y-axis (units of metres per sec) plotted against time (years since 2000).

In WASP-8 and WASP-34 the orbit of the companion is beginning to be constrained, while for WASP-10 and WASP-22 the timescale of the orbit appears to be longer. Further monitoring of these systems and other hot Jupiters (the plot also shows planets from the HAT and XO projects) might help to answer the question of whether these outer companions are the cause of hot Jupiters.