Gliese 667C Artist view
The findings of the University of Hertfordshire team are published on line.
The team started with data from three planet-hunting missions: Harps, AAPS, and HiRes, all of which had data on Tau Ceti.
Now, Hugh Jones of the University of Hertfordshire and colleagues have refined their "noise modelling" in order to subtract it, and thereby see the smallest signals hiding in the data - starting with Tau Ceti.
The trick to honing the technique was to put in "fake planets" - to add signals into the messy data that planets should add - and find ways to reduce the noise until the fake planets became more and more visible in the data.
"Putting all that together, we optimised a noise-modelling strategy which allows us to recover our fake signals - but in the process of doing that, we actually saw that we were finding signals as well," Prof Jones said - actual planets.
BBC News Science and Environment 18.12.2012