The measured Velocities for Upsilon Andreomdae are fit with a model containing three companions. The inner planet is subtracted (4.617-day sine wave) to render more visible the wobble caused by the outer two planets.

The inner planet was announced by Butler et al. in Jan 1997. A velocity plot containing a few months of data looks clean ( Fig 1 ), during a few months. However, a larger duration of velocities reveals additional velocity variations. Recognizing that the velocity variations could be due to a second companion, Butler and Marcy alerted the astronomical community to this possibility.

The second companion appeared to have an orbital period of about 3.5 years from the long-term excursions of the velocities. These velocity variations were also found by the AFOE team of Drs. R.Noyes, T.Brown, S.Korzennik, S.Horner, A.Contos, P Nisenson, et al.

A Keplerian orbital model was constructed to fit the data. However, the best solution still left velocity points scattering above and below the 2-planet theoretical curve. After subtracting the theoretical velocities that would be produced by a 2-planet system from the observed velocities, a periodic pattern was observed in the "residual" velocities, suggesting the possibility of one more companion. A 3-planet model was found to produce a dramtically superior fit to the data, reducing the residual velocities to the expected variations, with no periodicity in the residuals.

With the orbital solution for 3 planets that fit our observed velocities, we can subtract the predicted velocities for any one planet, to reveal the velocities caused by the remaining two planets:

PLANET B

PLANET C

PLANET D

 

 

Images of the Star Field around Upsilon Andromedae.

Simulated view of Planets around Upsilon Andromedae by Sylvain Korzennik.
 

Return to planet search home page...