A THREE-DIMENSIONAL ORBIT FOR CAPELLA

BRANHAM, RICHARD L. JR.

ASTRONOMICAL JOURNAL

Año: 2008 vol. 136 p. 963 - 973

0004-6256

Semidefinite programming is applied to 169 interferometric observations of Capella, made between 1919 and 1999, and 221 double-line radial velocities, obtained between 1896 and 1991, to calculate a three-dimensional orbit. The data are reduced with the robust L1 criterion. The orbit is nearly circular, eccentricity of 0.00508, with a semimajor axis of 0.056 and period of 104.039 days. The mass of the primary is calculated to be 3.049M, that of the secondary 2.569M, and the parallax of the system is calculated to be 74.85 mas. Another orbit is calculated, but using only the best data, Mark III interferometric observations, and Coralie radial velocities. Although the mean errors for this orbit are considerably smaller, reasons are given for preferring the orbit calculated from all of the data as opposed to only the best data: the residuals are more random, the parallax agrees better with van Leeuwen’s re-reduction of the Hipparcos parallax, and the Shannon uncertainty is lower.L1 criterion. The orbit is nearly circular, eccentricity of 0.00508, with a semimajor axis of 0.056 and period of 104.039 days. The mass of the primary is calculated to be 3.049M, that of the secondary 2.569M, and the parallax of the system is calculated to be 74.85 mas. Another orbit is calculated, but using only the best data, Mark III interferometric observations, and Coralie radial velocities. Although the mean errors for this orbit are considerably smaller, reasons are given for preferring the orbit calculated from all of the data as opposed to only the best data: the residuals are more random, the parallax agrees better with van Leeuwen’s re-reduction of the Hipparcos parallax, and the Shannon uncertainty is lower..056 and period of 104.039 days. The mass of the primary is calculated to be 3.049M, that of the secondary 2.569M, and the parallax of the system is calculated to be 74.85 mas. Another orbit is calculated, but using only the best data, Mark III interferometric observations, and Coralie radial velocities. Although the mean errors for this orbit are considerably smaller, reasons are given for preferring the orbit calculated from all of the data as opposed to only the best data: the residuals are more random, the parallax agrees better with van Leeuwen’s re-reduction of the Hipparcos parallax, and the Shannon uncertainty is lower.M, and the parallax of the system is calculated to be 74.85 mas. Another orbit is calculated, but using only the best data, Mark III interferometric observations, and Coralie radial velocities. Although the mean errors for this orbit are considerably smaller, reasons are given for preferring the orbit calculated from all of the data as opposed to only the best data: the residuals are more random, the parallax agrees better with van Leeuwen’s re-reduction of the Hipparcos parallax, and the Shannon uncertainty is lower.Hipparcos parallax, and the Shannon uncertainty is lower.