Previous   |   Session 68   |   Next  |   Author Index   |   Block Schedule

[68.22] A Comparison of Least-Squares and Bayesian Techniques in Fitting the Orbits of Extrasolar Planets

P. Driscoll, D Fischer (San Francisco State University)

The orbit of an extrasolar planet is obtained by fitting a Keplerian orbit to radial velocity, or astrometric data. We present a comparison of Levenberg-Marquardt Monte Carlo (LMMC) residual fitting to the Bayesian method of Markov Chain Monte Carlo (MCMC). The most challenging aspect of fitting orbits is quantifying the uncertainty in the orbital parameters. We investigate the ability of the two techniques to quantify the orbital parameters and their uncertainties for tens of synthetic radial velocity data sets with varying planetary periods, observation phase coverage, and time of first observation. We find that the LMMC best-fits correspond in general to the peaks of the MCMC posterior probability distributions for the orbital parameters. LMMC tends to under estimate the uncertainty in the orbital parameters, especially for data sets with large gaps, while the MCMC provides a more complete representation of the parameter posterior distributions. We acknowledge the NASA Grant NNG056G164G.

The author(s) of this abstract have provided an email address for comments about the abstract: driscoll@dickinson.edu

Previous   |   Session 68   |   Next

Bulletin of the American Astronomical Society, 37 #4
© 2005. The American Astronomical Soceity.