|
Abstract:
|
Determining distances is one of the most challenging problems in astronomy due to the degeneracy between apparent brightness, intrinsic luminosity, and distance. Several classes of variable stars break this degeneracy because their period of variation is proportional to their luminosity. Using variable stars as distance indicators requires fast and accurate period estimation algorithms. This is especially challenging for sparsely-sampled light curves. We develop a semi-parametric model for estimating periods of Miras, a class of periodic variable stars that display additional aperiodic variations. The model uses a sinusoidal basis for periodic variation and a gaussian process for aperiodic variation. We use the maximum likelihood to estimate the period and the parameters of the Gaussian process, while integrate out the effects of other nuisance parameters in the model with respect to a suitable prior distribution obtained from earlier studies. Since the likelihood is highly multimodal for period, we implement a hybrid method that applies the quasi-Newton algorithm and grid search. A large-scale, high-fidelity simulation is conducted to mimic the sampling quality of M33SSS.
|
