Improved period estimates of RR Lyrae stars using multitapering and the F-test

Obtaining period estimates of RR Lyrae variables often requires a multi-step process (e.g., Lomb-Scargle periodogram followed by refinement steps to converge on the best estimate of the period). We have discovered a new technique to acquire accurate pulsation frequencies for RR Lyrae stars in a more efficient and reliable way. The method uses mtNUFFT, an extension of the Thomson multitaper power spectrum estimator to quasi-even time-sampling, along with the F variance ratio test, to estimate the periods. We introduce this approach and demonstrate its effectiveness on a sample of 44 RR Lyrae stars from TESS. Overall, the F variance ratio test outperforms the Lomb-Scargle periodogram in both accuracy and computational efficiency. In particular, we see obvious improvements on period estimates in the phase-folded light curves of the RR Lyrae stars compared to literature values. The software for our new approach is available in an open-source software package, which should be useful not only for RR Lyrae variables, but also for other studies of stellar pulsations.