http://arxiv.org/abs/1705.08537
The simultaneous photometric and spectroscopic observations of the RR Lyrae variables in the globular cluster, M3, published in Jurcsik et al. (2017, Paper I) made it possible to perform Baade-Wesselink (BW) analysis of a large sample of Blazhko stars for the first time. The BW distances of Blazhko stars turned out to be unreliable, as significantly different distances were obtained for the stars of the Blazhko sample and also for the same star in different modulation phases. Even the results of small modulation-amplitude Blazhko stars may be doubtful. This result warns that the application of the BW method to Blazhko stars is not trustworthy.
Keeping the distance fixed for each Blazhko star in each modulation phase, a significant difference between the spectroscopic and the photometric radius ($R_{\textrm{sp}}$, $R_{\textrm{ph}}$) variations is detected. The phase and amplitude variations of $R_{\textrm{sp}}$ follow the changes of the light curve during the Blazhko cycle but the $R_{\textrm{ph}}$ curve seems to be not (or only marginally) affected by the modulation. The asynchronous behaviour of $R_{\textrm{sp}}$ and $R_{\textrm{ph}}$ supports the interpretation of the Blazhko effect as a depth-dependent phenomenon, as the spectroscopic radius variation reflects the radial displacement of the line-forming region high in the atmosphere, while the photospheric radius variation is derived from the information of the observed visual-band light emitted mostly by the lower photosphere. The stability of $R_{\textrm{ph}}$ may be interpreted as a strong argument against the non-radial-mode explanation of the Blazhko phenomenon.
J. Jurcsik and G. Hajdu
Thu, 25 May 17
20/44
Comments: 9 pages, 6 figures, accepted for publication in the Monthly Notices of the Royal Astronomical Society