Proposal #59

Simultaneous ground- and space-based photometric observations of the eclipsing massive binary HD 165246
-------------------------------------------------------------------------------------------------------------------------------------------------

The most massive stars, having spectral-type O, evolve extremely quickly, and their lifetime and evolution is dominated by their internal properties, such as the internal mixing of chemical species. These processes are ill constrained, causing significant uncertainties in various fields of modern astrophysics. To study this mixing, we resort to asteroseismology, since the stellar pulsations contain information from the otherwise hidden, deep interior (see Aerts et al. 2010, Springer, for a monograph on asteroseismology).

To this end, we have proposed 12 O-type stars to be observed during Campaign 9 of the K2 mission, with the Kepler satellite, encouraged by our successful proof-of-concept study of 5 O-type stars from K2 Campaign 0 data (Buysschaert et al. 2015, MNRAS, 453, 89). All targets have been accepted, and will be observed from April 07 until July 01, 2016. Among these 12 targets, is HD 165246, a well characterised binary system. Such a target offers the unique opportunity to combine asteroseismology and binary modelling. Moreover, the photometric eclipses permit eclipse mapping of the pulsations, where stellar pulsations are mapped to particular regions of the stellar surface during the ingress and egress of the eclipses (Reed et al. 2005, ApJ, 634, 602). This would be the very first application of this method.

A recent study, using NACO/SAM imaging, showed that HD 165246 consists of at least three components (Sana et al. 2014, ApJS, 215, 15). The two innermost components, consisting of an O8V and B7V star, are separated by 30 ± 15 mas, while the B-component is located 1.93 ± 0.04" further away. Mayer et al. 2013 (A&A, 550, A2) studied HD 165246A in detail, constraining many of the orbital parameters and the following linear ephemeris for the primary eclipse:
T_{min, eclipse} = HJD 2,452,383.8927(45) + 4.592706(14)d x E.
The fast rotation velocity of the primary component (vsini = 242.6 ± 2.7 km/s) leafs to an approximate 3:1 spin-orbit resonance.

The long timebase of the high-precision K2 photometry is ideal to study stellar pulsations, yet lacks much needed color information. This information is vital to study the geometry of the stellar pulsations. Indeed, differences in the spherical degree, l, of the modes, produce different amplitude ratios for a given frequency, when comparing light from different filterbands (e.g. Dupret et al. 2003, A&A, 398, 77). Combining this with eclipse mapping, should constrain most of the geometry of the stellar pulsations.

With this proposal, we request high-cadence, multi-color monitoring of HD 165246 from May 18 until May 26, simultaneously with the K2 observations. These ground-based observations would happen together with those from La Palma (Spain), with MAIA mounted on the 1.2-m Mercator telescope (Raskin et al. 2013, A&A, 559, A26). As such, we will obtain a high duty-cycle, thanks to this multi-site campaign. Since HD 165246 is located near the Galactic Plane, and because of the large field-of-view of all AAVSOnet telescopes, suitable reference stars are fairly easily obtained. Because MAIA observes simultaneously in the u-, g-, and r-filters, we wish to perform the AAVSOnet observations with the same filterbands. If necessary, however, we can switch the Sloan u-filter to a Johnson B-filter. In addition, it would be good to obtain a few multi-color photometric measurements of eclipses prior or after the monitoring ground-based campaign and during the K2 campaign. These observations would help with the possible eclipse mapping, since there is a 3:1 spin-orbit resonance, obscuring different regions of the stellar surface during consecutive eclipses.

Custom apertures will be defined for both HD 165246 and the reference stars. These apertures will enable us to work in relative brightness units for HD 165246. In a second step, we we will perform a detailed frequency analysis for each final lightcurve in each filter and the simultaneously taken K2 observations, by means of an iterative prewhitening procedure. Using the initial solution from Mayer et al. and the new K2 photometry, we will further constrain the parameter space using PHOEBE (PHysics Of Eclipsing Binaries; Prsa & Zwitter 2005, ApJ, 628, 426). Using the updated PHOEBE 2.0 routine, it is also possible to simulate the effects of the binary and the pulsations together in the lightcurve. The amplitude ratios of the pulsations, together with all additional seismic information, is then compared to stellar models computed with MESA (Paxton et al. 2011, ApJS, 192, 3) coupled to the pulsation code GYRE (Townsend & Teitler 2013, MNRAS, 435, 3406) to calculate seismic models. It is possible to take the binary evolution into account during the MESA modelling, as well as the various mechanisms of internal mixing with different efficiencies. Therefore, defining the internal properties of the massive O-star HD 165246Aa.