Pulsating white dwarfs

In 1968 the American astronomer Arlo Udell Landolt serendipitously detected rapid luminosity variations in the DA white dwarf HL Tau 76 (Landolt, 1968). Even though it was not initially recognized as such, this was the first pulsating white dwarf discovered. Since then the study of these objects has made great progress both

observationally, with over 200 pulsating white dwarfs found, and theoretically with accurate models to describe and exploit the observed pulsations. Today we known that pulsations are cause by non-radial gravity waves and arise as white dwarfs cool through narrow temperature ranges which define several instability strip (e.g Bergeron et al. 2004; Winget & Kepler 2008; Fontaine & Brassard 2008). Therefore it is thought that all white dwarfs will experience similar conditions as they cool through the instability strips, i.e. the appearance of pulsations is a natural phase in their evolution (Dziembowski & Koester, 1981; Winget et al., 1981). Pulsating white dwarfs are divided into three classes.

• DAVs or ZZ Ceti Stars. Named after the second pulsating white dwarf dis- covered (Lasker & Hesser, 1971), these are pulsating white dwarfs with H- dominated atmospheres. ZZ Cetis are by far the most common type of pul- sating white dwarfs. As DA white dwarfs cool through the temperature range 12,500₋11,000 K, partial recombination of H in their atmospheres causes a dramatic increase in the envelope opacity which, in turn reduces the flow of radiation causing the pulsations (Fontaine & Brassard, 2008). Recently Her- mes et al. (2013) discovered that extremely low-mass (ELM), He-core white dwarfs (M <0.25M⊙) also undergo pulsations. These white dwarfs also have

H-dominated atmospheres, but are often considered a separate class from ZZ Cetis. The pulsation periods of ZZ Cetis typically range from 100 to 1400s (Mukadam et al., 2004) and can even reach up to 1.7h in ELM white dwarfs.

• DBV or V777 Her. The existence of pulsating white dwarfs with He-dominated atmospheres was predicted by Winget et al. (1981), before the actual discovery of one a year later (Winget et al., 1982). Similarly to the driving mechanism in ZZ Cetis, pulsations in DBV are caused by partial recombination of the outer envelope. Because of the higher ionisation potential of He, this happens at a higher temperature than in ZZ Cetis: pulsations in DBs arise in the temperature range 29,000 ₋ 21,000K (Nitta et al., 2009). However, with only _≃20 known to date (Kilkenny et al., 2009; Nitta et al., 2009), a robust definition of a DBV instability strip is still an ongoing challenge.

• PG 1159 or GW Vir. These extremely hot (75,000K. Teff . 200,000K)

pre-white dwarfs have atmospheres composed of helium, carbon and oxygen in roughly comparable amounts. This peculiar atmosphere composition is thought to be the result of violent mixing events caused by He flashes in late stages of the progenitor evolution (Werner & Herwig, 2006). The origin of pulsations in PG 1159 is the so called κ-mechanism, a process caused by the

cyclic ionization of the K-shell electrons of carbon and oxygen (Quirion et al., 2007).

• Another class of variable white dwarfs are the recently discoveredhotDQ stars (Dufour et al. 2007a, Sect. 1.2.2). As mentioned in the introduction, these are hot (Teff ≈18,000−24,000K) white dwarfs with C dominated atmospheres.

Some studies speculated thathotDQs may constitute a new class of pulsators, but the origin of the observed variability is still matter of debate (e.g Fontaine et al. 2008; Lawrie et al. 2013).

Traditionally, the physical parameters of white dwarfs, including their effec- tive temperature (Teff) and surface gravity (logg) are determined from spectroscopic

analysis (Sect 1.2.2, Fig 1.5, Bergeron et al. 1992b). However, spectral information is restricted to the outermost layers of of the star. As a consequence, our understand- ing of white dwarfs is often, literally, superficial. Pulsating white dwarfs, however, provide a unique opportunity to look past the outermost atmosphere and probe the interior of these objects. Asteroseismological tools can be used to study pulsating white dwarfs and investigate the structure, composition and mass of both the core and envelope (Winget & Kepler 2008; Fontaine & Brassard 2008; Althaus et al. 2010). Furthermore, asteroseismology can be used to probe internal rotation pro- files (Charpinet et al., 2009), measure weak magnetic fields (Winget et al., 1991) and even search for planetary companions via pulse timing variations (Mullally et al., 2008).

Historically, this unambiguous identification of pulsating white dwarfs re- quired several hours of continuous high cadence photometry (e.g. Mukadam et al. 2004; Nitta et al. 2009), which is observationally expensive. Candidate selection has so far relied on colours and/orTeff and logg, estimated from model fits to spectra,

with efficiencies ranging from 30% to 80% (e.g. Mukadam et al. 2004).

In recent years, the opportunity to repeatedly survey large areas of the sky has rapidly advanced the field of time-domain astronomy. Time-domain exploration of the sky is at the forefront of modern astronomy with many wide-field surveys in operation, or soon to come on-line (e.g. CRTS, Drake 2014; PTF, Law et al. 2009; EVRYSCOPE, Law et al. 2015; Pan-STARRS, Morgan et al. 2014; Gaia, Walton 2014; LSST, Ivezic et al. 2011). In order to fully exploit these vast resources we will need to develop efficient a robust method to identify pulsating white dwarf candi- dates that reduces, or removes, the need for high-cadence identification photometry. Here, we investigate the feasibility of using multi-epoch photometry from large-area surveys to reliably identify pulsating white dwarf candidates based on

Stripe 82 of the Sloan Digital Sky Survey (SDSS). Several successful studies have made use of Stripe 82 multi-epoch observations to search for variable objects (eg. Sesar et al. 2007; Bramich et al. 2008; Becker et al. 2011). However these studies mainly focused on identifying large-amplitude variable sources (e.g eclipsing binaries or flaring stars) and the potential of identifying low amplitude variability (like that of pulsating white dwarfs) has not yet been explored. Starting from a sample of 400 high-confidence white dwarfs candidates from the catalogue of Gentile Fusillo et al. (2015), we recover, recalibrate and quality-control all available multi-epoch pho- tometry to identify variable candidates. Even though Stripe 82 offers only low and irregular cadence over a relatively limited area, our study demonstrates promising results. In the near future a similar methodology, applied to superior time-domain surveys (e.g. Pan-STARSS, Gaia and LSST) will completely change the way we identify variable stars, including pulsating white dwarfs.