Asteroseismology of the Pulsating sdB star and Suspected SNe Ia Progenitor KPD 1930+2752: A Whole Earth Telescope Campaign

Abstract

We propose to observe KPD 1930+2752, a pulsating subdwarf B star with a particularly rich period spectrum that is also a short period (~ 2h17m) close binary system (as revealed by the presence of an ellipsoidal variation in the lightcurve). The companion is almost certainly a white dwarf as there are no indications of eclipse or reflection effect. Radial velocity measurements indicate that the total mass of the system exceeds the Chandrasekhar limit, making KPD 1930+2752 the first known (non-X-ray source) candidate for a Type Ia supernova progenitor. The pair should merge within an astrophysically interesting timescale (~ 200 Myr) due to gravitational wave radiation. Additionally, the tidal distortions are indicative of the strong tidal force incurred on the pulsating star by its companion. This will allow us to investigate the effects such strong tidal forces have on pulsations. Due to its outstanding astrophysical interest, this object is the primary target of the WET network Xcov 23 campaign (which includes the present proposal). The goal is to fully resolve the complex pulsation pattern of the pulsating sdB star, as well as the details of the ellipsoidal modulation. The asteroseismological analysis of the pulsation data, along with a detailed analysis of the folded, high S/N ratio ellipsoidal variation will lead to strong physical constraints on this particularly interesting and, so far, unique stellar object.

Justification

KPD 1930+2752 is, by all accounts, a rather extraordinary object. Multiperiodic luminosity variations in the lightcurve of this apparently unassuming V=13.8 hot B subdwarf (sdB) star have been reported by Billeres et al. (2000). Figure 1 shows a portion of the lightcurve which revealed the presence of at least 44 short-period modes (significantly more than is usually seen in the pulsating sdB -- or EC14026 -- stars) with periods and amplitudes respectively in the ranges 145--332 s and 0.06--0.45\% of the mean brightness of the star (Fig. 2). However, the light curve is dominated (~ 1.39% of the mean brightness) by a nearly sinusoidal modulation with a much longer period of ~ 4108.9s. The latter variation has no equivalent among the other known EC14026 stars and is interpreted as the ellipsoidal deformation (see Fig. 3) of the sdB star in a close, but detached binary system containing a faint, unseen companion (most likely a white dwarf). The analysis of Maxted et al. 2000 confirmed that KPD 1930+2752 is, indeed, a binary system by measuring, from the H-alpha and He I-6678A spectral lines, radial velocity variations with the same period (~ 2h17m) as the orbital period derived from the ellipsoidal deformation (ie., twice the observed modulation period). In addition, they found from the kinematics (the projected orbital velocity is K=349.3 +/- 2.7 km/s) that the unseen companion is almost certainly a white dwarf (as suspected by Billeres et al. 2000) and that the total mass of the binary system (at least 1.47 +/- 0.01 M_sun exceeds the Chandrasekhar limit (just under 1.4 M_sun for a C/O core white dwarf). Since the binary system should merge within only ~200 Myr due to gravitational wave radiation, KPD 1930+2752 becomes the first good candidate of a Type Ia supernova (SNe Ia) progenitor that can be associated with the double-degenerate SNe Ia model. Type Ia supernovae are of utmost importance as they can be seen out to enormous distances and are used to derive cosmological parameters (e.g., the cosmological constant; Perlmutter et al. 1999). However, the physical origin of these detonations is still unclear, which leads to uncertainties on the calibration of SNe Ia light curves. Consequently, as the first detected SNe Ia progenitor in which the visible component is a pulsating sdB star with a particularly rich pulsation period spectrum, KPD 1930+2752 becomes an object of outstanding astrophysical interest. Pulsation computations show (see Fig. 4) that the shorter periods are consistent with a theoretical, low radial order, low degree p-mode period spectrum which is rotationally split with a period of 8217.8 s, the value expected for a star tidally locked to its companion as would be the case for a close binary system. This assumption seems confirmed by a recent measurement of vsin i=90.2 +/- 4.2 km/s from high spectral resolution VLT data (Heber 2002). In addition, when tidal forces dominate the Coriolis (or centrifugal) force, as is the case in KPD 1930+2752, it has been suggested that the pulsation axis may align with the companion (in a way similar to roAp stars where the pulsation axis is aligned with the magnetic axis). If true, the pulsation axis should precess with the orbit and each pulsation period should create multiple peaks in the Fourier domain. These secondary peaks, along with phase cues, provide signatures of a tipped pulsation axis and, in principle, can also be used to uniquely identify the pulsation modes (Reed et al. 2002; see http://sdbv.smsu.edu/mreed/IAU185/).

In order to fully exploit the extraordinary circumstances that KPD 1930+2752 is both an ellipsoidal variable and a rich p-mode pulsator, this star has been chosen as the primary target of the Whole Earth Telescope (WET; Nather et al. 1990) Xcov 23 campaign. Its complex temporal spectrum requires continuous coverage (to eliminate daily aliases) and a long time baseline (2 weeks in the present case) to 1) fully and unambiguously resolve the pulsations, 2) search for a tidally induced tipped pulsation axis using the method outlined in Reed et al. (2002), and 3) build a folded ellipsoidal lightcurve with a significantly improved S/N ratio. After many years of efforts, we have developed the capability to model successfully the nonadiabatic pulsation properties of EC14026 stars (see Charpinet et al. 2001, for a review). The rich, fully resolved pulsation period spectrum, along with constraints on mode identification brought by an eventual tipped pulsation axis, will allow us to perform a very detailed asteroseismological analysis of the pulsating sdB component, similar to the recent success obtained for the sdB pulsators PG 0014+067 (Brassard et al. 2001) and PG 1047+003 (Charpinet et al. 2002). In parallel, the quantitative analysis of the high S/N ratio folded ellipsoidal modulation will provide crucial constraints on the KPD 1930+2752 system that are completely independant of the sdB pulsation analysis. Put together, these studies promise to be extremely valuable for inferring the global structural parameters of the sdB star as well as those of the complete sdB+WD system. Since KPD 1930+2752 is the first known object that might lead to a SNe Ia detonation within a relatively short astrophysical timescale, the informations obtained on this object should have wide ranging scientific implications.

References

Figures

Figure 2: Upper half: Fourier amplitude spectrum of the complete light curve of KPD 1930+2752. The vertical scale has been multiplied by a factor of 3 for frequencies larger than 2 mHz in order to visualize better the numerous higher frequency components (due to pulsations) that have substantially smaller amplitudes. The dominant peak at low frequency corresponds to the ellipsoidal variation (see text). Lower half: Fourier amplitude spectrum (plotted upside down) of the residual light curve after substraction of the identified periods.}

Figure 3: Light curve of KPD 1930+2752 prewithened of the short-period variations and folded on the period of 8217.76 s (using 100 frequency bins). As usual, the curve is plotted twice for better visualization. This curve shows the ellipsoidal modulation of the mean brightness of the sdB star. The lowest curve is the residual after substraction of a fitted model of the ellipsoidal modulation (shown as the plain line).

Figure 4: Comparison of the observed period spectrum of KPD 1930+2752 with a theoretical period spectrum that has been rotationally split through the effect of first-order uniform rotation on a representative stellar model rotating with a period of 8218 s. Note that no attempt has been made to perform a detailed fit because of the difficulty to extract the exact periods from our single site data.}

Download the KPD 1930+2752 justification

Back to the XCov 23 page.