The Extent and Cause of the Pre-White Dwarf Instability Strip

One of the least understood aspects of white dwarf evolution is the process by which they are formed. The initial stages of white dwarf evolution are characterized by high luminosity, high effective temperature, and high surface gravity, making it difficult to constrain their properties through traditional spectroscopic observations. We are aided, however, by the fact that many H- and He-deficient pre-white dwarfs (PWDs) are multiperiodic g-mode pulsators. These stars fall into two classes: the variable planetary nebula nuclei (PNNV) and the "naked" GW Vir stars. Pulsations in PWDs provide a unique opportunity to probe their interiors, which are otherwise inaccesible to direct observation. Until now, however, the nature of the pulsation mechanism, the precise boundaries of the instability strip, and the mass distribution of the PWDs were complete mysteries. These problems must be addressed before we can apply knowledge of pulsating PWDs to improve understanding of white dwarf formation. This paper lays the groundwork for future theoretical investigations of these stars. In recent years, Whole Earth Telescope observations led to determination of mass and luminosity for the majority of the GW Vir pulsators. With these observations, we identify the common properties and trends PWDs exhibit as a class. We find that pulsators of low mass have higher luminosity, suggesting the range of instability is highly mass dependent. The observed trend of decreasing periods with decreasing luminosity matches a decrease in the maximum (standing-wave) g-mode period across the instability strip. We show that the red edge can be caused by the lengthening of the driving timescale beyond the maximum sustainable period. This result is general for ionization-based driving mechanisms, and it explains the mass dependence of the red edge. The observed form of the mass dependence provides a vital starting point for future theoretical investigations of the driving mechanism. We also show that the blue edge probably remains undetected because of selection effects arising from rapid evolution.