This work was supported by the National Science Foundation AST-1109394, PI R. Humphreys

Luminous Stars


Introduction to the Program

Luminous Blue Variables, Hypergiants, and Supernova Impostors

In 1965 Fritz Zwicky proposed his "Type V" class of supernovae, with relatively faint maxima but long durations. He cited only SN 1961v and eta Carinae, but more examples have since been recognized. Today we know that at least some (maybe all) of them were giant eruptions whose stars survived, not true supernovae. These are the most energetic non-terminal stellar explosions known, sometimes ejecting several M of material. Such episodes (it probably dominate the mass-loss budget for stars above 60 M) critically altering their evolution (Humphreys and Davidson 1979, 1994, Smith and Owocki 2007). But the theory is still largely conjectural, most likely involving poorly understood instabilities near the Eddington limit; see, e.g., A.S.P. Conf. 332 (2005).

Four famous examples are eta Car; SN 1954j (= Variable 12 in N2403); SN 1961v in N1058; and P Cygni's outburst 400 years ago(Humphreys, Davidson and Smith 1999). Their historical light curves are shown in Figure 1.


Light curves of the four historical SN impostors. Note the qualitative similarities.

An increasing number of objects initially classified as supernovae (SNe) in current surveys are not true SNe, i.e. the "SN impostors". Some of them appear to be pre-SN extreme events or giant eruptions like eta Car. Others may be related to the classical Luminous Blue Variables (LBVs) such as S Dor and AG Car, while those that are heavily obscured may be in a post-AGB or post red supergiant stage, i.e. the Intermediate Luminosity Red Transients (ILRTs). There is considerable diversity in their observed properties; maximum luminosity and duration of the outburst. This diversity is the sign of a little explored field in stellar astrophysics. Very little is known about the progenitors of the giant eruptions and their evolutionary state.

Many authors refer to these objects as LBVs, but most of the eruptions do not resemble the classical or normal LBV maximum light stage. In quiescence an LBV or S Doradus variable is a moderately evolved hot star, with a B-type supergiant or Of-type/late-WN classification. An LBV eruption causes the wind to become dense and opaque, with a large pseudo-photosphere at T ~ 7000-8000 K resembling the spectrum of an F-type supergiant. On an HR diagram the object thus appears to move toward the right. Since this alters the bolometric correction, the visual brightness increases by about 2 magnitudes while the total luminosity remains nearly constant (Wolf (1989),Humphreys and Davidson (1994) and numerous early references therein) or may decrease (Groh et al. 2009). Such an event can last for several years or even decades. Basic causes remain somewhat mysterious; most proposed explanations invoke an opacity-modified Eddington limit, subphotospheric gravity-mode instabilities, and super-Eddington winds (see Humphreys and Davidson(1994), Glatzel(2005), Owocki and Shaviv (2012)).

In rare cases, however, the luminosity substantially increases during outburst; these have been called giant eruption LBVs (Humphreys and Davidson 1994, eta Car variables (Humphreys, Davidson and Smith 1999), or eta Car analogs (Van Dyk 2005). The distinction between giant eruptions and the more common LBV or S Dor-type variability is often overlooked in the literature. They may be related and originate from similar types of stars, perhaps in the same evolutionary stage, but the physical cause may be different. Certainly the energetics of the eruptions and what we observe are different. There are numerous questions about the origin of the giant eruptions, their relation to normal LBV outbursts, and perhaps even to SNe. These eruptions are important. They may account for considerable mass loss and they indicate that some instability has been overlooked in stellar theory.

Very little is known about the progenitors of these giant eruptions and their evolutionary state. They have may come from a range of initial masses and different evolutionary paths and the origin of the instability may be different. The observational record is sparse because these stars are rare and their importance has only been fully recognized in recent years. A greatly improved census of the likely progenitor class, including the most luminous evolved stars, the LBVs, and the warm and cool hypergiants is now needed for a complete picture of the final pre-SN stages of very massive stars. Few LBVs and hypergiants are known in our galaxy due to their rarity, uncertainties in distance, and the infrequency of the LBV eruptions. Even the Magellanic Clouds do not provide a large enough sample to properly determine the relative numbers, duration, and properties of these unstable and eruptive variables. For these reasons we have begun a census of the evolved and unstable luminous star populations in several nearby resolved galaxies: M33 M31, NGC2403 M81 and M101.