LPV Section Home‎ > ‎

Type C Semiregulars: Waiting for the Spectacular

by David Turner

The Type C semiregular variables, or SRCs, are, like many fascinating objects in astronomy, relatively unstudied. But not unobserved, given that many AAVSO members, and especially the more experienced observers, have been observing them for years.

Their basic properties are easy enough to summarize, although the descriptions given for them in many textbooks
are, quite simply, wrong. As defined in the General Catalogue of Variable Stars (GCVS), the SRC class consists of semiregular supergiant variables of late spectral type (M, C, S or Me, Ce, Se, i.e. class M or their chemically peculiar carbon star equivalents), with M

SRCs on the HR Diagram

SRC stars are super giants located to the right (red) end of the Ia and Ib supergiant areas on the Hertzsrung-Russell diagram above.
u Cephei cited as an example. Their light amplitudes are of order one magnitude or more, and they undergo cycles of anywhere from 30 days to several thousand days. Most of the Milky Way SRCs may be multi-periodic, with primary periods of order 300-900 days and light amplitudes tied to luminosity, according to AAVSO observations. Closely related are the LC variables, irregularly variable M supergiants with visual amplitudes also on the order of one magnitude. A cited example is TZ Cas.
The primary periods of SRC variables fall in the range 300-900 days, and are tied to radial pulsation according to Stothers (1969, ApJ, 156, 541), who also pointed out that they appear to obey a period-luminosity relation different from that of Cepheids . More recent numerical, linear, nonadiabatic, pulsation models by Guo and Li (2002, ApJ, 565, 559) confirm and extend the earlier work of Stothers. Their secondary periods tend to be longer and probably originate from spot cycles on their surfaces.

  AAVSOers may recognize a more familiar prototype SRC variable in Betelgeuse, Alpha Orionis, literally the Shoulder of the Giant (bet el geuse), an object selected previously as the AAVSO star of the year but one that is probably not an ideal example of the class. Betelgeuse is not an easy star for which to make magnitude estimates because of its extreme visual brightness between 0 and 1.3. All of its reference stars lie well outside the standard field of view for observations without optical aid (i.e. "naked eye," no telescope, no binoculars, although with corrective lenses permitted), and calibrated photoelectric or CCD photometry is often difficult for such bright objects.

The SRC variables should not be confused with the more familiar types of long period variables (LPVs): Miras and Type A and B semiregulars. Those objects are all red giants, stars of spectral type M that are about as massive as the Sun and of varying chemical composition, typically with ages of order 1-10 billion years passing through the red giant or asymptotic giant branch (AGB) stage of evolution. They are found just about everywhere in our Galaxy, although more frequently in the disk.

The Type C semiregulars, on the other hand, are young, massive stars ~15-25 times more massive than the Sun with ages of only a few million years, and they are tightly constrained to the Galactic disk. Most such stars help delineate our Galaxy's spiral arms. All are M supergiants, typically the most poorly studied stars in our Galaxy. And, according to stellar evolutionary models for massive stars, they will end their lives as Type II supernovae after exhausting their various sources of nuclear fuel (Chevalier 1981, Fund. Cosmic Phys., 7, 1). Somewhat unexpectedly, perhaps, the most spectacular such events in recent years, SN 1987A in the Large Magellanic Cloud and SN 1993J in M81, originated from B3 I and G8-K5 I supergiants, respectively.

It is interesting to note that, while all SRCs are M supergiants or their chemically peculiar kin, e.g. supergiant carbon stars, is it also true that all M supergiants are Type C semiregular variables? The related class of LC variables does consist of M supergiants, for example, and conceivably they are simply variables like the SRCs where the variability is so poorly expressed that the main periodicity is difficult to detect. It is noted by Pierce, Jurcevic and Crabtree (2000, MNRAS, 313, 271) that M supergiant variables are more common than Cepheid variables in most galaxies, yet classical Cepheids outnumber the SRCs and LCs by a factor of 10 in the GCVS. Have we simply overlooked most of our Galaxy's nearby SRC variables because of their red colors and long pulsation periods?

Dedicated variable star observers are all familiar with the Purkinje effect, in which the sensitivity of the eyes at low light levels shifts to the blue end of the spectrum, so that staring at a red star over an extended period of time leads one to perceive it as being brighter than it actually is. Such difficulties make SRC variables challenging objects to observe, and yet it has not dissuaded the many observers who observe both them and the Miras and Type B and C semiregulars in routine fashion.

My interest in SRC variables began with the discovery that one of them, BC Cyg, was the brightest member of the heavily reddened young cluster Berkeley 87 and was listed in the GCVS with only the barest details regarding its light variability. Berkeley 87 is a relatively obscure open cluster in terms of literature attention, yet it happens to coincide with the strongest source of cosmic rays in the northern hemisphere. Other young clusters lie in the general direction of Berkeley 87, but its stars are the likely origin of the cosmic rays given their extreme peculiarity. Where else in the Galaxy can you find a M supergiant variable, a peculiar emission-line B supergiant, ST3 --- one of the Galaxy's few known Wolf-Rayet stars of the oxygen sequence, V439 Cyg --- an exotic variable that appears to be a Be star hidden from direct view spectroscopically by a surrounding nebular veil, and perhaps a few other exotic stars (?), all collected together in a sparsely populated open cluster? I know of no others myself, and I have been studying our Galaxy's obscure star clusters for more than thirty years.

I managed to complete a study of the long term variability of BC Cyg using the plate collection of the Harvard College Observatory in conjunction with a smaller subset of plates in the collection of the Russian Sternberg Institute examined by my Russian colleagues. BC Cyg displays regular pulsational variations of duration ~700 days, but more interesting is how the pulsation changes over the course of a century. Between 1900 and 2000 BC Cyg brightened by about 60%, while its pulsation period dropped from ~700 days to ~690 days. The most likely origin of such changes is evolution during the red supergiant stage, given that most evolutionary
models of stars of ~20 solar masses exhibit detectable changes on time scales as short as 35 years! If the picture deduced from the star's observed color variations is correct, the brightening of the star at blue/visual wavelengths actually corresponded to a decrease in its overall luminosity, which is highly temperature-dependent for such cool stars. In turn, the decrease in its luminosity and pulsation period resulted from an increase in surface temperature as a result of evolutionary pressures.

The recent picture is even more interesting, since it is here where AAVSO observations play an important role, especially the valuable observations of red variables supplied by long-time observer Paul Vedrenne. Visual observations of BC Cyg indicate that the star has recently faded in the blue/visual region. But decreased brightness at blue/visual wavelengths corresponds to an increase in overall luminosity, which means that the pulsation period of BC Cyg is probably increasing again as the star ascends the red supergiant branch in the H-R diagram and becomes more extended. The question is: how much further can the star evolve up the red supergiant branch before it turns into a Type II supernova? Will it occur within our lifetimes or some time in the distant future? Given the scanty state of information about such stars that is currently available, it is not a question that is easily answered.

Although the existing list of Galactic SRC variables in the GCVS numbers only 55, new photometric surveys will likely turn up many more in years to come. AAVSOers should not be dissuaded from observing them simply because of their red colors and long pulsation periods of 1-2 years, because it may turn out that we are observing the long time-delay fuse on a very promising fireworks display. The deduced mass of BC Cyg, ~19 solar masses, is remarkably close to the estimated masses for the progenitors of SN 1987A and SN 1993J! Perhaps in this instance we will have advance warning of the star's impending doom?

Originally published in AAVSO's "Eyepice Views", January, 2007