Mira Variables: Pulsation and Stellar Evolution

• Mira variables are pulsating red giants on the asymptotic giant branch characterized by 100–1000 day periods, high amplitude variations, and complex circumstellar environments.
• They exhibit fundamental-mode radial pulsations driven by the κ-mechanism that establish tight period–luminosity relations in the near-infrared, essential for accurate distance measurements.
• Observational studies reveal dynamic atmospheres with shock-induced molecular levitation, substantial dust-driven mass loss, and intricate binary interactions in systems like symbiotic Miras.

In astrophysics, Mira refers both to the prototype star $o$ Ceti and to the class of Mira variables: long-period, large-amplitude, fundamental-mode pulsating red giants at the tip of the asymptotic giant branch (AGB). They are cool, luminous, strongly mass-losing stars with extended molecular and dusty atmospheres, and they occupy a central position in late stellar evolution, circumstellar-wind physics, interacting binaries, and the extragalactic distance scale. In some binaries a Mira becomes the cool component of a symbiotic Mira, where an accreting white dwarf interacts with the Mira wind; in other contexts, short-period O-rich Miras serve as infrared standard candles and Type Ia supernova calibrators (Willson et al., 2012, Munari et al., 2010, Huang, 2024).

1. Definition, classification, and evolutionary setting

Mira variables are red, late-type giants with periods of roughly $100$–$1000$ days, visual amplitudes of at least about $2.5$ magnitudes, and spectral types $M$, $S$, or $C$. They are radially pulsating AGB stars in a very late evolutionary stage, and the broader Mira phase is associated with low- to intermediate-mass stellar evolution near the end of envelope retention. The literature summarized here places them on the thermally pulsing AGB, with a degenerate C–O core, hydrogen- and helium-burning shells, and deep convective envelopes (Willson et al., 2012).

Their surface chemistry divides them into O-rich and C-rich subclasses. O-rich Miras have $\mathrm{C/O}<1$ and show oxygen-bearing molecules and silicate-rich dust; C-rich Miras have $\mathrm{C/O}>1$ after third dredge-up episodes have altered the envelope composition. This chemical distinction is not merely spectroscopic: it changes atmospheric opacity, dust chemistry, wind properties, and the form of useful period–luminosity relations. The review literature also emphasizes that thermal pulses and dredge-up events can shift luminosity, radius, and pulsation period over long timescales, so a Mira’s period is not a static identifier of mass alone (Willson et al., 2012).

Miras are also defined by their circumstellar environment. They lose mass at high rates, often through dusty winds that produce OH, H$_2$O, and SiO masers, and many long-period Miras become OH/IR sources. Their visual light curves can change by factors of $100$0–$100$1, while bolometric variation is much smaller, indicating that atmospheric opacity changes, especially in molecular bands, dominate the visible modulation (Willson et al., 2012).

2. Pulsation and period–luminosity structure

The established picture is that Miras are fundamental-mode radial pulsators. Their oscillations are driven by the $100$2-mechanism in the H $100$3 H$100$4 and He $100$5 He$100$6 partial ionization zones, and their periods follow the usual period–mean density scaling,

$100$7

For models of red giants with Mira-like parameters, the review gives representative radial-mode relations

$100$8

for the fundamental and first overtone, respectively, with $100$9 in days and $1000$0, $1000$1 in solar units (Willson et al., 2012).

Observationally, Miras occupy the bright end of the fundamental-mode sequence in Magellanic Cloud period–luminosity diagrams. In the near-infrared, where extinction is reduced and molecular-opacity effects are less disruptive than in the optical, their period–luminosity relations are tight enough for distance work. A standard form is

$1000$2

and the current distance-ladder work concentrates on short-period O-rich Miras, usually $1000$3 d, because this regime is both observationally clean and close to linear in the near-IR (Huang, 2024).

Long-period Miras, typically $1000$4 d, are more complicated. The distance-ladder review associates them with higher luminosities and, in many cases, hot-bottom burning, which can steepen or otherwise distort the period–luminosity relation. This makes them less straightforward calibrators in the present framework, but also potentially powerful because they are substantially brighter and may eventually support direct measurements of $1000$5 without Type Ia supernovae (Huang, 2024).

3. Atmospheres, shocks, dust, and winds

Mira atmospheres are extended, dynamic, and strongly structured. Near-infrared spectro-interferometry of six Mira variables shows that the visibility function decreases in H$1000$6O and CO bands, confirming molecular atmospheres located above the continuum radii. The same observations indicate large-scale inhomogeneities or clumps that contribute a few percent of the total flux, with structure varying on timescales of about three months and above. Both 1D CODEX dynamic atmospheres and azimuthally averaged 3D CO5BOLD convection simulations reproduce the molecular-extension signature, supporting the view that pulsation- and shock-driven levitation lifts material to the radii required for molecule and dust formation (Wittkowski et al., 2016).

The atmospheric dynamics inferred from these studies are shock-dominated. The 3D AGB simulations, once azimuthally averaged, are consistent with roughly spherically expanding shocks of the same general nature as those in self-excited 1D pulsation models. This is an important result because it links resolved interferometric signatures directly to the global dynamics of the atmosphere rather than only to localized convection cells. In Miras, unlike red supergiants, these pulsation- and shock-induced dynamics can levitate the molecular atmosphere to extensions consistent with observation (Wittkowski et al., 2016).

ALMA long-baseline observations of $1000$7 Ceti resolve the millimetre continuum and inner wind at $1000$8 mas resolution. At 229.6 GHz, Mira A is consistent with a radio photosphere of brightness temperature $1000$9 K, and the data do not confirm a compact $2.5$0 mas region of enhanced brightness. Radiative-transfer modeling of SiO and H$2.5$1O line absorption and emission indicates that SiO-bearing gas starts to deplete beyond $2.5$2 at kinetic temperatures $2.5$3 K, implying that the inner dust shells are probably composed of grain types other than pure silicates. During those observations the atmosphere generally showed infall motion, with a shock front of velocity $2.5$4 km s$2.5$5 outside the radio photosphere (Wong et al., 2016).

These resolved results sit naturally within the standard wind-density picture,

$2.5$6

and within the broader observational range of Mira mass-loss rates, $2.5$7, reported for Mira variables in general (Wittkowski et al., 2016, Munari et al., 2010).

4. Mira stars in binaries and symbiotic systems

The nearest and best-studied interacting Mira system is Mira AB. ALMA resolved both components from 94 to 679 GHz and found that Mira A follows a single power law,

$2.5$8

from centimetre to submillimetre wavelengths, consistent with an optically thick thermal radio photosphere surrounding the evolved star. Mira B, by contrast, requires at least two ionized components: a compact, dense inner region producing a nearly thermal spectrum with $2.5$9 between submillimetre and $M$0 mm, and a larger, less dense, inhomogeneous ionized envelope with $M$1 at longer wavelengths. The same ALMA data also showed a significant $M$2 mas discrepancy with the previously predicted orbit, demonstrating sub-mas relative astrometry and implying that the orbital solution needs revision (Planesas et al., 2015).

A different binary configuration is the symbiotic Mira, a pulsating Mira embedded in a binary with a compact hot companion, almost always a white dwarf, that accretes from the Mira wind. Compared with the more common $M$3-type symbiotics, these $M$4-type systems have a Mira instead of a non-Mira red giant, much denser and more extended dusty winds, and larger separations, with wind accretion rather than Roche-lobe overflow. The white dwarf’s radiation and eruptions can reduce or disrupt dust formation in the inner wind and produce strong nebular emission lines (Munari et al., 2010).

The 2010 nova in V407 Cyg is an especially instructive symbiotic-Mira case. The system contains an accreting white dwarf and an O-rich Mira pulsating with a 745 day period. Although Miras with such a long period are generally OH/IR sources hidden by thick dust, V407 Cyg’s dust envelope is described as much thinner, probably because orbital motion, the white dwarf’s hard radiation field in quiescence, and violent mass ejection during outbursts inhibit dust formation in a large fraction of the Mira wind. The nova itself is classified as a very fast He/N nova erupting inside the dense wind of the cool giant companion (Munari et al., 2010).

Spectroscopy during the outburst turns the nova into a time-resolved probe of the Mira wind. The narrow H$M$5 component, attributed to Mira-wind gas ionized by the thermonuclear flash, decays on a recombination timescale of about four days, implying

$M$6

for the ionized wind component. The broad H$M$7 component, attributed to nova ejecta, narrows dramatically with time, from a FWHM of $M$8 km s$M$9 on day $S$0 to $S$1 km s$S$2 on day $S$3, demonstrating strong deceleration as the ejecta sweep up the Mira wind. At the same time, a sharp absorption component at $S$4 km s$S$5 persists, and [N II] emission with FWHM $S$6 km s$S$7 appears two months after maximum, showing that an outer neutral and unperturbed region of the Mira wind survives throughout the observed outburst. The authors therefore infer a stratified wind with inner shocked and ionized zones, an intermediate interaction region, and an outer neutral region that may even allow some ejecta to remain bound and be re-accreted later (Munari et al., 2010).

5. Distance-scale and cosmological applications

Miras have become an alternative distance ladder to Cepheids and the tip of the red giant branch because short-period O-rich Miras are bright in the near-IR, belong to older stellar populations, and can be identified and characterized with infrared observations alone. This makes them valuable in galaxies that do not host young Cepheids and in systems where TRGB stars are too faint or difficult to isolate. Their near-IR amplitudes are smaller than in the optical, their light curves are close to sinusoidal, and their spectral energy distributions peak near $S$8–$S$9, all of which improve period determination and mean-magnitude estimation (Huang, 2024).

Current calibration work uses geometric anchors. In NGC 4258, fitting the F160W period–luminosity relation to a gold sample of O-rich Miras gives zeropoints $C$0 mag or $C$1 mag, depending on the exact bandpass treatment. In the LMC, after transformation to F160W and analogous selection cuts, the corresponding calibration is $C$2 mag. These are mutually consistent and support the use of short-period O-rich Miras as standard candles in the near-IR (Huang, 2024).

With those zeropoints in hand, Mira distances can calibrate local Type Ia supernovae. For NGC 1559, host of SN 2005df, the Mira-based distance modulus is

$C$3

corresponding to $C$4 Mpc, and yields a supernova fiducial luminosity $C$5 mag. For M101, host of SN 2011fe, the Mira distance modulus is

$C$6

Using the standard low-redshift supernova relation

$C$7

these calibrations lead to

$C$8

from NGC 1559 alone, and

$C$9

when NGC 1559 and M101 are combined. The review argues that long-period Miras, because they can be $\mathrm{C/O}<1$0–$\mathrm{C/O}<1$1 magnitudes brighter, may eventually enable a two-rung ladder that excludes Type Ia supernovae altogether (Huang, 2024).

6. Modeling frontiers and unresolved questions

One open line of inquiry concerns whether Mira winds are purely dust-driven or require a dynamically important magnetic field. A hybrid MHD plus dust-driven equatorial wind model for $\mathrm{C/O}<1$2 Ceti found a fiducial solution with $\mathrm{C/O}<1$3 G, Alfvén radius $\mathrm{C/O}<1$4, dust parameter $\mathrm{C/O}<1$5, dust condensation radius $\mathrm{C/O}<1$6, and terminal wind speed $\mathrm{C/O}<1$7 km s$\mathrm{C/O}<1$8. In that framework Mira can plausibly be described as a magneto-dusty rotator rather than a purely dust-driven outflow. A second, hot-spot-related solution with dust forming farther out was found to be mathematically possible but extremely sensitive to perturbations, suggesting that such outer dust-shell formation is likely time-dependent and unstable (Thirumalai et al., 2012).

More broadly, the review literature identifies several unresolved problems: the detailed physics of Mira mass loss; the interaction of convection, shocks, and dust formation in fully 3D atmospheres; the origin of long-term secular period changes not obviously attributable to thermal pulses; and the diversity of evolutionary outcomes, including the fact that many AGB stars may not form visible planetary nebulae. Continued optical, infrared, radio, interferometric, and millimetre observations remain central because Miras are nearby laboratories where pulsation, atmospheric chemistry, wind launching, binary interaction, and distance-scale calibration can all be confronted with resolved data (Willson et al., 2012).

Taken together, these lines of work define Mira variables as more than large-amplitude pulsators. They are simultaneously probes of nonlinear stellar pulsation, circumstellar shock physics, dust and molecular chemistry, binary accretion and nova interaction, and infrared cosmological distance calibration. The present literature suggests that this combination of accessibility and physical complexity is precisely what makes Mira stars enduringly important to stellar astrophysics.