🔭 astrophysics

VLTI-GRAVITY measurements of cool evolved stars: II. Pulsation properties and mass-loss process of the Mira star R Car and the red supergiant VX Sgr

Using the largest VLTI-GRAVITY time-series dataset to date, this study reveals that while the Mira star R Car exhibits steady mass loss driven by stable, large-amplitude fundamental mode pulsations, the red supergiant VX Sgr undergoes extreme mass-loss events triggered by shifts in pulsation modes and strong atmospheric shocks.

Original authors: D. Jadlovský, M. Wittkowski, A. Chiavassa, K. Kravchenko, B. Freytag, S. Höfner, J. Krtička, C. Paladini, G. Rau, M. Brož, T. Granzer, M. Weber

Published 2026-04-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: D. Jadlovský, M. Wittkowski, A. Chiavassa, K. Kravchenko, B. Freytag, S. Höfner, J. Krtička, C. Paladini, G. Rau, M. Brož, T. Granzer, M. Weber

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the night sky is filled with aging giants. Some are like Mirrors (Mira variables), which are old, pulsating stars that breathe in a very steady, rhythmic way. Others are Red Super Giants, which are massive, chaotic behemoths that are on the verge of exploding.

This paper is like a high-definition, 7-year-long documentary filmed by a team of astronomers using the world's most powerful "eye" in the sky: the VLTI-GRAVITY telescope. They watched two specific giants: R Car (a steady Mira) and VX Sgr (a wild Red Super Giant).

Here is the story of what they found, explained simply.

1. The Telescope: A Giant Eye with a Zoom Lens

The astronomers didn't just take a picture; they used a technique called interferometry. Imagine trying to see the details on a coin from a mile away. You can't do it with one eye. But if you have two eyes (or two telescopes) far apart and you combine their view, you can see tiny details.

They combined four telescopes to create a virtual lens the size of a football field. This allowed them to see the "skin" (photosphere) and the "atmosphere" (the gas cloud around the star) of these distant stars in incredible detail, measuring how big they were down to the millimeter (in astronomical terms).

2. The Two Stars: The Metronome vs. The Storm

R Car: The Metronome

R Car is a Mira variable. Think of it as a cosmic metronome.

The Rhythm: It breathes in and out perfectly. It gets bright, then dim, then bright again, every 307 days.
The Atmosphere: As the star "breathes," its skin expands and contracts.
- The Surprise: The star's skin gets biggest just before it gets darkest. It's like a balloon that expands right before you let the air out.
- The Layers: The star isn't just a solid ball. It has layers like an onion.
  - The inner skin (photosphere) moves first.
  - The atomic layer (Titanium/Scandium) follows a few weeks later.
  - The water vapor layer follows even later.
  - The carbon monoxide (CO) layer is the outermost skin, stretching out to 1.5 to 1.7 times the size of the star itself.
The Verdict: R Car is a stable, predictable machine. Its steady breathing pushes dust and gas out into space, creating the "wind" that helps build new stars.

VX Sgr: The Storm

VX Sgr is a Red Super Giant. Think of it as a chaotic, massive storm cloud that is 1,500 times bigger than our Sun.

The Rhythm: It used to be wild, with huge, irregular bursts of brightness. But recently, it changed its tune.
The "Great Dimming" Event: In 2020–2021, VX Sgr had a massive tantrum.
- It experienced two huge shockwaves (like sonic booms) traveling through its atmosphere.
- These shocks pushed the star's outer layers (water and carbon gas) to expand to 2.2 times its normal size!
- The Smoking Gun: During this explosion, the astronomers saw a specific type of hydrogen light (Brγ) that usually only appears in binary stars (two stars orbiting each other). But the data showed this light was coming from inside VX Sgr's own atmosphere, proving it was a shockwave, not a companion star.
The Aftermath: After this massive event, the star calmed down. Its "breathing" changed from a deep, slow, massive beat (Fundamental Mode) to a quick, shallow, weak beat (First Overtone). It essentially switched from a deep drum to a high-pitched whistle.

3. The Big Picture: How Stars Lose Weight

Why does this matter? These stars are the universe's recyclers. They lose mass, and that mass becomes the dust and gas needed to make new stars and planets.

For the Metronome (R Car): The mass loss is a steady, gentle process. The regular breathing pushes the gas out slowly and consistently.
For the Storm (VX Sgr): The mass loss is episodic and violent. It doesn't just leak; it erupts. The paper suggests that Red Super Giants might lose most of their mass in these sudden, extreme "shock" events, rather than a steady wind.

4. The Computer Simulation: Did the Math Match?

The team compared their real observations with super-computer simulations (CO5BOLD models).

The Good News: The computer models got the "breathing" right. They predicted that the star's skin would expand before it gets dark, and they matched the behavior of the Carbon Monoxide gas.
The Bad News: The models were too "dry." They didn't predict as much water vapor as the astronomers saw. This suggests the computer models need to be updated to better understand how heat and light interact with water molecules in these extreme environments.

Summary

This paper is like a medical check-up for two different types of dying stars.

R Car is healthy and rhythmic, losing mass steadily like a slow, steady leak.
VX Sgr is a patient who had a sudden heart attack (the shock event), changed its heartbeat, and then settled into a new, quieter rhythm.

The discovery that Red Super Giants like VX Sgr can undergo these massive, sudden shock events helps us understand how they might eventually explode as supernovae, seeding the universe with the ingredients for life.

1. Problem Statement

The physical mechanisms driving mass loss in cool evolved stars—specifically Red Supergiants (RSGs) and Asymptotic Giant Branch (AGB) stars—remain poorly constrained, particularly regarding the initiation of mass loss in the near-surface atmospheric layers.

AGB Stars: Mass loss is generally understood to be driven by radial fundamental-mode (FM) pulsations and convection, which levitate material to distances where dust forms, followed by radiation pressure on dust grains.
RSGs: The mechanism is less clear. While they share observational similarities with AGBs (extended atmospheres, variability), their pulsation is often irregular, involving multiple modes (overtones, Long Secondary Periods). Recent events like the "Great Dimming" of Betelgeuse suggest that episodic mass-loss events driven by strong shocks may play a critical role, a hypothesis that requires further observational validation.
Gap: There is a lack of long-term, high-resolution time-series interferometric data comparing the temporal variability of extended atmospheric layers (molecular bands) against the photosphere and light curves for both AGB and RSG stars.

2. Methodology

The authors utilized the VLTI-GRAVITY instrument in the near-infrared K-band (2.0–2.4 µm) to conduct the largest time-series interferometric study of cool evolved stars to date.

Dataset: 54 snapshots collected over ~7 years (2018–2025):
- R Car: A typical oxygen-rich Mira variable (AGB), 18 observations.
- VX Sgr: An extreme RSG, 36 observations.
Observational Strategy: Used the small configuration of the 4 Auxiliary Telescopes (ATs) (baselines 11.3–33.9 m) to resolve the stellar disk and its immediate extensions without resolving substructures.
Data Analysis:
- Spectral Selection: Analyzed specific spectral regions to probe different atmospheric depths:
  - Continuum (2.22–2.28 µm): Photosphere ( $R_\star$ ).
  - Atomic Lines (Ti I, Sc I): Intermediate layers.
  - H $_2$ O Bands (1.98–2.05 µm): Extended molecular layers.
  - CO Bands (2.29–2.38 µm): Most extended layers.
- Modeling: Fitted visibility amplitudes ( $|V|^2$ ) using a Uniform Disc (UD) model with a scaling factor ( $A$ ) to account for spatially over-resolved background flux.
- Phase Analysis: Correlated angular diameter variations with visual light curves (AAVSO data) to determine phase shifts ( $\Delta\phi_{vis}$ ).
- Simulation Comparison: Compared observational data with CO5BOLD 3D radiative-hydrodynamics (RHD) simulations of an AGB model (st28gm05n057) to test theoretical predictions of pulsation and atmospheric extension.
- Complementary Data: Utilized optical spectra from the STELLA telescope to analyze radial velocities and emission lines (Balmer series, Br $\gamma$ ).

3. Key Contributions

Largest Time-Series Dataset: Provided the most extensive VLTI-GRAVITY time-series to date, covering multiple pulsation cycles for both an AGB star and an RSG.
Multi-Layer Variability: Successfully mapped the variability of the photosphere and distinct atmospheric layers (atomic, H $_2$ O, CO) simultaneously, quantifying their phase shifts relative to the light curve.
Discovery of Extreme Event in VX Sgr: Identified a massive mass-loss event in VX Sgr (2020–2021) analogous to Betelgeuse's Great Dimming, characterized by shock propagation, extreme atmospheric expansion, and a change in pulsation mode.
Detection of Br $\gamma$ in RSG: Uniquely detected Brackett $\gamma$ emission in the interferometric data of a single RSG (VX Sgr) during a shock event, distinguishing it from binary accretion scenarios.
Model Validation: Provided a critical benchmark for 3D RHD simulations, highlighting where models succeed (CO extension) and where they fail (H $_2$ O strength).

4. Key Results

A. R Car (Mira Variable)

Pulsation: Exhibits regular fundamental-mode (FM) pulsations with a period of $P \approx 307.4$ days.
Radius: Mean photospheric radius $R_\star = 280 \pm 25 R_\odot$ .
Phase Shifts: The photosphere and near-surface layers show maximum radii before minimum brightness (anti-correlated with light curve).
- Continuum max at $\phi_{vis} \approx 0.43$ .
- Ti I/Sc I max at $\phi_{vis} \approx 0.46$ .
- H $_2$ O max at $\phi_{vis} \approx 0.55$ .
- This implies a delay of ~2–3 weeks between layers, consistent with shock propagation velocities of 15–20 km/s.
Extension: CO layers extend to $1.3 - 1.7 R_\star$ .
Emission: Detected Br $\gamma$ emission near brightness maxima, attributed to shock propagation.

B. VX Sgr (Red Supergiant)

Dual-Mode Behavior: The star exhibits two distinct cycles:
1. Active Cycle: Large-amplitude FM pulsations ( $P \approx 748$ days, $\Delta R \approx 13\%$ of $R_\star$ ).
2. Quiescent Cycle: Low-amplitude pulsations near the first overtone (O1, $P \approx 316$ days, $\Delta R \approx 4\%$ of $R_\star$ ).
Radius: Active phase $R_\star \approx 1556 \pm 110 R_\odot$ ; Quiescent phase $R_\star \approx 1456 \pm 108 R_\odot$ .
The 2020–2021 Event:
- Preceded by two strong shocks (indicated by FWHM increases in optical spectra).
- Resulted in extreme expansion of H $_2$ O and CO layers up to $\sim 2.2 R_\star$ .
- Emission: Detected strong Balmer emission and Br $\gamma$ in interferometric data. Radial velocity analysis confirmed the emission originated from the stellar atmosphere (shock), not a binary companion.
- Mode Switch: Following the event, the pulsation mode shifted from FM to O1.

C. Comparison with CO5BOLD Simulations

Agreement: The AGB model successfully reproduced regular pulsations and the general trend of photospheric diameter preceding brightness minimum. CO layer extensions ( $\sim 1.4 R_\star$ ) matched observations well.
Discrepancy: The model significantly underestimated the extension and flux contribution of H $_2$ O bands compared to observations.
Cause: Likely due to the "gray Rosseland opacity" approximation in the simulation, which may result in higher upper-atmosphere temperatures, preventing sufficient H $_2$ O formation.

5. Significance and Conclusions

Mass-Loss Mechanisms: The study supports the view that AGB mass loss is driven by stable, large-amplitude FM pulsations. In contrast, RSG mass loss appears dominated by episodic extreme events linked to shock waves and changes in pulsation modes (from low-amplitude O1 to high-amplitude FM).
RSG Nature: The analysis of VX Sgr's luminosity and pulsation properties confirms it is an extreme massive RSG ( $20-40 M_\odot$ ), refuting previous hypotheses that it might be a super-AGB star or Thorne–Żytkow object.
Shock Physics: The detection of Br $\gamma$ and Balmer emission in VX Sgr provides direct evidence that strong shocks propagate through RSG photospheres, levitating material and potentially triggering mass ejection events similar to Betelgeuse's Great Dimming.
Future Modeling: The discrepancy in H $_2$ O modeling highlights the urgent need for 3D RHD simulations to incorporate wavelength-dependent opacity to accurately reproduce molecular formation and dust-driven winds in cool evolved stars.

This work establishes a new benchmark for understanding the onset of mass loss in evolved stars, demonstrating that while AGBs follow predictable pulsation cycles, RSGs undergo dramatic, episodic structural changes driven by shocks.