🔭 astrophysics

The CARMENES search for exoplanets around M dwarfs. A homogeneous catalogue of projected rotational velocities accounting for limb-darkening

This paper presents a homogeneous catalogue of projected rotational velocities ( $v \sin i$ ) for 392 M dwarfs observed by the CARMENES spectrograph, derived using a novel oversampled convolution method with limb-darkening that achieves significantly higher precision (6.8% median uncertainty) than previous literature.

Original authors: R. Varas, G. Morello, M. Zechmeister, P. J. Amado, F. J. Pozuelos, J. A. Caballero, A. Claret, C. Cifuentes, R. Morales, A. Quirrenbach, A. Reiners, I. Ribas, V. J. S. Béjar, M. Cortés-Contreras, A. P

Published 2026-04-20

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: R. Varas, G. Morello, M. Zechmeister, P. J. Amado, F. J. Pozuelos, J. A. Caballero, A. Claret, C. Cifuentes, R. Morales, A. Quirrenbach, A. Reiners, I. Ribas, V. J. S. Béjar, M. Cortés-Contreras, A. P. Hatzes, Th. Henning, I. Hermelo, H. L. Ruh, A. Schweitzer, H. M. Tabernero, M. R. Zapatero Osorio

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 you are trying to figure out how fast a spinning top is rotating, but you can only see it from a distance, and it's spinning on a tilted axis. You can't see the top itself clearly; you only see the blur of its edges. In astronomy, this "blur" is how we measure how fast stars spin.

This paper is about a team of astronomers who built a super-sharp, high-definition camera (a new mathematical method) to measure the spin of "M-dwarf" stars (the most common type of star in our galaxy, which are small, cool, and red).

Here is the breakdown of their work using simple analogies:

1. The Problem: The "Fuzzy" Photo

Stars are so far away we can't take a picture of them spinning. Instead, astronomers look at their light. When a star spins, one side moves toward us (making the light look slightly bluer) and the other moves away (making it look slightly redder). This smears out the star's "fingerprint" (its spectrum).

The faster the star spins, the more smeared out the fingerprint becomes. This smearing is called $v \sin i$ (velocity times the sine of the angle).

The Old Way:
Previous methods were like trying to measure that smudge with a ruler made of rubber. They used a "one-size-fits-all" rule to guess how much the star's surface gets darker at the edges (a phenomenon called limb-darkening). It was like assuming every spinning top has the exact same shape and shading, regardless of its size or color. This led to measurements that were often off by about 15%.

2. The Solution: The "High-Def" Upgrade

The authors, led by R. Varas, created a new method using the CARMENES telescope (a super-powerful eye in Spain). They improved the measurement in two clever ways:

The "Oversampling" Trick (The Super-Res Zoom):
Imagine trying to draw a smooth curve using only a few blocky pixels. If you zoom in, the curve looks jagged and bumpy. The old method was like drawing with big blocks.
The new method oversamples the data. It's like taking a low-res photo and using AI to fill in millions of extra pixels to make the curve perfectly smooth before measuring it. This removes the "jagged" errors that used to mess up the results.
The "Custom Tailored" Lens (Limb-Darkening):
Stars aren't uniformly bright; they are brighter in the center and darker at the edges, like a glowing orange that fades toward the peel. The amount of fading changes depending on how hot the star is and what color of light you are looking at.
The old method used a "generic" fading rule for everyone. The new method custom-tailors the rule for every single star and every single color of light. It's like switching from a one-size-fits-all hat to a hat custom-molded to your exact head shape.

3. The Results: A Perfectly Measured Catalogue

They tested this new method on 392 M-dwarf stars.

Precision: The old method had an error margin of about 15%. The new method reduced this to just 6.8%. That's like going from guessing the weight of a bag of flour within 3 pounds to guessing it within less than 1 pound.
New Discoveries: They found the spin rates for 36 stars that nobody had measured before.
The "Young" Stars: By measuring the spin more accurately, they could spot stars that are spinning unusually fast. In the universe, fast spinners are often "teenagers" (young stars) that haven't slowed down yet. They identified several of these young candidates, which helps us understand how planetary systems form around these stars.

4. Why Does This Matter?

Knowing how fast a star spins is like knowing the age of a car.

Age: Fast-spinning stars are usually young; slow-spinning stars are older.
Safety: Fast-spinning stars are often "angry" (very active with flares), which can strip the atmosphere off any planets orbiting them.
Detecting Planets: To find Earth-like planets, we need to know if the star's "wobble" is caused by a planet or just the star's own magnetic tantrums. Accurate spin measurements help us filter out the noise and find the real planets.

The Bottom Line

The authors didn't just build a better ruler; they built a smart, custom-fitting ruler that accounts for the unique shape and color of every star. They have now released a massive, high-precision catalogue of star spins for the entire scientific community to use, making our map of the galaxy's "spinners" more accurate than ever before.

1. Problem Statement

Stellar rotation is a critical parameter for understanding stellar age, magnetic activity, and the evolution of planetary systems. The projected rotational velocity ( $v \sin i$ ) is typically derived from the Doppler broadening of spectral lines. However, accurate determination is challenging because rotational broadening must be disentangled from other mechanisms like instrumental effects and stellar turbulence.

Previous studies often relied on simplified models that assumed a fixed linear limb-darkening coefficient (typically $u = 0.6$ ), independent of the star's spectral type or wavelength. The authors argue that this simplification introduces systematic biases. Furthermore, standard convolution methods on discrete wavelength grids can introduce numerical artifacts (kinks in $\chi^2$ curves) that degrade precision. There was a need for a homogeneous, high-precision catalogue of $v \sin i$ for M dwarfs that accounts for wavelength- and temperature-dependent limb-darkening.

2. Methodology

The authors developed an enhanced method implemented within the serval (SpEctrum Radial Velocity AnaLyser) pipeline, specifically for CARMENES (Calar Alto high-Resolution search for M dwarfs with Exoearths) VIS spectra. The method features two primary innovations:

Oversampled Convolution:
- Standard convolution on a discrete logarithmic wavelength grid causes step-wise changes in the kernel width as $v \sin i$ varies, creating "kinks" in the $\chi^2$ minimization curve.
- To mitigate this, the template spectrum is interpolated onto a finer grid (oversampling factor $N=5$ ), convolved with the rotational kernel, and then downsampled back to the native resolution. This smooths the discontinuities and significantly reduces scatter in $v \sin i$ estimates across different spectral orders.
Order-Specific Limb-Darkening:
- Instead of a fixed coefficient, the authors compute theoretical linear limb-darkening coefficients ( $u$ ) for each spectral order and effective temperature ( $T_{\text{eff}}$ ) using PHOENIX NewEra stellar atmosphere models and the ExoTETHyS software.
- These coefficients account for the wavelength dependence of limb-darkening across the CARMENES VIS range (520–960 nm).
- The method uses a linear limb-darkening law, which was found to provide sufficient precision compared to more complex laws (e.g., power-2), with differences below 0.5%.
Alternative Approach (Numerical Integration):
- The authors also implemented a direct numerical integration method (following Carvalho & Johns-Krull 2023) as a benchmark. While this method offers slightly higher precision for fast rotators ( $v \sin i \gtrsim 30$ km s $^{-1}$ ) and hotter stars ( $T_{\text{eff}} \approx 4000$ K), it is computationally expensive (approx. 3.5 times slower). The convolution method was deemed sufficient for large statistical samples.
Template Selection:
- Templates were selected based on the target's sub-spectral type, choosing the least active and slowest rotators available (e.g., Teegarden's star for M6.5+).
- The method compares observed spectra against these templates using $\chi^2$ minimization over a grid of trial $v \sin i$ values.

3. Key Contributions

Development of a Novel Algorithm: An oversampled convolution method incorporating realistic, order-specific limb-darkening coefficients to minimize numerical artifacts and systematic biases.
Homogeneous Catalogue: The creation of a uniform catalogue of $v \sin i$ measurements for 392 M dwarfs observed by CARMENES.
New Data: The inclusion of 36 new targets with previously unavailable $v \sin i$ measurements.
Re-evaluation of Existing Data: Significant updates to values for several targets where previous literature values were found to be unreliable or inconsistent with the equatorial velocity ( $v_{\text{eq}}$ ).
Identification of Young Candidates: The analysis helped identify and characterize young, fast-rotating M dwarfs based on discrepancies between $v \sin i$ and $v_{\text{eq}}$ , supported by activity indicators (X-ray, H $\alpha$ , UV).

4. Results

Precision Improvement: The new method achieved a median relative uncertainty of 6.8% for $v \sin i > 2$ km s $^{-1}$ . This is a substantial improvement over the 15.4% median uncertainty reported in previous literature.
Systematic Shifts: The inclusion of limb-darkening resulted in derived $v \sin i$ values that are systematically ~5.7% higher than previous literature values (which often used fixed coefficients). This shift is consistent with the theoretical expectation that limb-darkening narrows the effective rotational kernel.
Validation:
- Tests on synthetic spectra (2500–4000 K) showed that the oversampled convolution method reduces scatter by ~50% compared to non-oversampled methods.
- Comparisons between the convolution and numerical integration methods showed excellent agreement (median difference of 0.1%), validating the convolution approach for the majority of the sample.
Outliers and Reliability:
- 21 measurements were flagged as unreliable, often due to unresolved multiple systems or $v \sin i > v_{\text{eq}}$ (physically impossible).
- Six young candidates were identified as fast rotators consistent with their kinematic and activity properties.
- Discrepancies with the Jenkins et al. (2009) catalogue were resolved, with the authors recommending their updated values for seven specific targets.

5. Significance

This work provides the largest uniform catalogue of projected rotational velocities for M dwarfs to date. By correcting for systematic biases related to limb-darkening and numerical discretization, the authors have established a more accurate baseline for stellar rotation.

The significance of this catalogue extends to:

Gyrochronology: More precise $v \sin i$ values allow for better age estimates of M dwarfs, which is crucial for understanding the age of their planetary systems.
Exoplanet Detection: Accurate characterization of stellar rotation and activity is essential for distinguishing between planetary signals and stellar activity noise in radial velocity surveys.
Spin-Orbit Alignment: The data supports studies on the alignment between stellar rotation and planetary orbits.
Future Work: The methodology is designed to be extendable to CARMENES NIR data, promising further improvements in characterizing cool, low-mass stars.

The pipeline, data, and results are publicly available via a GitHub repository and the CDS archive, ensuring reproducibility and utility for the broader astronomical community.