🔭 astrophysics

Sensitivity of spectral lines to granulation: from the Sun to K-type stars

This study demonstrates that a line-by-line diagnostic derived from 3D magneto-convection simulations can effectively identify granulation-sensitive spectral lines in late-G and K-type stars, revealing that optimal line selections for mitigating radial-velocity jitter differ significantly from solar-optimized choices due to evolving convective velocities and ionization balances.

Original authors: V. Vasilyev, K. Sowmya, A. I. Shapiro, N. Kostogryz, D. Vukadinovic, V. Witzke, T. Bhatia, A. Collier Cameron, L. Gizon, S. K. Solanki

Published 2026-05-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: V. Vasilyev, K. Sowmya, A. I. Shapiro, N. Kostogryz, D. Vukadinovic, V. Witzke, T. Bhatia, A. Collier Cameron, L. Gizon, S. K. Solanki

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

The Big Problem: The Star's "Shaky Hand"

Imagine you are trying to spot a tiny, dark speck (an Earth-like planet) passing in front of a bright light bulb (a star). To do this, astronomers measure the star's wobble. But there's a problem: the surface of the star isn't smooth and still. It's like a pot of boiling water.

Hot bubbles rise up (granules), and cooler water sinks down (lanes). This churning motion makes the star's surface jitter. This jitter creates a "noise" in the measurements that is about 10 times louder than the tiny signal of an Earth-sized planet. It's like trying to hear a whisper while someone is shaking the microphone.

The Solution: Picking the Right "Microphones"

The authors of this paper propose a clever trick. They realized that not all "colors" (spectral lines) of light coming from the star react to this boiling surface in the same way.

Think of the star's light as a choir singing a chord.

Some singers (spectral lines) are very sensitive to the shaking; their voices wobble a lot.
Other singers are very steady; their voices stay on pitch even when the stage shakes.

If you listen to the whole choir at once, the shaking messes up the measurement. But if you can identify the "steady singers" and ignore the "wobbly ones," you can filter out the noise and hear the planet's signal much better.

What They Did: Simulating Different Stars

In a previous study, the team figured out which "singers" were steady for our Sun. In this new paper, they asked: "Does this work for cooler stars, like K-type dwarfs?"

K-type stars are popular targets for finding Earth-like planets because they are stable and long-lived. However, they are cooler than the Sun, which changes the physics of their "boiling pot."

The team used powerful supercomputer simulations to create 3D models of:

The Sun (a hot star).
A G9 star (slightly cooler).
A K4 star (even cooler).

They didn't just look at the stars; they simulated the actual churning motion of the gas on their surfaces and calculated how thousands of different iron lines in their light would react to that churning.

The Key Discovery: The "Ionization Switch"

The most important finding is that the rules for picking "steady singers" change as the star gets cooler.

1. The Sun (Hotter):

Neutral Iron (Fe I): These lines are a bit wobbly in strength but okay in position.
Ionized Iron (Fe II): These lines are very wobbly in position.

2. The Cooler Stars (K-type):
As the star cools down, the chemistry changes. Neutral iron becomes the "majority species" (the most common form), while ionized iron becomes rare.

Neutral Iron (Fe I): In cooler stars, these lines become the super-steady singers. They barely move when the surface boils. They are the best candidates to use for finding planets.
Ionized Iron (Fe II): These become the super-wobbly singers. They react strongly to the boiling surface and should be avoided.

The "Recipe" Doesn't Transfer

The paper warns astronomers: You cannot use the same list of "good lines" for a K-star that you use for the Sun.

It's like baking a cake. If you have a recipe that works perfectly for a hot oven (the Sun), you can't just use that exact same recipe for a cooler oven (a K-star) and expect the cake to turn out the same. The ingredients (the physics of the lines) react differently to the temperature.

Specifically, the team found that:

Lines that were stable in the Sun might become unstable in a K-star.
Lines that were unstable in the Sun might become the best choice for a K-star.

The "Where" Matters

The authors also figured out why this happens. They looked at how deep inside the star each line forms.

Lines that form deep in the star's atmosphere (where the gas is moving faster and more violently) tend to be more wobbly.
Lines that form higher up (where things are calmer) tend to be steadier.

In cooler stars, the "steady" lines (Neutral Iron) tend to form in layers that are less affected by the violent churning, while the "wobbly" lines (Ionized Iron) are more sensitive to the turbulence.

The Bottom Line

To find Earth-like planets around cooler stars, astronomers need to build custom "masks" (lists of lines) specifically for those stars. They should focus on Neutral Iron lines and ignore the Ionized Iron lines. If they stick to the old Sun-based lists, they might miss the planet or get a false alarm because they are listening to the wrong "singers" in the choir.

This research provides a physics-based guide to help astronomers tune their instruments to hear the quiet whisper of an Earth-like planet over the roar of the star's boiling surface.

Technical Summary: Sensitivity of spectral lines to granulation: from the Sun to K-type stars

Problem Statement
Extreme-precision radial-velocity (EPRV) searches for Earth analogs are increasingly limited by stellar "jitter" driven by photospheric convection (granulation) and magnetic activity, rather than instrumental noise. Granulation induces time-variable spectral line profiles and Doppler shifts that can obscure planetary signals, producing fluctuations of $\sim 1$ m/s on Sun-like stars—orders of magnitude larger than the $\sim 10$ cm/s signal expected from an Earth-mass planet. While line-by-line approaches offer a path to mitigate this noise by exploiting differential sensitivities of spectral lines to convective inhomogeneities, existing diagnostics are largely solar-centric. A physics-based, spectral-type-dependent diagnostic is required to determine if line-selection strategies optimized for the Sun are portable to cooler late-G and K dwarfs, which are prime targets for habitable-zone planet searches.

Methodology
The authors extend the framework established in Paper I (Sowmya et al. 2025) from the Sun to G9 and K4 dwarfs. The study utilizes:

3D Simulations: Near-surface magneto-hydrodynamic (MHD) simulations of the Sun, G9, and K4 dwarfs using the MURaM code. These simulations include self-consistently generated small-scale turbulent dynamos and span sub-surface convective layers through the photosphere.
Spectral Synthesis: High-resolution spectra are synthesized using the MPS-ATLAS code in Local Thermodynamic Equilibrium (LTE). The synthesis employs a "1.5D" approach, solving radiative transfer along parallel rays through the 3D velocity and thermodynamic fields.
Spatial Sampling: Instead of computationally expensive time-series analysis, the study treats spatial variability within a single simulation snapshot as an ergodic proxy for temporal variability. Spectra are spatially averaged over $16 \times 16$ pixel sub-domains to generate a grid of 1,024 independent realizations per line.
Line Selection: The analysis focuses on neutral (Fe i) and singly ionized (Fe ii) iron lines spanning a wide range of excitation potentials ( $E_{low}$ ) and oscillator strengths ( $\log gf$ ).
Sensitivity Metrics: Two parameters quantify granulation sensitivity for each line:
1. Velocity Sensitivity ( $\sigma_v$ ): The spatial root-mean-square (RMS) scatter of the line center-of-gravity (COG) wavelength, scaled to velocity units. This traces convective shifts and asymmetries.
2. Strength Sensitivity ( $S_W$ ): The fractional scatter of the equivalent width ( $W$ ), tracing thermodynamic fluctuations and opacity variations.
Formation Depth Analysis: The study computes RV-weighted contribution functions to determine the characteristic line-formation temperature ( $T_{1/2}$ ), specifically weighting layers that contribute most to the Doppler signal (primarily line flanks).

Key Results

Species-Dependent Trends: As effective temperature ( $T_{eff}$ $T_{e f f}$ ) decreases from the Sun to K4, convective velocities weaken, and the ionization balance shifts.
- Fe i (Neutral): In cooler stars, Fe i becomes the majority species. These lines exhibit lower velocity sensitivity ( $\sigma_v$ ) and smaller fractional strength variability ( $S_W$ ) compared to the solar case. They form a compact, low-sensitivity cluster in the $(\sigma_v, S_W)$ plane.
- Fe ii (Ionized): Fe ii becomes a minority species in K dwarfs. These lines show the opposite trend: they retain relatively high velocity sensitivity and exhibit larger and more dispersed strength variability ( $S_W$ ) in cooler models compared to the Sun.
Non-Portability of Solar Selections: There is no simple linear scaling between solar and cooler-star sensitivities. Specifically, solar $S_W$ values are poor predictors of strength stability in K dwarfs. Line selections optimized on the Sun based on equivalent-width stability cannot be assumed to transfer to late-G or K dwarfs.
Formation Depth Correlation: Velocity sensitivity ( $\sigma_v$ $σ_{v}$ ) correlates with the RV-weighted formation temperature ( $T_{1/2}$ $T_{1/2}$ ). Lines forming deeper in the photosphere (higher $T_{1/2}$ $T_{1/2}$ ) generally exhibit larger $\sigma_v$ $σ_{v}$ because they sample regions with larger vertical velocity fluctuations.
- For Fe ii, $\sigma_v$ increases systematically with $T_{1/2}$ .
- For Fe i, the relationship is structured by excitation potential and equivalent width; lower-excitation, stronger lines tend to form higher (cooler) and show lower sensitivity.
Impact of Line Strength: Within a given species and excitation range, weaker lines (smaller equivalent widths) tend to form deeper and show higher velocity sensitivity than stronger lines.

Significance and Claims
The paper establishes that granulation sensitivity is not a universal property of a spectral line but is fundamentally dependent on the stellar spectral type and the local ionization balance.

Diagnostic Validity: The line-by-line sensitivity diagnostic derived from 3D simulations remains robust for separating "robust" (low sensitivity) and "sensitive" (high sensitivity) lines in late-G and K dwarfs.
Strategic Implication: For EPRV surveys targeting K dwarfs, spectral masks should be constructed using spectral-type-specific line weights. Fe i lines are identified as natural candidates for up-weighting due to their reduced granulation-driven variability, while Fe ii lines, particularly weak transitions, are candidates for down-weighting or exclusion.
Limitation of Current Practices: The authors caution that relying on solar-optimized line lists, especially those prioritizing equivalent-width stability over velocity sensitivity, may fail to minimize RV jitter in cooler stars due to the changing physics of line formation and ionization.

The study concludes that while the overall level of convective RV noise decreases toward cooler stars, the differential behavior of line families changes significantly, necessitating a shift from solar-centric to spectral-type-aware line selection strategies.