🔭 astrophysics

Evidence for stellar contamination and water absorption in NGTS-5b's transmission spectra with GTC/OSIRIS

This study utilizes GTC/OSIRIS optical transmission spectroscopy to reveal that NGTS-5b possesses a relatively clear, water-rich atmosphere with sub-solar metallicity, while demonstrating that unocculted stellar spots significantly contaminate the spectral data and explain discrepancies between transit observations.

Original authors: Wan-Hao Wang, Guo Chen, Chengzi Jiang, Enric Palle, Felipe Murgas, Hannu Parviainen

Published 2026-03-25

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Wan-Hao Wang, Guo Chen, Chengzi Jiang, Enric Palle, Felipe Murgas, Hannu Parviainen

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 trying to listen to a whisper in a crowded room. That's essentially what astronomers are doing when they study the atmosphere of a distant planet like NGTS-5b. They aren't looking at the planet directly; instead, they are listening to the "whisper" of starlight as it filters through the planet's atmosphere during a transit (when the planet passes in front of its star).

This paper is the story of how a team of scientists finally figured out what NGTS-5b is made of, but not before realizing the "crowd" (the star itself) was making a lot of noise.

Here is the breakdown of their discovery in simple terms:

1. The Setup: A Warm "Mini-Neptune"

NGTS-5b is a "warm sub-Saturn." Think of it as a giant gas ball, slightly smaller than Saturn, orbiting very close to its star. Because it's so close, it's quite hot (about 950 Kelvin, or 1,250°F). Scientists wanted to know: What is its atmosphere made of? Is it cloudy? Is it full of water vapor? Is it a clear, blue sky?

2. The Problem: The Star is a "Flickering Lightbulb"

To study the planet, the team used a giant telescope in the Canary Islands (GTC) to watch the planet cross its star on two different nights (Night 1 and Night 2).

However, stars aren't perfect, steady lightbulbs. They have sunspots (cooler, darker patches) and faculae (hotter, brighter patches), just like our Sun.

The Analogy: Imagine trying to measure the tint of a window by looking through it at a lightbulb. If the lightbulb has a smudge on it, or if the smudge moves around, your measurement of the window's tint will be wrong.
The Issue: On Night 1, the star was very "spotty" (active). On Night 2, it was calmer. When the scientists looked at the data, the two nights gave them completely different pictures of the planet's atmosphere. It was like the planet looked like a cloudy swamp on one night and a clear desert on the next.

3. The Detective Work: Separating the Noise

The scientists realized they couldn't just ignore the star's "smudges." They had to build a mathematical model that accounted for the star's changing activity.

The "Hybrid" Solution: They created a model that treated the planet and the star as a team. They figured out that the star had a lot of dark spots on Night 1 (covering about 20-29% of the visible surface) and fewer spots on Night 2.
The Result: Once they mathematically "cleaned" the star's noise from the data, the two different nights finally agreed with each other. The confusion vanished.

4. The Discovery: A Clear, Water-Rich World

After cleaning up the data, the true nature of NGTS-5b's atmosphere emerged:

It's Clear: The atmosphere isn't hidden behind a thick, opaque blanket of clouds. It's relatively clear, allowing us to see deep into it.
It's Wet: The strongest signal they found was water vapor (H₂O). The planet has a lot of water in its atmosphere.
The Chemistry: They found that the planet has a very low ratio of Carbon to Oxygen (less than 0.22). This is interesting because it suggests the planet might have formed in a specific region of its solar system where oxygen was more abundant than carbon.

5. Why This Matters

This paper teaches us two major lessons:

Stars are messy: If you don't account for the star's spots and activity, you might think a planet has a totally different atmosphere than it actually does. It's a reminder that we can't just look at the planet; we have to understand the star it orbits.
NGTS-5b is a "Goldilocks" candidate: It sits in a sweet spot of temperature and size where we expect to find clear atmospheres with water. While we need better data (like from the James Webb Space Telescope) to confirm the exact amounts of other gases, we now know this world is likely a water-rich, clear-skied giant.

In a nutshell: The scientists looked at a planet through a dirty window (the star). They realized the window was dirty and moved around between observations. Once they cleaned the window, they saw that the planet is a clear, water-filled world, not the confusing mess they thought it was.

1. Problem Statement

Transmission spectroscopy is a primary method for characterizing exoplanet atmospheres, but its interpretation is frequently complicated by stellar surface heterogeneity (spots and faculae). Unocculted active regions on the host star can introduce wavelength-dependent contamination (the Transit Light Source effect), mimicking or masking planetary atmospheric features.

Target: NGTS-5b, a warm sub-Saturn ( $M_p \approx 0.229 M_{Jup}$ , $R_p \approx 1.136 R_{Jup}$ , $T_{eq} \approx 952$ K) orbiting a K2-type dwarf.
Challenge: Previous studies of similar systems have shown discrepancies between transit spectra taken at different epochs, often attributed to varying stellar activity levels. There was a need to determine if NGTS-5b's atmosphere is clear or cloudy and to identify key molecular species (like water) while disentangling them from stellar noise.

2. Methodology

The authors conducted a comprehensive analysis using data from two transit events observed with the GTC/OSIRIS instrument (10.4m Gran Telescopio Canarias).

Observations:
- Night 1 (2020-03-10): Observed with the R1000R grism (513–1043 nm).
- Night 2 (2021-03-21): Observed with the R1000B grism (364–789 nm).
- Simultaneous observation of a reference star to correct for telluric and instrumental variations.
Data Reduction:
- Standard reduction (bias, flat-field, cosmic ray removal) followed by optimal extraction.
- Generation of white-light and spectroscopic light curves (binned into ~5 nm channels).
- Correction for common-mode systematics.
Light Curve Modeling:
- Used batman for transit modeling and Gaussian Processes (GP) via george to model correlated noise (systematics) using state vectors (time, position, seeing, rotation).
- Joint fitting of white-light curves to derive precise orbital parameters ( $i$ , $a/R_*$ , $R_p/R_*$ ).
Atmospheric Retrieval:
- Used petitRADTRANS for forward modeling and PyMultiNest for Bayesian spectral retrieval.
- Models Tested:
  1. Flat-line (no atmosphere).
  2. Pure atmospheric models (Equilibrium and Free chemistry).
  3. Pure stellar contamination models (Spots/Faculae).
  4. Hybrid Models: Combining atmospheric physics with epoch-dependent stellar contamination (unocculted spots).
- Chemistry: Tested both chemical equilibrium (parameterized by metallicity $Z$ and C/O ratio) and free chemistry (independent abundances for 15 species).

3. Key Contributions

Quantification of Stellar Contamination: The study provides strong evidence that the discrepancies between the two transit spectra (observed ~1 year apart) are driven by varying levels of unocculted stellar spots, rather than intrinsic atmospheric changes.
Hybrid Retrieval Framework: Successfully demonstrated the necessity of jointly modeling atmospheric features and stellar contamination to avoid biased atmospheric inferences.
Optical Constraints on a Warm Sub-Saturn: Provided the first optical transmission spectrum for NGTS-5b, constraining its atmospheric properties despite the limitations of optical wavelengths compared to infrared (JWST).

4. Key Results

A. Stellar Activity and Contamination

Evidence: Bayesian model comparison strongly favors hybrid models (Atmosphere + Spots) over pure atmospheric models ( $\Delta \ln Z > 15$ ).
Epoch Variability:
- Night 1: Characterized by higher stellar activity, with a spot coverage fraction ( $f_{spot}$ ) of 20–29% and a large temperature contrast ( $\Delta T_{spot} \approx -1000$ to $-1200$ K).
- Night 2: Lower activity, with $f_{spot} \approx 4–23\%$ and a smaller temperature contrast.
Impact: Correcting for this contamination significantly reduced the $\chi^2$ discrepancy between the two nights (from 103.03 to 73.18).

B. Atmospheric Properties

After correcting for stellar contamination, the atmosphere of NGTS-5b is characterized as relatively clear (low cloud/haze coverage) with significant water absorption.

Water Abundance:
- Free Chemistry Retrieval: Strong evidence for H $_2$ O presence ( $\Delta \ln Z = 5.91$ ).
- Abundance: $\log X_{H_2O} = -0.79^{+0.14}_{-0.17}$ (volume mixing ratio $\approx 10^{-1.6}$ ).
- This is comparable to values found in other warm sub-Saturns observed by JWST (e.g., WASP-80b, WASP-107b).
Metallicity and C/O Ratio (Equilibrium Chemistry):
- Metallicity: Sub-solar, $0.10^{+0.34}_{-0.05} \times$ Solar.
- C/O Ratio: Sub-solar, $< 0.22$ (90% upper limit).
- Note: The authors caution that these values may be biased due to the limited optical wavelength coverage, which lacks strong carbon-bearing species features.
Temperature:
- Retrieved isothermal temperatures ( $T_{iso} \approx 600–700$ K) are significantly lower than the equilibrium temperature ($952$ K). This is a common feature in transmission spectroscopy, potentially due to terminator limb asymmetry or 1D model limitations.
Other Species:
- Potassium (K) and Iron Hydride (FeH) were tentatively detected in the free chemistry model, but their constraints are less robust than H $_2$ O.
- Other species (CH $_4$ , CO, CO $_2$ , etc.) remain unconstrained due to the lack of infrared data.

5. Significance

Validation of Stellar Contamination Models: The paper reinforces that ignoring stellar activity can lead to incorrect conclusions about planetary atmospheres. The "clear sky" nature of NGTS-5b was only revealed after rigorously accounting for the "dirty" stellar surface.
Atmospheric Diversity: The high water abundance in NGTS-5b contrasts with other similar-mass planets (like HAT-P-18b) that show low water, suggesting significant diversity in the formation and migration histories of warm sub-Saturns.
Future Directions: The study highlights the critical need for infrared observations (JWST) to detect carbon and oxygen-bearing species to break degeneracies in metallicity and C/O ratios. It also underscores the value of blue-optical ground-based observations (like GTC/OSIRIS) for characterizing stellar contamination, which is most pronounced at these wavelengths.

In conclusion, NGTS-5b hosts a relatively clear, water-rich atmosphere, but precise characterization of its bulk composition requires multi-wavelength follow-up to disentangle planetary signals from stellar noise.