Resolved stellar disk
Cell-by-cell integration of PHOENIX photosphere and spot intensities with center-to-limb variation in the stellar scenario. Stellar rotation and the Rossiter–McLaughlin effect are excluded from this study.
Research software
Starspot
Controlled simulations of stellar activity and planetary atmospheric signals in exoplanet transmission spectroscopy.
Scientific Question
Can a cool starspot produce a spectral signal that resembles absorption by a planetary atmosphere? Starspot creates controlled experiments in which the physical origin of every signal is known.
TOI-5205b Case Study
Two isolated scenarios share orbital parameters, an H2O template and instrumental processing: a planet with an H2O atmosphere transiting a uniform star, and a constant-radius planet transiting a 3400 K photosphere with an unocculted 2500 K spot. The stellar scenario includes center-to-limb variation.
Simulation framework
From the stellar disk to the Doppler map
Starspot generates time-resolved synthetic spectra from a defined physical configuration, preserving every input and output needed to reproduce a run.
Physical spectra
PHOENIX photosphere and spot models, wavelength-dependent planetary radius, molecular templates and instrumental resolution.
Reproducible analysis
VIS/NIR spectral orders, controlled noise regimes, cross-correlation functions and stellar- versus planetary-frame diagnostics.
Spectral comparison
Similar bands do not imply the same origin
At low resolution, the atmospheric and stellar scenarios share broad regions of variation but are not equivalent. After removing their median levels, their spectral shapes have a Pearson correlation of r = 0.269 over the full interval. Their infrared shapes are much more similar: r = 0.920 for λ > 1.3 μm and r = 0.914 for λ > 1.5 μm. These correlations describe shape similarity, not equal amplitudes or statistical indistinguishability.
Doppler kinematics
Reference frames distinguish the signals when noise is low
In these controlled simulations, the planetary atmosphere follows the orbital velocity and becomes centered near 0 km s−1 in the planetary frame. Starspot contamination remains tied to the star and acquires the opposite trajectory after the same transformation.
Current results
Discrimination depends on signal quality
The internal diagnostic z below measures peak contrast relative to the CCF background. Values are the median ± standard deviation across ten noise realizations. This is a comparative recovery metric, not a formal statistical significance.
| Scenario | Channel | CARMENES | E-MARCOT |
|---|---|---|---|
| Planetary atmosphere | VIS | 1.30 ± 0.72 | 3.64 ± 0.93 |
| Planetary atmosphere | NIR | 2.31 ± 1.08 | 6.79 ± 1.14 |
| 2500 K spot | VIS | 1.01 ± 0.53 | 0.98 ± 0.52 |
| 2500 K spot | NIR | 1.28 ± 0.62 | 3.10 ± 0.78 |
The different Doppler paths are recovered clearly.
One simulated transit does not recover a robust separation.
An idealized noise scaling of 3.5/15 improves recovery, particularly for the planetary NIR signal. Coverage, resolution and observing sequence remain those adopted for CARMENES; this is not a complete instrumental performance prediction.
The atmospheric and stellar contributions are simulated separately; recovery with both present simultaneously remains future work. The model omits telluric absorption and observational systematics. These results are not an observational detection of H2O in TOI-5205b.
A strong response to an H2O template measures spectral similarity; by itself, it does not establish a planetary origin or an atmospheric detection.
Software Status
This controlled study combines Starspot v6.1 PH and Starspot v6.1 H2O, with modular Python components, numerical validation and traceable run configurations. Reproduction requires the matching code, configurations and external scientific inputs; the PHOENIX models are stored separately.
Access
The current code package is available for selected academic review and scientific collaboration. It is not offered as a public download.
Research Context
Starspot and the TOI-5205b case study were developed by Gerardo Mancebo for his completed Master’s thesis in Astronomy and Astrophysics at Valencian International University (VIU), under the supervision of Dr. Giuseppe Morello.