Exoplanets, -moons, -comets

Modeling Doppler Shifts In Radial-Velocity Data With Deep Learning Toward Earth-mass Exoplanet Detection

By Keith Cowing

Status Report

astro-ph.IM

June 19, 2026

Filed under astro-ph.IM, astronomy, Doppler-shift, Earth-sized planet, exoplanet, imaging, Spectroscopy

Modeling Doppler Shifts In Radial-Velocity Data With Deep Learning Toward Earth-mass Exoplanet Detection — Shell representations for flux and temperature at the time of a maximum DS of 20 m/s. The masked shell is defined as the element-wise product between the spectral shell and the associated density-weight map. In the shell construction, a cell is assigned zero only when no spectral pixels populate the corresponding bin. Although the same selected spectral pixels are used for both the flux- and temperature-based shells, they are projected onto different shell spaces, (F, ∂F/∂v) and (T, ∂T/∂v), so the occupancy of the shell cells is not identical between the two representations. Top row: Flux representations, with the first two panels unmasked and the last two masked. Bottom row: Same configuration for the temperature based representation. — astro-ph.IM

Detecting the tiny Doppler shifts induced by Earth-mass planets in stellar radial-velocity measurements remains extremely challenging due to stellar activity. Many deep-learning methods performing well on simulated data remain difficult to apply reliably on real stellar spectra.

The aim of this work is to develop a deep-learning framework that generalizes to real, unseen spectra and improves the detectability of Earth-mass planets in radial-velocity data. We train artificial neural networks on HARPS-N solar spectra with injected planetary signals, using physics-motivated spectral representations based on flux and line-formation temperature, together with their velocity gradients.

Two training strategies are explored: hold-out testing and cross-validation. Model robustness is enhanced through genetic-algorithm-based hyperparameter optimization, and predictive uncertainty is quantified using Monte Carlo dropout.

Our most precise neural network model reliably retrieves, under the cross-validation strategy, the amplitudes, phases, and orbital periods of planetary signals with amplitudes greater than or equal to 25 cm/s and periods between 10 and 550 days.

In addition, in all cases tested here, the successfully recovered signals correspond to the most significant peaks in the periodograms of the Doppler-shift predictions. Temperature-based spectral-shell representations consistently outperform flux-based shells. We also release doppleriann, a Python package implementing the proposed framework.

Our results demonstrate that combining physically motivated spectral representations with deep learning provides a promising pathway toward the detection of Earth-mass planets in radial-velocity data from real observations, supported by a modeling framework that is both physically grounded and statistically rigorous, incorporating uncertainty quantification and optimized training strategies.

Isidro Gómez-Vargas, Xavier Dumusque, Yinan Zhao, Khaled Al Moulla, Michael Cretignier

Comments: 20 pages, 14 figures. Accepted for publication in Astronomy & Astrophysics
Subjects: Instrumentation and Methods for Astrophysics (astro-ph.IM); Earth and Planetary Astrophysics (astro-ph.EP); Machine Learning (cs.LG)
Cite as: arXiv:2606.18464 [astro-ph.IM] (or arXiv:2606.18464v1 [astro-ph.IM] for this version)
https://doi.org/10.48550/arXiv.2606.18464
Focus to learn more
Submission history
From: Isidro Gómez-Vargas
[v1] Tue, 16 Jun 2026 20:16:16 UTC (5,164 KB)
https://arxiv.org/abs/2606.18464
Astrobiology,

Keith Cowing

Biologist, Explorers Club Fellow, ex-NASA Space Biologist and Payload integrator, Editor of NASAWatch.com and Astrobiology.com, Lapsed climber, Explorer, Synaesthete, Former Challenger Center board member 🖖🏻

Follow on Twitter