Photometric search for exomoons by using convolutional neural networks

2021/08/01 by Lukas Weghs
Chemistry · Computer Science · Physics and Astronomy · #Artificial intelligence #Artificial neural network #Astro and Planetary Science #Astrobiology #Astronomy #Chemistry #Circumstellar habitable zone #Computer science #Convolutional neural network #Exoplanet #Gamma-ray bursts and supernovae #Kepler #Light curve #Oscillation (cell signaling) #Physics #Set (abstract data type) #Stars #Stellar, planetary, and galactic studies #astro-ph.EP #astro-ph.IM #cs.LG

paper · pdf · doi:10.1002/asna.202114007

published as Astronomische Nachrichten, 2021 · 11 pages, 4 figures

openalex publication_date 2021/08/01 · arxiv created 2021/11/03 · arxiv updated 2021/12/08 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

Abstract Until now, there is no confirmed moon beyond our solar system (exomoon). Exomoons offer us new possibly habitable places which might also be outside the classical habitable zone. But until now, the search for exomoons needs much computational power because classical statistical methods are employed. It is shown that exomoon signatures can be found by using deep learning and convolutional neural networks (CNNs), respectively, trained with synthetic light curves combined with real light curves with no transits. It is found that CNNs trained by combined synthetic and observed light curves may be used to find moons bigger or equal to roughly 2–3 Earth radii in the Kepler data set or comparable data sets. Using neural networks in future missions like Planetary Transits and Oscillation of stars might enable the detection of exomoons.

Citations

• Europa, tidally heated oceans, and habitable zones around giant planets
• THE ROCHE LIMIT FOR CLOSE-ORBITING PLANETS: MINIMUM DENSITY, COMPOSITION CONSTRAINTS, AND APPLICATION TO THE 4.2 hr PLANET KOI 1843.03
• Mass‐Radius Relationships for Solid Exoplanets
• Exomoon habitability constrained by illumination and tidal heating
• Gradient-based learning applied to document recognition
• ImageNet classification with deep convolutional neural networks
• Evidence of an Upper Bound on the Masses of Planets and Its Implications for Giant Planet Formation
• Evidence for a Large Exomoon Orbiting Kepler-1625b
• TensorFlow: A system for large-scale machine learning
• The subsurface habitability of small, icy exomoons