arXiv ScienceSearch

arXiv · 2102.10326

Automated identification of transiting exoplanet candidates in NASA Transiting Exoplanets Survey Satellite (TESS) data with machine learning methods

Abstract

A novel artificial intelligence (AI) technique that uses machine learning (ML) methodologies combines several algorithms, which were developed by ThetaRay, Inc., is applied to NASA's Transiting Exoplanets Survey Satellite (TESS) dataset to identify exoplanetary candidates. The AI/ML ThetaRay system is trained initially with Kepler exoplanetary data and validated with confirmed exoplanets before its application to TESS data. Existing and new features of the data, based on various observational parameters, are constructed and used in the AI/ML analysis by employing semi-supervised and unsupervised machine learning techniques. By the application of ThetaRay system to 10,803 light curves of threshold crossing events (TCEs) produced by the TESS mission, obtained from the Mikulski Archive for Space Telescopes, the algorithm yields about 50 targets for further analysis, and we uncover three new exoplanetary candidates by further manual vetting. This study demonstrates for the first time the successful application of the particular combined multiple AI/ML-based methodologies to a large astrophysical dataset for rapid automated classification of TCEs.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Leon Ofman, Amir Averbuch, Adi Shliselberg, Idan Benaun, David Segev, Aron Rissman. 2021-08-27. Automated identification of transiting exoplanet candidates in NASA Transiting Exoplanets Survey Satellite (TESS) data with machine learning methods. https://doi.org/10.1016/j.newast.2021.101693

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

Volcanic Satellites and Ion Escape in the Magnetospheres of Ultra-Cool and Brown Dwarf Stars

Radio emissions at $\sim$ GHz frequencies of ultra-cool dwarf and brown dwarf stars suggest the presence of radiation belts not unlike Jupiter's. We investigate the possibility the inferred magnetospheric plasma at the primary star is sourced by an active planet or via ion escape modeled as a weak ionospheric outflow. We consider indirect methods to estimate the magnetospheric plasma mass flow from auroral radio emission and apply them to the ultra-cool dwarf LSR J1835+3259. We find that an ionospheric outflow is a viable source of the order of $10^{5}$ kg s$^{-1}$, if the ionospheric effective Pedersen conductance is lower than $\sim 0.02$ mho. On the other hand, a tidally-heated volcanic satellite with the same mass and radius as Io and an orbit with a semimajor axis lower than about $10$ stellar radii, whose eccentricity ($e \sim 10^{-3}$) is maintained by perturbations by other planets in the system, is found to be a viable source of magnetospheric plasma without strong limitations on the ionospheric conductance. The volcanic satellite scenario is also naturally in sync with the recent finding that ultra-cool dwarfs with distant substellar or stellar companions are remarkably more likely to be detected as radio emitters.

astro-ph.EP

Tidal Demise: The Evolution and Fate of a Hypothetical Venus Moon

Venus possesses no natural satellite, raising the question whether a formed moon could have survived. We explore the tidal evolution of a Venus-moon system, coupling Venus's spin to the satellite's orbit under tides from the moon and Sun. We survey spin period ($P_0 = 5$--100~hr), moon mass ($M_m = 0.01$--$10~M_{\rm Moon}$), eccentricity, quality factor, and initial semi-major axis under both constant-$Q$ and constant time lag models. Survival depends on competition between outward migration ($\propto M_m$) and synchronous radius expansion ($\propto M_m^2$): for circular orbits around rapidly spinning Venus ($P_0 \lesssim 12$~hr), a lunar-mass satellite survives the age of the Solar System in both models. For $P_0 \lesssim 10$~hr, eccentricity pumping can destabilize low-mass satellites, while for $P_0 \gtrsim 15$~hr or $M_m \gtrsim 2~M_{\rm Moon}$ the synchronous radius overtakes the orbit and drives Roche destruction within $\sim$0.03--1.7~Gyr in the constant-$Q$ model. The constant time lag model instead permits quasi-synchronous survival for massive moons at fast spin. Explaining Venus's present state requires satisfying two constraints simultaneously: loss of the satellite and despinning of an initially rapid rotator. Both are met only within a restricted region of parameter space, favoring moderate post-impact spin periods and lunar-to-super-lunar masses. Giant impact simulations predict spin periods $\gtrsim$12~hr for Venus's present rotation, placing a lunar-mass satellite at the survival boundary. For last-impact conditions within this region, the present absence of a Venusian satellite arises through tidal evolution alone; a subsequent catastrophic stripping event, while capable of removing a moon, is not required.

astro-ph.EP

Evidence for an Extended Hydrogen Outflow on WASP-12 b

Recent observations of atmospheric escape from planets orbiting early-type stars indicate that planets with higher Roche filling factors have significantly higher mass loss rates. For three of these planets, full-orbit observations have revealed the presence of large leading and trailing tails of escaping planetary material. These 3D outflow geometries encode information about the underlying outflow physics, and can also be used to constrain their stellar wind properties and to predict their long-term orbital evolution. We present new evidence for extended H$α$ and H$β$ absorption from escaping hydrogen for a fourth planet, WASP-12 b. We observed WASP-12 b with the Keck Planet Finder on Keck I over a period of approximately eight hours centered on the transit. We find a H$α$ and H$β$ absorption signal in the stellar rest frame that is strongest during the transit (H$α$ amplitude: $2.66\pm0.25\%$, H$β$ amplitude: $2.88\pm0.43\%$), with evidence of both pre- and post-transit absorption. This measurement is consistent with outflow models where the escaping gas overflows the planet's Roche lobe and is advected into the stellar rest frame before being tidally sheared into extended tails. This outflow behavior is qualitatively similar to previous H$α$ observations of this planet as well as observations of extended outflows from other gas giants orbiting early-type stars.

astro-ph.EP