Panopticon: a novel deep learning model to detect single transit events with no prior data filtering in PLATO light curves
Aix Marseille Univ（艾克斯-马赛大学）；CNRS（法国国家科学研究中心）；CNES（法国国家空间研究中心）；Institut Origines（起源研究所）；LAM（马赛天体物理实验室）；Institute for Astronomy, KU Leuven（鲁汶大学天文学研究所）；AIM（AIM研究所）；CEA（法国原子能和替代能源委员会）；Université Paris-Saclay（巴黎-萨克雷大学）；Université Paris Diderot（巴黎狄德罗大学）；Sorbonne Paris Cité（巴黎城市大学）
H. G. Vivien, M. Deleuil, N. Jannsen, J. De Ridder, D. Seynaeve, M. -A. Carpine, Y. Zerah
To prepare for the analyses of the future PLATO light curves, we develop a deep learning model, Panopticon, to detect transits in high precision photometric light curves. Since PLATO's main objective is the detection of temperate Earth-size planets around solar-type stars, the code is designed to detect individual transit events. The filtering step, required by conventional detection methods, can affect the transit, which could be an issue for long and shallow transits. To protect transit shape and depth, the code is also designed to work on unfiltered light curves. We trained the model on a set of simulated PLATO light curves in which we injected, at pixel level, either planetary, eclipsing binary, or background eclipsing binary signals. We also include a variety of noises in our data, such as granulation, stellar spots or cosmic rays. The approach is able to recover 90% of our test population, including more than 25% of the Earth-analogs, even in the unfiltered light curves. The model also recovers the transits irrespective of the orbital period, and is able to retrieve transits on a unique event basis. These figures are obtained when accepting a false alarm rate of 1%. When keeping the false alarm rate low (<0.01%), it is still able to recover more than 85% of the transit signals. Any transit deeper than 180ppm is essentially guaranteed to be recovered. This method is able to recover transits on a unique event basis, and does so with a low false alarm rate. Thanks to light curves being one-dimensional, model training is fast, on the order of a few hours per model. This speed in training and inference, coupled to the recovery effectiveness and precision of the model make it an ideal tool to complement, or be used ahead of, classical approaches.
In its long-duration observation phase, the PLATO satellite will observe two non-overlapping fields for a total of 4 yr. The exact duration of each pointing will be determined 2 yr before launch. Previous estimates of PLATO's yield of Earth-sized planets in the habitable zones (HZs) around solar-type stars ranged between 6 and 280. We use the PLATO Solar-like Light curve Simulator (PSLS) to simulate light curves with transiting planets around bright (m_V > 11) Sun-like stars at a cadence of 25 s, roughly representative of the >15,000 targets in PLATO's high-priority P1 sample (mostly F5-K7 dwa…
Performance of AutoRegressive Planet Search methodology on simulated PLATO lightcurves
Penn State University（宾夕法尼亚州立大学）；Università di Padova（帕多瓦大学）；INAF, Osservatorio Astronomico di Padova（意大利国家天体物理研究所，帕多瓦天文台）；INAF, Osservatorio Astrofisico di Catania（意大利国家天体物理研究所，卡塔尼亚天文台）；University of Oxford（牛津大学）
Recently, we introduced PLanetary Atmospheric Tool for Observer Noobs (PLATON), a Python package that calculates model transmission spectra for exoplanets and retrieves atmospheric characteristics based on observed spectra. We now expand its capabilities to include the ability to compute secondary eclipse depths. We have also added the option to calculate models using the correlated-$k$ method for radiative transfer, which improves accuracy without sacrificing speed. Additionally, we update the opacities in PLATON--many of which were generated using old or proprietary line lists--using the mos…
Detecting transiting exoplanets in simulated PLATO data: A comparison of light curve filter and transit search algorithms
University of Oxford（牛津大学）；University of Cambridge（剑桥大学）；Università di Padova（帕多瓦大学）；INAF - Osservatorio Astronomico di Padova（意大利国家天体物理研究所帕多瓦天文台）；Center for Computational Astrophysics, Flatiron Institute（平顿研究所计算天体物理学中心）
One of the main science motivations for the ESA PLAnetary Transit and Oscillations (PLATO) mission is to measure exoplanet transit radii with 3% precision. In addition to flares and starspots, stellar oscillations and granulation will enforce fundamental noise floors for transiting exoplanet radius measurements. We simulate light curves of Earth-sized exoplanets transiting continuum intensity images of the Sun taken by the HMI instrument aboard SDO to investigate the uncertainties introduced on the exoplanet radius measurements by stellar granulation and oscillations. After modeling the solar …
Following up the Kepler field with PLATO: Transit Timing Performance
University of Warwick（华威大学）；Centre for Exoplanets and Habitability, University of Warwick（华威大学系外行星与宜居性中心）；Centre for Space Domain Awareness, University of Warwick（华威大学太空领域感知中心）；LIRA, Observatoire de Paris, Université PSL, Sorbonne Université, Université Paris Cité, CY Cergy Paris Université, CNRS（巴黎天文台、PSL大学、索邦大学、巴黎西岱大学、CY塞尔吉巴黎大学、法国国家科学研究中心LIRA实验室）；Institute for Astronomy, KU Leuven（荷语鲁汶大学天文学研究所）；Isaac Newton Group of Telescopes（艾萨克·牛顿望远镜组）
We assess broadband color filters for the two fast cameras on the PLAnetary Transits and Oscillations (PLATO) of stars space mission with respect to exoplanetary atmospheric characterization. We focus on Ultra Hot Jupiters and Hot Jupiters placed 25pc and 100pc away from the Earth and low mass low density planets placed 10pc and 25pc away. Our analysis takes as input literature values for the difference in transit depth between the broadband lower (500 to 675nm) wavelength interval (hereafter referred to as blue) and the upper (675-1125nm) broadband wavelength interval (hereafter referred to a…
Detecting false positives with PLATO using double-aperture photometry and centroid shifts
LIRA, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, Université Paris Diderot, Sorbonne Paris Cité（巴黎天文台）；Max Planck Institute for Solar System Research（马克斯·普朗克太阳系统研究所）；Deutsches Zentrum für Luft- und Raumfahrt（德国航空航天中心）；Aix Marseille Univ, CNRS, CNES, LAM（艾克斯-马赛大学）；Univ. Grenoble Alpes, CNRS, IPAG（格勒诺布尔阿尔卑斯大学）
The space mission PLATO will usher in a new era of exoplanetary science by expanding our current inventory of transiting systems and constraining host star ages, which are currently highly uncertain. This capability might allow PLATO to detect changes in planetary system architecture with time, particularly because planetary scattering due to Lagrange instability may be triggered long after the system was formed. Here, we utilize previously published instability timescale prescriptions to determine PLATO's capability to detect a trend of decreasing planet frequency with age for systems with eq…
Astrophysical false positives that mimic planetary transit are one of the main limitation to exoplanet transit surveys. In this proceeding, we review the issue of the false positive in transit survey and the possible complementary observations to constrain their presence. We also review the false-positive rate of both Kepler and CoRoT missions and present the basics of the planet-validation technique. Finally, we discuss the interest of observing bright stars, as PLATO 2.0 and TESS will do, in the context of the false positives. According to simulations with the Besançon galactic model, we fin…
For the upcoming PLAnetary Transits and Oscillation of stars (PLATO) satellite mission, a large number of target stars are required to yield a statistically significant number of planet transits. Locating the centres of the long duration observational phase (LOP) fields closer to the Galactic plane will increase the target star numbers but also the astrophysical false positives (FPs) from blended eclipsing binary systems. We utilise the Binary Stellar Evolution and Population Synthesis (BiSEPS) code, to create a complete synthetic stellar and planetary population for the proposed southern LOP …
Time-series transit photometry from the Kepler space telescope has allowed for the discovery of thousands of exoplanets. We explore the potential of yet improved future missions such as PLATO 2.0 in detecting solar system analogues. We use real-world solar data and end-to-end simulations to explore the stellar and instrumental noise properties. By injecting and retrieving planets, rings and moons of our own solar system, we show that the discovery of Venus- and Earth-analogues transiting G-dwarfs like our Sun is feasible at high S/N after collecting 6yrs of data, but Mars and Mercury will be d…
Variability classification of TESS targets in LOPS2, the first long-term pointing field of PLATO. Version 1 of the public variability catalogue
KU Leuven（荷语鲁汶大学）；MIT Kavli Institute for Astrophysics & Space Research, Massachusetts Institute of Technology（麻省理工学院卡弗里天体物理与空间研究所）；Astronomical Institute of the Czech Academy of Sciences（捷克科学院天文研究所）；Newcastle University（纽卡斯尔大学）；University of Southern Queensland（南昆士兰大学）
While the solar system contains about 20 times more moons than planets, no moon has been confirmed around any of the thousands of extrasolar planets known so far. Tools for an uncomplicated identification of the most promising exomoon candidates could be beneficial to streamline follow-up studies.} Here we study three exomoon indicators that emerge if well-established planet-only models are fitted to a planet-moon transit light curve: transit timing variations (TTVs), transit duration variations (TDVs), and apparent planetary transit radius variations (TRVs). We re-evaluate under realistic con…
The increasing number of transiting exoplanets sparked a significant interest in discovering their moons. Most of the methods in the literature utilize timing analysis of the raw light curves. Here we propose a new approach for the direct detection of a moon in the transit light curves via the so called Scatter Peak. The essence of the method is the valuation of the local scatter in the folded light curves of many transits. We test the ability of this method with different simulations: Kepler "short cadence", Kepler "long cadence", ground-based millimagnitude photometry with 3-min cadence, and…
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 …
Short-period planets suffer from extreme tidal forces from their parent stars causing them to deform and attain non-spherical shapes. The planet shapes, modeled here as triaxial ellipsoids, can have an impact on the observed transit light curves and the planetary parameters derived. We investigate the detectability of tidal deformation in short-period planets from their transit light curves and the instrumental precision needed. We show how the detection of deformation from the light curve allows us to obtain an observational estimate of the second fluid Love number which gives valuable insigh…
Rapid planetary rotation can cause the equilibrium shape of a planet to be oblate. While planetary oblateness has mostly been probed by examining the subtle ingress and egress features in photometric transit light curves, we investigate the effect of oblateness on the spectroscopic Rossiter-McLaughlin (RM) signals. We found that a giant planet, with planet-to-star radius ratio of 0.15 and Saturn-like oblateness of 0.098, can cause spectroscopic signatures with amplitudes up to 1.1 ms$^{-1}$ which is detectable by high-precision spectrographs such as ESPRESSO. We also found that the spectroscop…
We present Pandora, a new software to model, detect, and characterize transits of extrasolar planets with moons in stellar photometric time series. Pandora uses an analytical description of the transit light curve for both the planet and the moon in front of a star with atmospheric limb darkening and it covers all cases of mutual planet-moon eclipses during transit. The orbital motion of the star-planet-moon system is computed with a high accuracy as a nested Keplerian problem. We have optimized Pandora for computational speed to make it suitable for large-scale exomoon searches in the new era…
The GPU Phase Folding and Deep Learning Method for Detecting Exoplanet Transits
The Harker School（哈克学校）；Science Talent Training Center（科学人才培养中心）；Shanghai Astronomical Observatory（上海天文台）；Princeton University（普林斯顿大学）；University of Geneva（日内瓦大学）
Photometric surveys such as Kepler have the precision to identify exoplanet and eclipsing binary candidates from only a single transit. K2, with its 75d campaign duration, is ideally suited to detect significant numbers of single-eclipsing objects. Here we develop a Bayesian transit-fitting tool ("Namaste: An Mcmc Analysis of Single Transit Exoplanets") to extract orbital information from single transit events. We achieve favourable results testing this technique on known Kepler planets, and apply the technique to 7 candidates identified from a targeted search of K2 campaigns 1, 2 and 3. We fi…
The orbital sampling effect (OSE) appears in phase-folded transit light curves of extrasolar planets with moons. Analytical OSE models have hitherto neglected stellar limb darkening and non-zero transit impact parameters and assumed that the moon is on a circular, co-planar orbit around the planet. Here, we present an analytical OSE model for eccentric moon orbits, which we implement in a numerical simulator with stellar limb darkening that allows for arbitrary transit impact parameters. We also describe and publicly release a fully numerical OSE simulator (PyOSE) that can model arbitrary incl…
We present a new method to detect planetary transits from time-series photometry, the Transit Least Squares (TLS) algorithm. TLS searches for transit-like features while taking the stellar limb darkening and planetary ingress and egress into account. We have optimized TLS for both signal detection efficiency (SDE) of small planets and computational speed. TLS analyses the entire, unbinned phase-folded light curve. We compensate for the higher computational load by (i.) using algorithms like "Mergesort" (for the trial orbital phases) and by (ii.) restricting the trial transit durations to a sma…
Université de Genève（日内瓦大学）；Monash University（莫纳什大学）；Disaitek；Universität Bern（伯尔尼大学）
Transit Timing Variations (TTVs) can provide useful information for systems observed by transit, by putting constraints on the masses and eccentricities of the observed planets, or even constrain the existence of non-transiting companions. However, TTVs can also prevent the detection of small planets in transit surveys, or bias the recovered planetary and transit parameters. Here we show that Kepler-1972 c, initially the "not transit-like" false positive KOI-3184.02, is an Earth-sized planet whose orbit is perturbed by Kepler-1972 b (initially KOI-3184.01). The pair is locked in a 3:2 Mean-mot…
We present the Signal Detection using Random-Forest Algorithm (SIDRA). SIDRA is a detection and classification algorithm based on the Machine Learning technique (Random Forest). The goal of this paper is to show the power of SIDRA for quick and accurate signal detection and classification. We first diagnose the power of the method with simulated light curves and try it on a subset of the Kepler space mission catalogue. We use five classes of simulated light curves (CONSTANT, TRANSIT, VARIABLE, MLENS and EB for constant light curves, transiting exoplanet, variable, microlensing events and eclip…
Context. Transit detection algorithms are mathematical tools used for detecting planets in the photometric data of transit surveys. In this work we study their application to space-based surveys. Aims: Space missions are exploring the parameter space of the transit surveys where classical algorithms do not perform optimally, either because of the challenging signal-to-noise ratio of the signal or its non-periodic characteristics. We have developed an algorithm addressing these challenges for the mission CoRoT. Here we extend the application to the data from the space mission Kepler. We aim at …
Space-based missions such as Kepler, and soon TESS, provide large datasets that must be analyzed efficiently and systematically. Recent work by Shallue & Vanderburg (2018) successfully used state-of-the-art deep learning models to automatically classify Kepler transit signals as either exoplanets or false positives; our application of their model yielded 95.8% accuracy and 95.5% average precision. Here we expand upon that work by including additional scientific domain knowledge into the network architecture and input representations to significantly increase overall model performance to 97.5% …
We investigate in this paper the astrophysical false-positive configuration in exoplanet-transit surveys that involves eclipsing binaries and giant planets which present only a secondary eclipse, as seen from the Earth. To test how an eclipsing binary configuration can mimic a planetary transit, we generate synthetic light curve of three examples of secondary-only eclipsing binary systems that we fit with a circular planetary model. Then, to evaluate its occurrence we model a population of binaries in double and triple system based on binary statistics and occurrence. We find that 0.061% +/- 0…
State of the art exoplanet transit surveys are producing ever increasing quantities of data. To make the best use of this resource, in detecting interesting planetary systems or in determining accurate planetary population statistics, requires new automated methods. Here we describe a machine learning algorithm that forms an integral part of the pipeline for the NGTS transit survey, demonstrating the efficacy of machine learning in selecting planetary candidates from multi-night ground based survey data. Our method uses a combination of random forests and self-organising-maps to rank planetary…
Current and future space-based observatories such as the Transiting Exoplanet Survey Satellite (TESS) and PLATO are set to provide an enormous amount of new data on oscillating stars, and in particular stars that oscillate similar to the Sun. Solar-like oscillators constitute the majority of known oscillating stars and so automated analysis methods are becoming an ever increasing necessity to make as much use of these data as possible. Here we aim to construct an algorithm that can automatically determine if a given time series of photometric measurements shows evidence of solar-like oscillati…
A large fraction of the smallest transiting planet candidates discovered by the Kepler and CoRoT space missions cannot be confirmed by a dynamical measurement of the mass using currently available observing facilities. To establish their planetary nature, the concept of planet validation has been advanced. This technique compares the probability of the planetary hypothesis against that of all reasonably conceivable alternative false-positive (FP) hypotheses. The candidate is considered as validated if the posterior probability of the planetary hypothesis is sufficiently larger than the sum of …
Inaccurate limb-darkening models can be a significant source of error in the analysis of the light curves for transiting exoplanet and eclipsing binary star systems. To test the accuracy of published limb-darkening models, I have compared limb-darkening profiles predicted by stellar atmosphere models to the limb-darkening profiles measured from high-quality light curves of 43 FGK-type stars in transiting exoplanet systems observed by the Kepler and TESS missions. The comparison is done using the parameters $h^{\prime}_1 = I_λ(\frac{2}{3})$ and $h^{\prime}_2 = h^{\prime}_1 - I_λ(\frac{1}{3})$, …
The CoRoT mission during its flight-phase 2007-2012 delivered the light-curves for over 2000 eclipsing binaries. Data from the Kepler mission have proven the existence of several transiting circumbinary planets. Albeit light-curves from CoRoT have typically lower precision and shorter coverage, CoRoT's number of targets is similar to Kepler, and some of the known circumbinary planets could potentially be detected in CoRoT data as well. The aim of this work has been a revision of the entire CoRoT data-set for the presence of circumbinary planets, and the derivation of limits to the abundances o…
DELOS: Contrastive Deep Learning for Low-SNR Blind Transit Searches in Kepler Photometry
Shanghai Astronomical Observatory, Chinese Academy of Sciences, Shanghai 200030, China（上海天文台，中国科学院，上海200030，中国）；School of Astronomy and Space Sciences, University of Chinese Academy of Sciences, Beijing 101408, China（中国科学院大学天文与空间科学学院，北京101408，中国）；Science Talent Training Center, Gainesville, FL, 32606 USA（科学人才培训中心，佛罗里达州盖恩斯维尔，32606 USA）
One of the strongest ${\rm Na~I}$ features was observed in WASP-96b. To confirm this novel detection, we provide a new 475-825nm transmission spectrum obtained with Magellan/IMACS, which indeed confirms the presence of a broad sodium absorption feature. We find the same result when reanalyzing the 400-825nm VLT/FORS2 data. We also utilize synthetic data to test the effectiveness of two common detrending techniques: (1) a Gaussian processes (GP) routine, and (2) common-mode correction followed by polynomial correction (CMC+Poly). We find that both methods poorly reproduce the absolute transit d…
We present K2SC (K2 Systematics Correction), a Python pipeline to model instrumental systematics and astrophysical variability in light curves from the K2 mission. K2SC uses Gaussian process regression to model position-dependent systematics and time-dependent variability simultaneously, enabling the user to remove both (e.g., for transit searches) or to remove systematics while preserving variability (for variability studies). For periodic variables, K2SC automatically computes estimates of the period, amplitude and evolution timescale of the variability. We apply K2SC to publicly available K…
We report transmission spectroscopy of the bloated hot Jupiter WASP-74b in the wavelength range from 4000 to 6200 Å. We observe two transit events with the Very Large Telescope FOcal Reducer and Spectrograph (VLT FORS2) and present a new method to measure the exoplanet transit depth as a function of wavelength. The new method removes the need for a reference star in correcting the spectroscopic light curves for the impact of atmospheric extinction. It also provides improved precision, compared to other techniques, reaching an average transit depth uncertainty of 211 ppm for a solar-type star o…
TIC65910228b: A single-transit discovery of a massive long-period warm Jupiter with TESS
Pontificia Universidad Católica de Chile（智利天主教 Pontificia 大学）；Universidad Adolfo Ibáñez（阿道夫·伊巴涅斯大学）；Millennium Institute for Astrophysics（天体物理学千年研究所）；European Southern Observatory (ESO)（欧洲南方天文台）；Astronomical Institute of the Czech Academy of Sciences（捷克科学院天文研究所）；Astronomical Institute of Charles University（查理大学天文研究所）；Max-Planck-Institut für Astronomie（马克斯·普朗克天体物理研究所）；Space Telescope Science Institute（太空望远镜科学研究所）
Data from transit light curves, radial velocity and transit timing observations can be used to probe the interiors of exoplanets beyond the mean density, by measuring the Love numbers $h_2$ and $k_2$. The first indirect estimate of $k_2$ for an exoplanet from radial velocity and transit timing variations observations has been performed by taking advantage of the years-spanning baseline. Not a single measurement of $h_2$ has been achieved from transit light curves, mostly because the photometric precision of current observing facilities is still too low. We show that the Imaging Spectrograph in…
