Machine learning for exoplanet detection in high-contrast spectroscopy Combining cross correlation maps and deep learning on medium-resolution integral-field spectra

Abstract

The advent of high-contrast imaging instruments combined with medium-resolution spectrographs allows spectral and temporal dimensions to be combined with spatial dimensions to detect and potentially characterize exoplanets with higher sensitivity. We develop a new method to effectively leverage the spectral and spatial dimensions in integral-field spectroscopy (IFS) datasets using a supervised deep-learning algorithm to improve the detection sensitivity to high-contrast exoplanets. We begin by applying a data transform whereby the IFS datasets are replaced by cross-correlation coefficient tensors obtained by cross-correlating our data with young gas giant spectral template spectra. This transformed data is then used to train machine learning (ML) algorithms. We train a 2D CNN and 3D LSTM with our data. We compare the ML models with a non-ML algorithm, based on the STIM map of arXiv:1810.06895. We test our algorithms on simulated young gas giants in a dataset that contains no known exoplanet, and explore the sensitivity of algorithms to detect these exoplanets at contrasts ranging from 1e-3 to 1e-4 at different radial separations. We quantify the sensitivity using modified receiver operating characteristic curves (mROC). We discover that the ML algorithms produce fewer false positives and have a higher true positive rate than the STIM-based algorithm, and the true positive rate of ML algorithms is less impacted by changing radial separation. We discover that the velocity dimension is an important differentiating factor. Through this paper, we demonstrate that ML techniques have the potential to improve the detection limits and reduce false positives for directly imaged planets in IFS datasets, after transforming the spectral dimension into a radial velocity dimension through a cross-correlation operation.

arXiv commentAccepted for publication in A&A on 23/04/2024. Total 15 pages of text, 7 figures

References in corpus (43)

• Evolutionary models for cool brown dwarfs and extrasolar giant planets. The case of HD 20945
• Angular Differential Imaging: a Powerful High-Contrast Imaging Technique
• The orbital motion, absolute mass, and high-altitude winds of exoplanet HD209458b
• The Gemini Planet Imager: First Light
• Discovery of a planetary-mass companion within the gap of the transition disk around PDS 70
• SPHERE: the exoplanet imager for the Very Large Telescope
• Discovery and spectroscopy of the young Jovian planet 51 Eri b with the Gemini Planet Imager
• Two accreting protoplanets around the young star PDS 70
• Detection of Carbon Monoxide and Water Absorption Lines in an Exoplanet Atmosphere
• Detection of water absorption in the day side atmosphere of HD 189733 b using ground-based high-resolution spectroscopy at 3.2 microns
• Fundamental limitations of high contrast imaging set by small sample statistics
• Imaging Spectroscopy for Extrasolar Planet Detection
• Retrieving Temperatures and Abundances of Exoplanet Atmospheres with High-Resolution Cross-Correlation Spectroscopy
• Orbital and atmospheric characterization of the planet within the gap of the PDS 70 transition disk
• Combining high-dispersion spectroscopy (HDS) with high contrast imaging (HCI): Probing rocky planets around our nearest neighbors
• Retrieving scattering clouds and disequilibrium chemistry in the atmosphere of HR 8799e
• VIP: Vortex Image Processing package for high-contrast direct imaging
• The Near-Infrared Spectrograph (NIRSpec) on the James Webb Space Telescope III. Integral-field spectroscopy
• Statistics of Long Period Gas Giant Planets in Known Planetary Systems
• Substructure in the Circumstellar Disk around the Young Star AU Mic (GJ 803)
• First light of the VLT planet finder SPHERE. IV. Physical and chemical properties of the planets around HR8799
• Surveying the reach and maturity of machine learning and artificial intelligence in astronomy
• Searching for Exoplanets Using Artificial Intelligence
• Medium-resolution integral-field spectroscopy for high-contrast exoplanet imaging: Molecule maps of the $\beta$ Pictoris system with SINFONI
• 1 to 2.4 micron Near-IR spectrum of the Giant Planet $\beta$ Pictoris b obtained with the Gemini Planet Imager
• Evidence for a circumplanetary disk around protoplanet PDS 70 b
• Radial Velocity Measurements of HR 8799 b and c with Medium Resolution Spectroscopy
• Direct Imaging and Astrometric Detection of a Gas Giant Planet Orbiting an Accelerating Star
• A Clear View of a Cloudy Brown Dwarf Companion from High-Resolution Spectroscopy
• In-depth study of moderately young but extremely red, very dusty substellar companion HD206893B
• Physical and orbital properties of Beta Pictoris b
• Interpreting High-Resolution Spectroscopy of Exoplanets Using Cross-Correlations and Supervised Machine Learning
• Retrieving the C and O Abundances of HR 7672~AB: a Solar-Type Primary Star with a Benchmark Brown Dwarf
• The Enhanced Resolution Imager and Spectrograph for the VLT
• Supervised detection of exoplanets in high-contrast imaging sequences
• STIM map: detection map for exoplanets imaging beyond asymptotic Gaussian residual speckle noise
• Molecular mapping of the PDS70 system: No molecular absorption signatures from the forming planet PDS70 b
• Direct emission spectroscopy of exoplanets with the medium resolution imaging spectrometer on board JWST MIRI: I. Molecular mapping and sensitivity to instrumental effects
• Characterization of low-mass companion HD 142527 B
• Simulated performance of the molecular mapping for young giant exoplanets with the Medium Resolution Spectrometer of JWST/MIRI
• Imaging exoplanets with coronagraphic instruments
• Machine Learning for Exoplanet Detection in High-Contrast Spectroscopy: Revealing Exoplanets by Leveraging Hidden Molecular Signatures in Cross-Correlated Spectra with Convolutional Neural Networks
• NA-SODINN: a deep learning algorithm for exoplanet image detection based on residual noise regimes

Cited by in corpus (1)

• Machine Learning for Exoplanet Detection in High-Contrast Spectroscopy: Revealing Exoplanets by Leveraging Hidden Molecular Signatures in Cross-Correlated Spectra with Convolutional Neural Networks