Search

Inferring atmospheric properties of exoplanets from observed spectra is key to understanding their formation, evolution, and habitability. Since traditional Bayesian approaches to atmospheric retrieval (e.g., nested sampling) are computationally expensive, a growing number of machine learning (ML) methods such as neural posterior estimation (NPE) have been proposed. We seek to make ML-based atmospheric retrieval (1) more reliable and accurate with verified results, and (2) more flexible with respect to the underlying neural networks and the choice of the assumed noise models. First, we adopt flow matching posterior estimation (FMPE) as a new ML approach to atmospheric retrieval. FMPE maintains many advantages of NPE, but provides greater architectural flexibility and scalability. Second, we use importance sampling (IS) to verify and correct ML results, and to compute an estimate of the Bayesian evidence. Third, we condition our ML models on the assumed noise level of a spectrum (i.e., error bars), thus making them adaptable to different noise models. Both our noise level-conditional FMPE and NPE models perform on par with nested sampling across a range of noise levels when tested on simulated data. FMPE trains about 3 times faster than NPE and yields higher IS efficiencies. IS successfully corrects inaccurate ML results, identifies model failures via low efficiencies, and provides accurate estimates of the Bayesian evidence. FMPE is a powerful alternative to NPE for fast, amortized, and parallelizable atmospheric retrieval. IS can verify results, thus helping to build confidence in ML-based approaches, while also facilitating model comparison via the evidence ratio. Noise level conditioning allows design studies for future instruments to be scaled up, for example, in terms of the range of signal-to-noise ratios., Comment: Accepted for publication in Astronomy & Astrophysics

The main challenge of exoplanet high-contrast imaging (HCI) is to separate the signal of exoplanets from their host stars, which are many orders of magnitude brighter. HCI for ground-based observations is further exacerbated by speckle noise originating from perturbations in the Earth's atmosphere and imperfections in the telescope optics. Various data post-processing techniques are used to remove this speckle noise and reveal the faint planet signal. Often, however, a significant part of the planet signal is accidentally subtracted together with the noise. In the present work, we use explainable machine learning to investigate the reason for the loss of the planet signal for one of the most used post-processing methods: Principal Component Analysis (PCA). We find that PCA learns the shape of the telescope point spread function for high numbers of PCA components. This representation of the noise captures not only the speckle noise, but also the characteristic shape of the planet signal. Building upon these insights, we develop a new post-processing method (4S) that constrains the noise model to minimize this signal loss. We apply our model to 11 archival HCI datasets from the VLT-NACO instrument in the L'-band and find that our model consistently outperforms PCA. The improvement is largest at close separations to the star ($\leq 4 \lambda /D$) providing up to 1.5 magnitudes deeper contrast. This enhancement enables us to detect the exoplanet AF Lep b in data from 2011, 11 years before its subsequent discovery. We present updated orbital parameters for this object., Comment: Submitted to AJ, 26 pages, 15 figures, comments welcome, code available on ReadtheDocs: https://fours.readthedocs.io/en/latest/

Atmospheric retrievals (AR) characterize exoplanets by estimating atmospheric parameters from observed light spectra, typically by framing the task as a Bayesian inference problem. However, traditional approaches such as nested sampling are computationally expensive, thus sparking an interest in solutions based on machine learning (ML). In this ongoing work, we first explore flow matching posterior estimation (FMPE) as a new ML-based method for AR and find that, in our case, it is more accurate than neural posterior estimation (NPE), but less accurate than nested sampling. We then combine both FMPE and NPE with importance sampling, in which case both methods outperform nested sampling in terms of accuracy and simulation efficiency. Going forward, our analysis suggests that simulation-based inference with likelihood-based importance sampling provides a framework for accurate and efficient AR that may become a valuable tool not only for the analysis of observational data from existing telescopes, but also for the development of new missions and instruments., Comment: Accepted at the "AI to Accelerate Science and Engineering (AI2ASE)" workshop at AAAI 2024

The investigation of the atmospheres of closely separated, directly imaged gas giant exoplanets is challenging due to the presence of stellar speckles that pollute their spectrum. To remedy this, the analysis of medium- to high-resolution spectroscopic data via cross-correlation with spectral templates (cross-correlation spectroscopy) is emerging as a leading technique. We aim to define a robust Bayesian framework combining, for the first time, three widespread direct-imaging techniques, namely photometry, low-resolution spectroscopy, and medium-resolution cross-correlation spectroscopy in order to derive the atmospheric properties of close-in directly imaged exoplanets. Our framework CROCODILE (cross-correlation retrievals of directly imaged self-luminous exoplanets) naturally combines the three techniques by adopting adequate likelihood functions. To validate our routine, we simulated observations of gas giants similar to the well-studied $\beta$~Pictoris~b planet and we explored the parameter space of their atmospheres to search for potential biases. We obtain more accurate measurements of atmospheric properties when combining photometry, low- and medium-resolution spectroscopy into atmospheric retrievals than when using the techniques separately as is usually done in the literature. We find that medium-resolution ($R \approx 4000$) K-band cross-correlation spectroscopy alone is not suitable to constrain the atmospheric properties of our synthetic datasets; however, this problem disappears when simultaneously fitting photometry and low-resolution ($R \approx 60$) spectroscopy between the Y and M bands. Our framework allows the atmospheric characterisation of directly imaged exoplanets using the high-quality spectral data that will be provided by the new generation of instruments such as VLT/ERIS, JWST/MIRI, and ELT/METIS.

The detection and characterization of Earth-like exoplanets around Sun-like stars is a primary science motivation for the Habitable Worlds Observatory. However, the current best technology is not yet advanced enough to reach the 10^-10 contrasts at close angular separations and at the same time remain insensitive to low-order aberrations, as would be required to achieve high-contrast imaging of exo-Earths. Photonic technologies could fill this gap, potentially doubling exo-Earth yield. We review current work on photonic coronagraphs and investigate the potential of hybridized designs which combine both classical coronagraph designs and photonic technologies into a single optical system. We present two possible systems. First, a hybrid solution which splits the field of view spatially such that the photonics handle light within the inner working angle and a conventional coronagraph that suppresses starlight outside it. Second, a hybrid solution where the conventional coronagraph and photonics operate in series, complementing each other and thereby loosening requirements on each subsystem. As photonic technologies continue to advance, a hybrid or fully photonic coronagraph holds great potential for future exoplanet imaging from space., Comment: Conference Proceedings of SPIE: Techniques and Instrumentation for Detection of Exoplanets XI, vol. 12680 (2023)

Atmospheric retrievals (AR) of exoplanets typically rely on a combination of a Bayesian inference technique and a forward simulator to estimate atmospheric properties from an observed spectrum. A key component in simulating spectra is the pressure-temperature (PT) profile, which describes the thermal structure of the atmosphere. Current AR pipelines commonly use ad hoc fitting functions here that limit the retrieved PT profiles to simple approximations, but still use a relatively large number of parameters. In this work, we introduce a conceptually new, data-driven parameterization scheme for physically consistent PT profiles that does not require explicit assumptions about the functional form of the PT profiles and uses fewer parameters than existing methods. Our approach consists of a latent variable model (based on a neural network) that learns a distribution over functions (PT profiles). Each profile is represented by a low-dimensional vector that can be used to condition a decoder network that maps $P$ to $T$. When training and evaluating our method on two publicly available datasets of self-consistent PT profiles, we find that our method achieves, on average, better fit quality than existing baseline methods, despite using fewer parameters. In an AR based on existing literature, our model (using two parameters) produces a tighter, more accurate posterior for the PT profile than the five-parameter polynomial baseline, while also speeding up the retrieval by more than a factor of three. By providing parametric access to physically consistent PT profiles, and by reducing the number of parameters required to describe a PT profile (thereby reducing computational cost or freeing resources for additional parameters of interest), our method can help improve AR and thus our understanding of exoplanet atmospheres and their habitability., Comment: Accepted for publication in Astronomy & Astrophysics

Looking to the future of exo-Earth imaging from the ground, core technology developments are required in visible extreme adaptive optics (ExAO) to enable the observation of atmospheric features such as oxygen on rocky planets in visible light. UNDERGROUND (Ultra-fast AO techNology Determination for Exoplanet imageRs from the GROUND), a collaboration built in Feb. 2023 at the Optimal Exoplanet Imagers Lorentz Workshop, aims to (1) motivate oxygen detection in Proxima Centauri b and analogs as an informative science case for high-contrast imaging and direct spectroscopy, (2) overview the state of the field with respect to visible exoplanet imagers, and (3) set the instrumental requirements to achieve this goal and identify what key technologies require further development., Comment: SPIE Proceeding: 2023 / 12680-67

NASA is engaged in planning for a Habitable Worlds Observatory (HabWorlds), a coronagraphic space mission to detect rocky planets in habitable zones and establish their habitability. Surface liquid water is central to the definition of planetary habitability. Photometric and polarimetric phase curves of starlight reflected by an exoplanet can reveal ocean glint, rainbows and other phenomena caused by scattering by clouds or atmospheric gas. Direct imaging missions are optimised for planets near quadrature, but HabWorlds' coronagraph may obscure the phase angles where such optical features are strongest. The range of accessible phase angles for a given exoplanet will depend on the planet's orbital inclination and/or the coronagraph's inner working angle (IWA). We use a recently-created catalog relevant to HabWorlds of 164 stars to estimate the number of exo-Earths that could be searched for ocean glint, rainbows, and polarization effects due to Rayleigh scattering. We find that the polarimetric Rayleigh scattering peak is accessible in most of the exo-Earth planetary systems. The rainbow due to water clouds at phase angles of ${\sim}20-60^\circ$ would be accessible with HabWorlds for a planet with an Earth equivalent instellation in ${\sim}{46}$ systems, while the ocean glint signature at phase angles of ${\sim}130-170^\circ$ would be accessible in ${\sim}{16}$ systems, assuming an IWA${=}62$ mas ($3\lambda/D$). Improving the IWA${=}41$ mas ($2\lambda/D$) increases accessibility to rainbows and glints by factors of approximately 2 and 3, respectively. By observing these scattering features, HabWorlds could detect a surface ocean and water cycle, key indicators of habitability., Comment: MNRAS accepted, 9 pages, 8 figures, 3 tables

Over the past decade, hundreds of nights have been spent on the worlds largest telescopes to search for and directly detect new exoplanets using high-contrast imaging (HCI). Thereby, two scientific goals are of central interest: First, to study the characteristics of the underlying planet population and distinguish between different planet formation and evolution theories. Second, to find and characterize planets in our immediate Solar neighborhood. Both goals heavily rely on the metric used to quantify planet detections and non-detections. Current standards often rely on several explicit or implicit assumptions about the noise. For example, it is often assumed that the residual noise after data post-processing is Gaussian. While being an inseparable part of the metric, these assumptions are rarely verified. This is problematic as any violation of these assumptions can lead to systematic biases. This makes it hard, if not impossible, to compare results across datasets or instruments with different noise characteristics. We revisit the fundamental question of how to quantify detection limits in HCI. We focus our analysis on the error budget resulting from violated assumptions. To this end, we propose a new metric based on bootstrapping that generalizes current standards to non-Gaussian noise. We apply our method to archival HCI data from the NACO-VLT instrument and derive detection limits for different types of noise. Our analysis shows that current standards tend to give detection limit that are about one magnitude too optimistic in the speckle-dominated regime. That is, HCI surveys may have excluded planets that can still exist., Comment: After first iteration with the referee, resubmitted to AJ. Comments welcome!

High-contrast imaging of exoplanets hinges on powerful post-processing methods to denoise the data and separate the signal of a companion from its host star, which is typically orders of magnitude brighter. Existing post-processing algorithms do not use all prior domain knowledge that is available about the problem. We propose a new method that builds on our understanding of the systematic noise and the causal structure of the data-generating process. Our algorithm is based on a modified version of half-sibling regression (HSR), a flexible denoising framework that combines ideas from the fields of machine learning and causality. We adapt the method to address the specific requirements of high-contrast exoplanet imaging data obtained in pupil tracking mode. The key idea is to estimate the systematic noise in a pixel by regressing the time series of this pixel onto a set of causally independent, signal-free predictor pixels. We use regularized linear models in this work; however, other (non-linear) models are also possible. In a second step, we demonstrate how the HSR framework allows us to incorporate observing conditions such as wind speed or air temperature as additional predictors. When we apply our method to four data sets from the VLT/NACO instrument, our algorithm provides a better false-positive fraction than PCA-based PSF subtraction, a popular baseline method in the field. Additionally, we find that the HSR-based method provides direct and accurate estimates for the contrast of the exoplanets without the need to insert artificial companions for calibration in the data sets. Finally, we present first evidence that using the observing conditions as additional predictors can improve the results. Our HSR-based method provides an alternative, flexible and promising approach to the challenge of modeling and subtracting the stellar PSF and systematic noise in exoplanet imaging data., Comment: Accepted for publication in Astronomy & Astrophysics

The detection of exoplanets in high-contrast imaging (HCI) data hinges on post-processing methods to remove spurious light from the host star. So far, existing methods for this task hardly utilize any of the available domain knowledge about the problem explicitly. We propose a new approach to HCI post-processing based on a modified half-sibling regression scheme, and show how we use this framework to combine machine learning with existing scientific domain knowledge. On three real data sets, we demonstrate that the resulting system performs clearly better (both visually and in terms of the SNR) than one of the currently leading algorithms. If further studies can confirm these results, our method could have the potential to allow significant discoveries of exoplanets both in new and archival data., Comment: Accepted at the "Machine Learning and the Physical Sciences" workshop at NeurIPS 2020. 9 pages, 6 figures

In the last few years, machine learning techniques, in particular convolutional neural networks, have been investigated as a method to replace or complement traditional matched filtering techniques that are used to detect the gravitational-wave signature of merging black holes. However, to date, these methods have not yet been successfully applied to the analysis of long stretches of data recorded by the Advanced LIGO and Virgo gravitational-wave observatories. In this work, we critically examine the use of convolutional neural networks as a tool to search for merging black holes. We identify the strengths and limitations of this approach, highlight some common pitfalls in translating between machine learning and gravitational-wave astronomy, and discuss the interdisciplinary challenges. In particular, we explain in detail why convolutional neural networks alone cannot be used to claim a statistically significant gravitational-wave detection. However, we demonstrate how they can still be used to rapidly flag the times of potential signals in the data for a more detailed follow-up. Our convolutional neural network architecture as well as the proposed performance metrics are better suited for this task than a standard binary classifications scheme. A detailed evaluation of our approach on Advanced LIGO data demonstrates the potential of such systems as trigger generators. Finally, we sound a note of caution by constructing adversarial examples, which showcase interesting "failure modes" of our model, where inputs with no visible resemblance to real gravitational-wave signals are identified as such by the network with high confidence., Comment: First two authors contributed equally; appeared at Phys. Rev. D