Astrophysical Images Research Articles

Increasing the raw contrast of VLT/SPHERE with the dark hole technique

Context. Imaging exoplanetary systems is essential to characterizing exoplanet atmospheres and orbits as well as circumstellar disks and to studying planet-disk interactions to understand the planet formation processes. Imaging exoplanets or circumstellar disks in the visible and near-infrared is challenging, however, because these objects are very faint relative to their star, even though only fractions of an arcsecond away. Coronagraphic instruments have already allowed the imaging of a few exoplanets, but their performance is limited by wavefront aberrations. Adaptive optics systems partly compensate for the Earth’s atmosphere turbulence, but they cannot fully control the wavefront. Some of the starlight leaks through the coronagraph and forms speckles in the astrophysical image. Focal plane wavefront control, used as a second stage after the adaptive optics system, has been proposed to minimize the speckle intensity within an area called the dark hole. Aims. We previously demonstrated the on-sky performance of dark hole techniques, pairwise probing coupled with electric field conjugation, using the apodized pupil Lyot coronagraph of the VLT/SPHERE instrument. In this paper, we probe their performance using the SPHERE four-quadrant phase mask coronagraph, and we demonstrate the interest of combining dark hole techniques and reference differential imaging. Methods. We used these dark hole techniques on-sky to create a dark hole in the narrow band around 1.7 |j.m observing HR 4796. We then recorded broadband images of HR 4796 and a reference star at the H band. Results. The dark hole techniques improved the H-band detection limit by a factor of three. The dark hole was stable from one star to a nearby star enabling reference differential imaging. Conclusions. This stability offers two new strategies of observation. First, one can quickly create a dark hole observing a bright star before pointing to a faint target star. Furthermore, one can couple dark hole techniques and reference differential imaging. A very interesting point is that the performance of these methods does not depend on the astrophysical signal.

Transfer learning for galaxy feature detection: Finding giant star-forming clumps in low-redshift galaxies using Faster Region-based Convolutional Neural Network

Abstract Giant star-forming clumps (GSFCs) are areas of intensive star-formation that are commonly observed in high-redshift (z ≳ 1) galaxies but their formation and role in galaxy evolution remain unclear. Observations of low-redshift clumpy galaxy analogues are rare but the availability of wide-field galaxy survey data makes the detection of large clumpy galaxy samples much more feasible. Deep Learning (DL), and in particular Convolutional Neural Networks (CNNs), have been successfully applied to image classification tasks in astrophysical data analysis. However, one application of DL that remains relatively unexplored is that of automatically identifying and localizing specific objects or features in astrophysical imaging data. In this paper, we demonstrate the use of DL-based object detection models to localize GSFCs in astrophysical imaging data. We apply the Faster Region-based Convolutional Neural Network object detection framework (FRCNN) to identify GSFCs in low-redshift (z ≲ 0.3) galaxies. Unlike other studies, we train different FRCNN models on observational data that was collected by the Sloan Digital Sky Survey and labelled by volunteers from the citizen science project ‘Galaxy Zoo: Clump Scout’. The FRCNN model relies on a CNN component as a ‘backbone’ feature extractor. We show that CNNs, that have been pre-trained for image classification using astrophysical images, outperform those that have been pre-trained on terrestrial images. In particular, we compare a domain-specific CNN – ‘Zoobot’ – with a generic classification backbone and find that Zoobot achieves higher detection performance. Our final model is capable of producing GSFC detections with a completeness and purity of ≥0.8 while only being trained on ∼5000 galaxy images.

Status of DUNE Offline Computing

We summarize the status of Deep Underground Neutrino Experiment (DUNE) Offline Software and Computing program. We describe plans for the computing infrastructure needed to acquire, catalog, reconstruct, simulate and analyze the data from the DUNE experiment and its prototypes in pursuit of the experiment’s physics goals of precision measurements of neutrino oscillation parameters, detection of astrophysical neutrinos, measurement of neutrino interaction properties and searches for physics beyond the Standard Model. In contrast to traditional HEP computational problems, DUNE’s Liquid Argon Time Projection Chamber data consist of simple but very large (many GB) data objects which share many characteristics with astrophysical images. We have successfully reconstructed and simulated data from 4% prototype detector runs at CERN. The data volume from the full DUNE detector, when it starts commissioning late in this decade will present memory management challenges in conventional processing but significant opportunities to use advances in machine learning and pattern recognition as a frontier user of High Performance Computing facilities capable of massively parallel processing. Our goal is to develop infrastructure resources that are flexible and accessible enough to support creative software solutions as HEP computing evolves.

Can diffusion model conditionally generate astrophysical images?

ABSTRACT Generative adversarial networks (GANs) are frequently utilized in astronomy to construct an emulator of numerical simulations. Nevertheless, training GANs can prove to be a precarious task, as they are prone to instability and often lead to mode collapse problems. Conversely, the diffusion model also has the ability to generate high-quality data without adversarial training. It has shown superiority over GANs with regard to several natural image data sets. In this study, we undertake a quantitative comparison between the denoising diffusion probabilistic model (DDPM) and StyleGAN2 (one of the most robust types of GANs) via a set of robust summary statistics from scattering transform. In particular, we utilize both models to generate the images of 21 cm brightness temperature mapping, as a case study, conditionally based on astrophysical parameters that govern the process of cosmic reionization. Using our new Fréchet Scattering Distance (FSD) as the evaluation metric to quantitatively compare the sample distribution between generative models and simulations, we demonstrate that DDPM outperforms StyleGAN2 on varied sizes of training sets. Through Fisher forecasts, we demonstrate that on our data sets, StyleGAN2 exhibits mode collapses in varied ways, while DDPM yields a more robust generation. We also explore the role of classifier-free guidance in DDPM and show the preference for a non-zero guidance scale only when the training data is limited. Our findings indicate that the diffusion model presents a promising alternative to GANs in the generation of accurate images. These images can subsequently provide reliable parameter constraints, particularly in the realm of astrophysics.

Transport in Astrophysics: II. Diffusion with Advection in Expanding Nebulae

The structure across an expanding shell in which drift and diffusion redistribute material requires careful consideration in order to correlate the surface brightness with the physical parameters. A new solution in terms of Fourier series is suggested for 1D diffusion in the presence of a drift velocity. The astrophysical parameters are chosen in agreement with the astronomical data for the planetary nebula A39. The new solution is then inserted into the existing theory for the astrophysical image which allows dealing with the intensity of radiation emitted in a spherical layer.

Robust Frequency Estimation Under Additive Mixture Noise

In many applications such as multiuser radar communications and astrophysical imaging processing, the encountered noise is usually described by the finite sum of -stable variables. In this paper, a new parameter estimator is developed, in the presence of this new heavy-tailed noise. Since the closed-form PDF of the -stable variable does not exist except and , we take the sum of the Cauchy () and Gaussian () noise as an example, namely, additive Cauchy-Gaussian (ACG) noise. The probability density function (PDF) of the mixed random variable, can be calculated by the convolution of the Cauchy's PDF and Gaussian's PDF. Because of the complicated integral in the PDF expression of the ACG noise, traditional estimators, e.g., maximum likelihood, are analytically not tractable. To obtain the optimal estimates, a new robust frequency estimator is devised by employing the Metropolis-Hastings (M-H) algorithm. Meanwhile, to guarantee the fast convergence of the M-H chain, a new proposal covariance criterion is also devised, where the batch of previous samples are utilized to iteratively update the proposal covariance in each sampling process. Computer simulations are carried out to indicate the superiority of the developed scheme, when compared with several conventional estimators and the Cramér-Rao lower bound.

Feature augmentation for the inversion of the Fourier transform with limited data

We investigate an interpolation/extrapolation method that, given scattered observations of the Fourier transform, approximates its inverse. The interpolation algorithm takes advantage of modeling the available data via a shape-driven interpolation based on variably scaled Kernels (VSKs), whose implementation is here tailored for inverse problems. The so-constructed interpolants are used as inputs for a standard iterative inversion scheme. After providing theoretical results concerning the spectrum of the VSK collocation matrix, we test the method on astrophysical imaging benchmarks.

Change-point Detection and Image Segmentation for Time Series of Astrophysical Images

Abstract Many astrophysical phenomena are time-varying, in the sense that their intensity, energy spectrum, and/or the spatial distribution of the emission suddenly change. This paper develops a method for modeling a time series of images. Under the assumption that the arrival times of the photons follow a Poisson process, the data are binned into 4D grids of voxels (time, energy band, and x-y coordinates), and viewed as a time series of non-homogeneous Poisson images. The method assumes that at each time point, the corresponding multiband image stack is an unknown 3D piecewise constant function including Poisson noise. It also assumes that all image stacks between any two adjacent change points (in time domain) share the same unknown piecewise constant function. The proposed method is designed to estimate the number and the locations of all of the change points (in time domain), as well as all of the unknown piecewise constant functions between any pairs of the change points. The method applies the minimum description length principle to perform this task. A practical algorithm is also developed to solve the corresponding complicated optimization problem. Simulation experiments and applications to real data sets show that the proposed method enjoys very promising empirical properties. Applications to two real data sets, the XMM observation of a flaring star and an emerging solar coronal loop, illustrate the usage of the proposed method and the scientific insight gained from it.

High-Resolution Multi-Spectral Imaging With Diffractive Lenses and Learned Reconstruction

Spectral imaging is a fundamental diagnostic technique with widespread application. Conventional spectral imaging approaches have intrinsic limitations on spatial and spectral resolutions due to the physical components they rely on. To overcome these physical limitations, in this paper, we develop a novel multi-spectral imaging modality that enables higher spatial and spectral resolutions. In the developed computational imaging modality, we exploit a diffractive lens, such as a photon sieve, for both dispersing and focusing the optical field, and achieve measurement diversity by changing the focusing behavior of this lens. Because the focal length of a diffractive lens is wavelength-dependent, each measurement is a superposition of differently blurred spectral components. To reconstruct the individual spectral images from these superimposed and blurred measurements, model-based fast reconstruction algorithms are developed with deep and analytical priors using alternating minimization and unrolling. Finally, the effectiveness and performance of the developed technique is illustrated for an application in astrophysical imaging under various observation scenarios in the extreme ultraviolet (EUV) regime. The results demonstrate that the technique provides not only diffraction-limited high spatial resolution, as enabled by diffractive lenses, but also the capability of resolving close-by spectral sources that would not otherwise be possible with the existing techniques. This work enables high resolution multi-spectral imaging with low cost designs for a variety of applications and spectral regimes.

Artifact-less coded aperture imaging in the x-ray band with multiple different random patterns

The coded aperture imaging technique is a useful method of x-ray imaging in observational astrophysics. However, the presence of imaging noise or so-called artifacts in a decoded image is a drawback of this method. We propose a coded aperture imaging method using multiple different random patterns for significantly reducing the image artifacts. This aperture mask contains multiple different patterns each of which generates a different artifact distribution in its decoded image. By summing all decoded images of the different patterns, the artifact distributions are cancelled out, and we obtain a remarkably accurate image. We demonstrate this concept with imaging experiments of a monochromatic 16-keV hard x-ray beam at the synchrotron photon facility SPring-8, using the combination of a complementary metal-oxide-semiconductor image sensor and an aperture mask that has four different random patterns composed of holes with a diameter of 27 μm and a separation of 39 μm. The entire imaging system is installed in a 25-cm-long compact size and achieves an angular resolution of <30 arc sec (full-width at half-maximum). In addition, we show by Monte Carlo simulation that the artifacts can be reduced more effectively if the number of different patterns increases to 8 or 16.

Vetting the optical transient candidates detected by the GWAC network using convolutional neural networks

ABSTRACT The observation of the transient sky through a multitude of astrophysical messengers has led to several scientific breakthroughs in the last two decades, thanks to the fast evolution of the observational techniques and strategies employed by the astronomers. Now, it requires to be able to coordinate multiwavelength and multimessenger follow-up campaigns with instruments both in space and on ground jointly capable of scanning a large fraction of the sky with a high-imaging cadency and duty cycle. In the optical domain, the key challenge of the wide field-of-view telescopes covering tens to hundreds of square degrees is to deal with the detection, identification, and classification of hundreds to thousands of optical transient (OT) candidates every night in a reasonable amount of time. In the last decade, new automated tools based on machine learning approaches have been developed to perform those tasks with a low computing time and a high classification efficiency. In this paper, we present an efficient classification method using convolutional neural networks (CNNs) to discard many common types of bogus falsely detected in astrophysical images in the optical domain. We designed this tool to improve the performances of the OT detection pipeline of the Ground Wide field Angle Cameras (GWAC) telescopes, a network of robotic telescopes aiming at monitoring the OT sky down to R = 16 with a 15 s imaging cadency. We applied our trained CNN classifier on a sample of 1472 GWAC OT candidates detected by the real-time detection pipeline.

Local Seeing Measurement for Increasing Astrophysical Observatory Quality Images Using an Autonomous Wireless Sensor Network.

Astrophysical observatories (AOs) are used to acquire high-quality images from the sky. However, AOs are amenable to distortion phenomena such as seeing. In this paper, we consider specifically the local seeing produced from indoor and outdoor temperature variations. Local seeing contributes to the captured image quality, therefore it must be measured. Local seeing has been considered, to the best of our knowledge, in observatories using ad hoc solutions, typically with high cost and complexity. This paper presents the complete development of an autonomous wireless sensor network (WSN) composed of temperature-measuring for real-time local seeing measurement. Therefore, a WSN is deployed using Zigbee as a data communication exchange. As a result, a long continuous-operating system is constructed and tested in a real AO infrastructure. Finally, we calculate a preliminary dome local seeing, from the acquired temperature data, achieving reasonable results.

Computing for the DUNE Long-Baseline Neutrino Oscillation Experiment

This paper is based on a talk given at Computing in High Energy Physics in Adelaide, South Australia, Australia in November 2019. It is partially intended to explain the context of DUNE Computing for computing specialists. The Deep Underground Neutrino Experiment (DUNE) collaboration consists of over 180 institutions from 33 countries. The experiment is in preparation now, with commissioning of the first 10kT fiducial volume Liquid Argon TPC expected over the period 2025-2028 and a long data taking run with 4 modules expected from 2029 and beyond. An active prototyping program is already in place with a short test-beam run with a 700T, 15,360 channel prototype of single-phase readout at the Neutrino Platform at CERN in late 2018 and tests of a similar sized dual-phase detector scheduled for mid-2019. The 2018 test-beam run was a valuable live test of our computing model. The detector produced raw data at rates of up to 2GB/s. These data were stored at full rate on tape at CERN and Fermilab and replicated at sites in the UK and Czech Republic. In total 1.2 PB of raw data from beam and cosmic triggers were produced and reconstructed during the six week testbeam run. Baseline predictions for the full DUNE detector data, starting in the late 2020’s are 30-60 PB of raw data per year. In contrast to traditional HEP computational problems, DUNE’s Liquid Argon TPC data consist of simple but very large (many GB) 2D data objects which share many characteristics with astrophysical images. This presents opportunities to use advances in machine learning and pattern recognition as a frontier user of High Performance Computing facilities capable of massively parallel processing.

High-resolution speckle imaging

Mathematics and technology combine to make a virtue of the graininess imposed on astrophysical images by atmospheric turbulence.

Noise-gating to Clean Astrophysical Image Data

Abstract I present a family of algorithms to reduce noise in astrophysical images and image sequences, preserving more information from the original data than is retained by conventional techniques. The family uses locally adaptive filters (“noise gates”) in the Fourier domain to separate coherent image structure from background noise based on the statistics of local neighborhoods in the image. Processing of solar data limited by simple shot noise or by additive noise reveals image structure not easily visible in the originals, preserves photometry of observable features, and reduces shot noise by a factor of 10 or more with little to no apparent loss of resolution. This reveals faint features that were either not directly discernible or not sufficiently strongly detected for quantitative analysis. The method works best on image sequences containing related subjects, for example movies of solar evolution, but is also applicable to single images provided that there are enough pixels. The adaptive filter uses the statistical properties of noise and of local neighborhoods in the data to discriminate between coherent features and incoherent noise without reference to the specific shape or evolution of those features. The technique can potentially be modified in a straightforward way to exploit additional a priori knowledge about the functional form of the noise.

Generative adversarial networks recover features in astrophysical images of galaxies beyond the deconvolution limit

Abstract Observations of astrophysical objects such as galaxies are limited by various sources of random and systematic noise from the sky background, the optical system of the telescope and the detector used to record the data. Conventional deconvolution techniques are limited in their ability to recover features in imaging data by the Shannon–Nyquist sampling theorem. Here, we train a generative adversarial network (GAN) on a sample of 4550 images of nearby galaxies at 0.01 &amp;lt; z &amp;lt; 0.02 from the Sloan Digital Sky Survey and conduct 10× cross-validation to evaluate the results. We present a method using a GAN trained on galaxy images that can recover features from artificially degraded images with worse seeing and higher noise than the original with a performance that far exceeds simple deconvolution. The ability to better recover detailed features such as galaxy morphology from low signal to noise and low angular resolution imaging data significantly increases our ability to study existing data sets of astrophysical objects as well as future observations with observatories such as the Large Synoptic Sky Telescope (LSST) and the Hubble and James Webb space telescopes.

Convolution kernels for multi-wavelength imaging

Astrophysical images issued from different instruments and/or spectral bands often require to be processed together, either for fitting or comparison purposes. However each image is affected by an instrumental response, also known as point-spread function (PSF), that depends on the characteristics of the instrument as well as the wavelength and the observing strategy. Given the knowledge of the PSF in each band, a straightforward way of processing images is to homogenise them all to a target PSF using convolution kernels, so that they appear as if they had been acquired by the same instrument. We propose an algorithm that generates such PSF-matching kernels, based on Wiener filtering with a tunable regularisation parameter. This method ensures all anisotropic features in the PSFs to be taken into account. We compare our method to existing procedures using measured Herschel/PACS and SPIRE PSFs and simulated JWST/MIRI PSFs. Significant gains up to two orders of magnitude are obtained with respect to the use of kernels computed assuming Gaussian or circularised PSFs. A software to compute these kernels is available at https://github.com/aboucaud/pypher

Astrophysical imaging of Kerr black holes with scalar hair

We address the astrophysical imaging of a family of deformed Kerr black holes (BHs). These are stationary, asymptotically flat black hole (BH) spacetimes, that are solutions of General Relativity minimally coupled to a massive, complex scalar field: Kerr BHs with scalar hair (KBHsSH). Such BHs bifurcate from the vacuum Kerr solution and can be regarded as a horizon within a rotating boson star. In a recent letter, it was shown that KBHsSH can exhibit very distinct shadows from the ones of their vacuum counterparts. The setup therein, however, considered the light source to be a celestial sphere sufficiently far away from the BH. Here, we analyse KBHsSH surrounded by an emitting torus of matter, simulating a more realistic astrophysical environment, and study the corresponding lensing of light as seen by a very far away observer, to appropriately model ground-based observations of Sgr A*. We find that the differences in imaging between KBHsSH and comparable vacuum Kerr BHs remain, albeit less dramatic than those observed for the corresponding shadows in the previous setup. In particular, we highlight two observables that might allow differentiating KBHsSH and Kerr BHs. The first is the angular size of the photon ring (in a Kerr spacetime) or lensing ring (in a KBHSH spacetime), the latter being significantly smaller for sufficiently non-Kerr-like spacetimes. The second is the existence of an edge in the intensity distribution (the photon ring in Kerr spacetime). This edge can disappear for very non-Kerr-like KBHsSH. It is plausible, therefore, that sufficiently precise Very Long Baseline Interferometric observations of BH candidates can constrain this model.

Enhancing retinal images by nonlinear registration

Being able to image the human retina in high resolution opens a new era in many important fields, such as pharmacological research for retinal diseases, researches in human cognition, nervous system, metabolism and blood stream, to name a few. In this paper, we propose to share the knowledge acquired in the fields of optics and imaging in solar astrophysics in order to improve the retinal imaging in the perspective to perform a medical diagnosis. The main purpose would be to assist health care practitioners by enhancing the spatial resolution of the retinal images and increase the level of confidence of the abnormal feature detection. We apply a nonlinear registration method using local correlation tracking to increase the field of view and follow structure evolutions using correlation techniques borrowed from solar astronomy technique expertise. Another purpose is to define the tracer of movements after analyzing local correlations to follow the proper motions of an image from one moment to another, such as changes in optical flows that would be of high interest in a medical diagnosis.

Microphysics in Astrophysical Plasmas

Although macroscale features dominate astrophysical images and energetics, the physics is controlled through microscale transport processes (conduction, diffusion) that mediate the flow of mass, momentum, energy, and charge. These microphysical processes manifest themselves in key (all) boundary layers and also operate within the body of the plasma. Crucially, most plasmas of interest are rarefied to the extent that classical particle collision length- and time-scales are long. Collective plasma kinetic phenomena then serve to scatter or otherwise modify the particle distribution functions and in so-doing govern the transport at the microscale level. Thus collisionless plasmas are capable of supporting thin shocks, current sheets which may be prone to magnetic reconnection, and the dissipation of turbulence cascades at kinetic scales. This paper lays the foundation for the accompanying collection that explores the current state of knowledge in this subject. The richness of plasma kinetic phenomena brings with it a rich diversity of microphysics that does not always, if ever, simply mimic classical collision-dominated transport. This can couple the macro- and microscale physics in profound ways, and in ways which thus depend on the astrophysical context.