Advances on the classification of radio image cubes
Bernoulli Institute for Mathematics, Computer Science and Artificial Intelligence（伯努利数学、计算机科学与人工智能研究所）；University of Groningen（格罗宁根大学）；University of Stellenbosch（斯泰伦博斯大学）；ASTRON（荷兰射电天文研究所）
Steven Ndung'u, Trienko Grobler, Stefan J. Wijnholds, Dimka Karastoyanova, George Azzopardi
备注：
21 page review paper submitted to New astronomy reviews journal for review
对于平方千米阵（Square Kilometre Array, SKA）这类系统而言，现代射电望远镜每天将生成艾字节量级的数据集。海量数据集是未知且罕见的天体物理现象的来源，能够带来新发现。尽管如此，这一目标只有通过运用高强度机器智能来补充人工辅助和传统统计技术才有可能实现。近年来，聚焦于人工智能在射电天文学中应用的科学出版物激增，这些研究解决了源提取、形态分类和异常检测等挑战。本研究对机器智能技术在射电图像中的应用进行了简洁而全面的综述，重点关注射电星系的形态分类。研究旨在对相关论文进行详细梳理，基于射电天文学中的数据复杂度、数据预处理和方法创新性三个维度对文献进行总结。计算机智能在射电天文学中的快速发展和应用，已经在艰巨的数据处理自动化领域引发了一场革命和新的范式转变。然而，要在射电天文学中最优地利用人工智能，需要在标注数据集创建方面持续开展协作努力。此外，为了快速定位具有相似或不同物理特征的射电星系，需要对已识别的射电源建立索引。但这一问题在现有文献中尚未得到充分解决，因此是一个有待进一步研究的开放领域。
英文摘要：
Modern radio telescopes will daily generate data sets on the scale of exabytes for systems like the Square Kilometre Array (SKA). Massive data sets are a source of unknown and rare astrophysical phenomena that lead to discoveries. Nonetheless, this is only plausible with the exploitation of intensive machine intelligence to complement human-aided and traditional statistical techniques. Recently, there has been a surge in scientific publications focusing on the use of artificial intelligence in radio astronomy, addressing challenges such as source extraction, morphological classification, and anomaly detection. This study presents a succinct, but comprehensive review of the application of machine intelligence techniques on radio images with emphasis on the morphological classification of radio galaxies. It aims to present a detailed synthesis of the relevant papers summarizing the literature based on data complexity, data pre-processing, and methodological novelty in radio astronomy. The rapid advancement and application of computer intelligence in radio astronomy has resulted in a revolution and a new paradigm shift in the automation of daunting data processes. However, the optimal exploitation of artificial intelligence in radio astronomy, calls for continued collaborative efforts in the creation of annotated data sets. Additionally, in order to quickly locate radio galaxies with similar or dissimilar physical characteristics, it is necessary to index the identified radio sources. Nonetheless, this issue has not been adequately addressed in the literature, making it an open area for further study.
Magnetic fields in galaxies exist on various spatial scales. Large-scale magnetic fields are thought to be generated by the $α-Ω$ dynamo. Small-scale galactic magnetic fields (1 kpc and below) can be generated by tangling the large-scale field or by the small-scale turbulent dynamo. The analysis of field structures with the help of polarized radio continuum emission is hampered by the effect of Faraday dispersion (due to fluctuations in magnetic field and/or thermal electron density) that shifts signals from large to small scales. At long observation wavelengths large-scale magnetic fields may…
The new generation of low radio-frequency interferometric arrays have enabled the imaging of the solar corona at high spectro-temporal resolutions and sensitivity. In this article we introduce and implement a formalism to generate flux density and brightness temperature ($T_{\mathrm B}$) maps from such images, using independently obtained disc-integrated solar flux density dynamic spectra. These images collectively generate a 4D data cube, with axes spanning angular coordinates ($θ,ϕ$), frequency ($ν$) and time ($t$). This 4D data cube is the most informative data product which can be generate…
Estimating the spectral resolution of a radio data cube from its noise properties
INAF - Osservatorio Astronomico di Cagliari（意大利国家天体物理研究所卡利亚里天文台）；Department of Physics and Astronomy, Chalmers University of Technology（查尔姆斯理工大学物理与天文系）；Onsala Space Observatory（翁萨空间观测站）；ASTRON, the Netherlands Institute for Radio Astronomy（荷兰射电天文研究所ASTRON）；Instituto de Astrofísica de Andalucía (CSIC)（安达卢西亚天体物理研究所（西班牙高等科学研究委员会））；International Centre for Radio Astronomy Research (ICRAR), The University of Western Australia（西澳大利亚大学国际射电天文研究中心）；Departamento de Física Teórica, Universidad Autónoma de Madrid & Astro-UAM（马德里自治大学理论物理系及Astro-UAM）；Kapteyn Astronomical Institute, University of Groningen（格罗宁根大学卡普坦天文研究所）
The current generation of radio and millimeter telescopes, particularly the Atacama Large Millimeter Array (ALMA), offers enormous advances in observing capabilities. While these advances represent an unprecedented opportunity to facilitate scientific understanding, the increased complexity in the spatial and spectral structure of these ALMA data cubes lead to challenges in their interpretation. In this paper, we perform a feasibility study for applying topological data analysis and visualization techniques never before tested by the ALMA community. Through techniques based on contour trees, w…
3D visualization is an important data analysis and knowledge discovery tool, however, interactive visualization of large 3D astronomical datasets poses a challenge for many existing data visualization packages. We present a solution to interactively visualize larger-than-memory 3D astronomical data cubes by utilizing a heterogeneous cluster of CPUs and GPUs. The system partitions the data volume into smaller sub-volumes that are distributed over the rendering workstations. A GPU-based ray casting volume rendering is performed to generate images for each sub-volume, which are composited to gene…
radio-astro-tools: linking radio astronomical data to the astronomical Python ecosystem
Dominion Radio Astrophysical Observatory, Herzberg Astronomy & Astrophysics, National Research Council Canada（加拿大国家研究委员会赫茨伯格天文学与天体物理学多明尼奥射电天文台）；Department of Astronomy, University of Florida（佛罗里达大学天文学系）；Aperio Software（阿佩里奥软件公司）；Dept. of Physics, University of Alberta（阿尔伯塔大学物理系）；Department of Physics and Astronomy, Haverford College（哈弗福德学院物理与天文学系）；Max-Planck-Institut für Astronomie（马克斯·普朗克天文学研究所）；National Radio Astronomy Observatory（美国国家射电天文台）；Department of Astronomy, The Ohio State University（俄亥俄州立大学天文学系）；Center for Cosmology and Astroparticle Physics (CCAPP)（宇宙学与基本粒子物理中心）；Univ. Grenoble Alpes, CNRS, IPAG（格勒诺布尔阿尔卑斯大学，法国国家科学研究中心，IPAG）
The new wide-field radio telescopes, such as: ASKAP, MWA, and SKA; will produce spectral-imaging data-cubes (SIDC) of unprecedented volume. This requires new approaches to managing and servicing the data to the end-user. We present a new integrated framework based on the JPEG2000/ISO/IEC 15444 standard to address the challenges of working with extremely large SIDC. We also present the developed j2k software, that converts and encodes FITS image cubes into JPEG2000 images, paving the way to implementing the pre- sented framework.
Spectral datasets obtained at radio frequencies and optical/IR wavelengths are increasing in complexity as new facilities and instruments come online, resulting in an increased need to visualize and quantitatively analyze the velocity structures. As the visible structure in spectral data cubes is not purely spatial, additional insight is required to relate structures in 2D space plus line-of-sight velocity to their true three-dimensional (3D) structures. This can be achieved through the use of models that are converted to velocity-space representations. We have used the S2PLOT programming libr…
Kapteyn Astronomical Institute, University of Groningen（格罗宁根大学卡普廷天文研究所）；Data Management & Bionetrics (DMB), University of Twente（特文特大学数据管理与生物识别系）；Bernoulli Institute for Mathematics, Computer Science and Artificial Intelligence, University of Groningen（格罗宁根大学伯努利数学、计算机科学与人工智能研究所）
The 21 cm spectral line emission of atomic neutral hydrogen (HI) is one of the primary wavelengths observed in radio astronomy. However, the signal is intrinsically faint and the HI content of galaxies depends on the cosmic environment, requiring large survey volumes and survey depth to investigate the HI Universe. As the amount of data coming from these surveys continues to increase with technological improvements, so does the need for automatic techniques for identifying and characterising HI sources while considering the tradeoff between completeness and purity. This study aimed to find the…
YOLO-CIANNA: Galaxy detection with deep learning in radio data: II. Winning the SKA SDC2 using a generalized 3D-YOLO network
LUX, Observatoire de Paris, Université PSL, Sorbonne Université, CNRS（巴黎天文台卢克斯实验室，巴黎文理研究大学，索邦大学，法国国家科学研究中心）；Canadian Institute for Theoretical Astrophysics, University of Toronto（加拿大理论天体物理研究所，多伦多大学）；Research School of Astronomy & Astrophysics, Australian National University（澳大利亚国立大学天文与天体物理研究学院）；Université de Strasbourg, CNRS UMR 7550, Observatoire astronomique de Strasbourg（斯特拉斯堡大学，法国国家科学研究中心UMR 7550，斯特拉斯堡天文台）；DIO, Observatoire de Paris, CNRS, PSL（巴黎天文台数据与信息办公室，法国国家科学研究中心，巴黎文理研究大学）；IDRIS, CNRS（法国国家科学研究中心集成系统研究所）；Collège de France（法兰西公学院）；Department of Physics & Electronics, Rhodes University（罗兹大学物理与电子系）
We present a framework to interactively volume-render three-dimensional data cubes using distributed ray-casting and volume bricking over a cluster of workstations powered by one or more graphics processing units (GPUs) and a multi-core CPU. The main design target for this framework is to provide an in-core visualization solution able to provide three-dimensional interactive views of terabyte-sized data cubes. We tested the presented framework using a computing cluster comprising 64 nodes with a total of 128 GPUs. The framework proved to be scalable to render a 204 GB data cube with an average…
We report on an exploratory project aimed at performing immersive 3D visualization of astronomical data, starting with spectral-line radio data cubes from galaxies. This work is done as a collaboration between the Department of Physics and Astronomy and the Department of Computer Science at the University of Manitoba. We are building our prototype using the 3D engine Unity, because of its ease of use for integration with advanced displays such as a CAVE environment, a zSpace tabletop, or virtual reality headsets. We address general issues regarding 3D visualization, such as: load and convert a…
KMOS is a multi-object near-infrared integral field spectrometer with 24 deployable pick-off arms. Data processing is inevitably complex. We discuss specific issues and requirements that must be addressed in the data reduction pipeline, the calibration, the raw and processed data formats, and the simulated data. We discuss the pipeline architecture. We focus on its modular style and show how these modules can be used to build a classical pipeline, as well as a more advanced pipeline that can account for both spectral and spatial flexure as well as variations in the OH background. A novel aspec…
We investigate whether the method of wavelet-based Faraday rotation measure (RM) Synthesis can help us to identify structures of regular and turbulent magnetic fields in extended magnetized objects, such as galaxies and galaxy clusters. Wavelets allow us to reformulate the RM synthesis method in a scale-dependent way and to visualize the data as a function of Faraday depth and scale. We present observational tests to recognize magnetic field structures. A region with a regular magnetic field generates a broad "disk" in Faraday space ("Faraday spectrum"), with two "horns" when the distribution …
The sheer volume of data anticipated to be captured by future radio telescopes, such as, The Square Kilometer Array (SKA) and its precursors present new data challenges, including the cost and technical feasibility of data transport and storage. Image and data compression are going to be important techniques to reduce the data size. We provide a quantitative analysis of the effects of JPEG2000's lossy wavelet image compression algorithm on the quality of the radio astronomy imagery data. This analysis is completed by evaluating the completeness, soundness and source parameterisation of the Duc…
Hyperspectral imaging is an ubiquitous technique in solar physics observations and the recent advances in solar instrumentation enabled us to acquire and record data at an unprecedented rate. The huge amount of data which will be archived in the upcoming solar observatories press us to compress the data in order to reduce the storage space and transfer times. The correlation present over all dimensions, spatial, temporal and spectral, of solar data-sets suggests the use of a 3D base wavelet decomposition, to achieve higher compression rates. In this work, we evaluate the performance of the rec…
This paper summarizes some of the major calibration and image reconstruction techniques used in radio interferometry and describes them in a common mathematical framework. The use of this framework has a number of benefits, ranging from clarification of the fundamentals, use of standard numerical optimization techniques, and generalization or specialization to new algorithms.
The advent of enhanced technologies in radio interferometry and the perspective of the SKA telescope bring new challenges in image reconstruction. One of these challenges is the spatio-spectral reconstruction of large (Terabytes) data cubes with high fidelity. This contribution proposes an alternative implementation of one such 3D prototype algorithm, MUFFIN (MUlti-Frequency image reconstruction For radio INterferometry), which combines spatial and spectral analysis priors. Using a recently proposed primal dual algorithm, this new version of MUFFIN allows a parallel implementation where comput…
This work presents a method for determining the accuracy of a source finder algorithm for spectral line radio astronomy data and the Source Finder Accuracy Evaluator (SFAE), a program that implements this method. The accuracy of a source finder is defined in terms of its completeness, reliability, and accuracy of the parameterisation of the sources that were found. These values are calculated by executing the source finder on an image with a known source catalogue, then comparing the output of the source finder to the known catalogue. The intended uses of SFAE include determining the most accu…
In this lecture, we describe a number of advanced gain calibration techniques. In particular, self-calibration is an important tool in interferometric imaging at all wavelengths. It allows the observer to determine and remove residual phase and amplitude errors that remain in the data after normal calibration while simultaneously producing a more accurate and more sensitive image of the target source. We describe the basic technique of self-calibration and attempt to dispel some common misconceptions. We proceed to give a range of useful tips, and provide continuum, spectral line, and mosaic s…
MeerKATHI is the current development name for a radio-interferometric data reduction pipeline, assembled by an international collaboration. We create a publicly available end-to-end continuum- and line imaging pipeline for MeerKAT and other radio telescopes. We implement advanced techniques that are suitable for producing high-dynamic-range continuum images and spectroscopic data cubes. Using containerization, our pipeline is platform-independent. Furthermore, we are applying a standardized approach for using a number of different of advanced software suites, partly developed within our group.…
Bayesian component separation and power spectrum estimation for 21 cm intensity mapping data cubes
University of Western Cape（西开普敦大学）；Jodrell Bank Centre for Astrophysics, University of Manchester（曼彻斯特大学乔德雷克天体物理学中心）；South African Radio Astronomy Observatory（南非射电天文台）；Institute of Cosmology and Gravitation, University of Portsmouth（朴茨茅斯宇宙学与引力研究所）
The new widefield radio telescopes, such as: ASKAP,MWA, LOFAR, eVLA and SKA; will produce spectral-imaging data-cubes (SIDC) of unprecedented volumes in the order of hundreds of Petabytes. Servicing such data as images to the end-user may encounter challenges unforeseen during the development of IVOA SIAP. We discuss the requirements for extremely large SIDC, and in this light we analyse the applicability of approach taken in the ISO/IEC 15444 (JPEG2000) standards.
Advances in astronomy are intimately linked to advances in digital signal processing (DSP). This special issue is focused upon advances in DSP within radio astronomy. The trend within that community is to use off-the-shelf digital hardware where possible and leverage advances in high performance computing. In particular, graphics processing units (GPUs) and field programmable gate arrays (FPGAs) are being used in place of application-specific circuits (ASICs); high-speed Ethernet and Infiniband are being used for interconnect in place of custom backplanes. Further, to lower hurdles in digital …
Enhancing interoperable data access to radio data has become a science priority within the International Virtual Observatory Alliance (IVOA). This lead to the foundation of the IVOA Radio astronomy Interest Group. Several radio astronomers and project scientists enrolled in various projects (NRAO, ASKAP, LOFAR, JIVE, ALMA, SKA, INAF, NenuFAR, etc.) have joined. Together they are paving the way to a better integration of their services in the virtual observatory (VO) infrastructure and propose extension of IVOA standards to help achieving this goal. Calibrated radio datasets such as cubes, imag…
The coming years will see routine use of solar data of unprecedented spatial and spectral resolution, time cadence, and completeness in the wavelength domain. To capitalize on the soon to be available radio facilities such as the expanded OVSA, SSRT and FASR, and the challenges they present in the visualization and synthesis of the multi-frequency datasets, we propose that realistic, sophisticated 3D active region and flare modeling is timely now and will be a forefront of coronal studies over the coming years. Here we summarize our 3D modeling efforts, aimed at forward fitting of imaging spec…
Fossil record methods based on spectral synthesis techniques have matured over the past decade, and their application to integrated galaxy spectra fostered substantial advances on the understanding of galaxies and their evolution. Yet, because of the lack of spatial resolution, these studies are limited to a global view, providing no information about the internal physics of galaxies. Motivated by the CALIFA survey, which is gathering Integral Field Spectroscopy over the full optical extent of 600 galaxies, we have developed an end-to-end pipeline which: (i) partitions the observed data cube i…
For decades now, scientific data volumes have experienced relentless, exponential growth. As a result, legacy astronomical data formats are straining under a burden not conceived when these formats were first introduced. With future astronomical projects ensuring this trend, ASTRON and the LOFAR project are exploring the use of the Hierarchical Data Format, version 5 (HDF5), for LOFAR radio data encapsulation. Most of LOFAR's standard data products will be stored using the HDF5 format. In addition, HDF5 analogues for traditional radio data structures such as visibility data and spectral image …
CASA, the Common Astronomy Software Applications package, is the primary data processing software for the Atacama Large Millimeter/submillimeter Array (ALMA) and NSF's Karl G. Jansky Very Large Array (VLA), and is frequently used also for other radio telescopes. The CASA software can process data from both single-dish and aperture-synthesis telescopes, and one of its core functionalities is to support the data reduction and imaging pipelines for ALMA, VLA and the VLA Sky Survey (VLASS). CASA has recently undergone several exciting new developments, including an increased flexibility in Python …
iDaVIE v1.0: A virtual reality tool for interactive analysis of astronomical data cubes
Inter-University Institute for Data Intensive Astronomy, University of Cape Town（开普敦大学数据密集型天文学跨大学研究所）；Department of Astronomy, University of Cape Town（开普敦大学天文系）；Armagh Observatory and Planetarium（阿马赫天文台和天文馆）；INAF - Istituto di Radioastronomia（意大利国家天体物理研究所射电天文研究所）；Astrophysical Observatory of Catania, National Institute for Astrophysics（卡塔尼亚天体物理观测站，意大利国家天体物理研究所）；Kapteyn Astronomical Institute, University of Groningen（格罗宁根大学卡普坦天文研究所）
With the advent of large scale surveys the manual analysis and classification of individual radio source morphologies is rendered impossible as existing approaches do not scale. The analysis of complex morphological features in the spatial domain is a particularly important task. Here we discuss the challenges of transferring crowdsourced labels obtained from the Radio Galaxy Zoo project and introduce a proper transfer mechanism via quantile random forest regression. By using parallelized rotation and flipping invariant Kohonen-maps, image cubes of Radio Galaxy Zoo selected galaxies formed fro…
Flux Cube Reconstruction from Slitless Spectroscopy
Space Telescope Science Institute（太空望远镜科学研究所）；E.O. Lawrence Berkeley National Laboratory（欧内斯特·奥兰多·劳伦斯伯克利国家实验室）；Department of Physics, University of California Berkeley（加州大学伯克利分校物理系）；Department of Physics and Astronomy, University of Hawai‘i at Mānoa（夏威夷大学马诺阿分校物理与天文系）
We present the analysis of radio interferometric 2-s images from a MeerKAT observation of the repeating fast radio burst FRB121102 on September 2019, during which 11 distinct pulses have been previously detected using high time and frequency resolution data cubes. In this work, we detected 6 out of the 11 bursts in the image plane at 1.48 GHz with a minimum peak signal-to-noise ratio (S/N) of 5 σ and a fluence detection limit of 0.512 Jy ms. These constitute the first detections of a fast radio burst (FRB) or a radio transient using 2-s timescale images with MeerKAT data. Analysis of the fitte…
We present GALARIO, a computational library that exploits the power of modern graphical processing units (GPUs) to accelerate the analysis of observations from radio interferometers like ALMA or the VLA. GALARIO speeds up the computation of synthetic visibilities from a generic 2D model image or a radial brightness profile (for axisymmetric sources). On a GPU, GALARIO is 150 faster than standard Python and 10 times faster than serial C++ code on a CPU. Highly modular, easy to use and to adopt in existing code, GALARIO comes as two compiled libraries, one for Nvidia GPUs and one for multicore C…
A discussion of the history of scientific computing for Radio Astronomy in the Cavendish Laboratory of the University of Cambridge in the decades after the Second World War. This covers the development of the aperture synthesis technique for Radio Astronomy and how that required using the new computing technology developed by the University's Mathematical Laboratory: the EDSAC, EDSAC 2 and TITAN computers. It looks at the scientific advances made by the Radio Astronomy group, particularly the assembling of evidence which contradicted the Steady State Hypothesis. It also examines the software a…
