Deep convolutional neural networks have been a popular tool for image generation and restoration. The performance of these networks is related to the capability of learning realistic features from a large dataset. In this work, we applied the problem of inpainting non-Gaussian signal, in the context of Galactic diffuse emissions at the millimetric and sub-millimetric regimes, specifically Synchrotron and Thermal Dust emission. Both of them are affected by contamination at small angular scales due to extra-galactic radio sources (the former) and to dusty star-forming galaxies (the latter). We consider the performances of a nearest-neighbors inpainting technique and compare it with two novels methodologies relying on generative Neural Networks. We show that the generative network is able to reproduce the statistical properties of the ground truth signal more consistently with high confidence level. The Python Inpainter for Cosmological and AStrophysical SOurces (PICASSO) is a package encoding a suite of inpainting methods described in this work and has been made publicly available.
We address the problem of line confusion in intensity mapping surveys and explore the possibility to mitigate line foreground contamination by progressively masking the brightest pixels in the observed map. We consider experiments targeting CO(1-0) at $z=3$, Ly$α$ at $z=7$, and CII at $z=7$, and use simulated intensity maps, which include both clustering and shot noise components of the signal and possible foregrounds, in order to test the efficiency of our method. We find that for CO and Ly$α$ it is quite possible to remove most of the foreground contribution from the maps via only 1%-3% pixe…
21cm intensity mapping experiments aim to observe the diffuse neutral hydrogen (HI) distribution on large scales which traces the Cosmic structure. The Square Kilometre Array (SKA) will have the capacity to measure the 21cm signal over a large fraction of the sky. However, the redshifted 21cm signal in the respective frequencies is faint compared to the Galactic foregrounds produced by synchrotron and free-free electron emission. In this article, we review selected foreground subtraction methods suggested to effectively separate the 21cm signal from the foregrounds with intensity mapping simul…
The intensity mapping of Ly-alpha emission during the epoch of reionization (EoR) will be contaminated by foreground emission lines from lower redshifts. We calculate the mean intensity and power spectrum of Ly-alpha emission at z~7, and estimate the uncertainties according to the relevant astrophysical processes. We find that the low-redshift emission lines from 6563 A H-alpha, 5007 A [OIII] and 3727 A [OII] will be the strong contaminants on the observed Ly-alpha power spectrum. We make use of both the star formation rate (SFR) and luminosity functions (LF) to estimate the mean intensity and…
We make use of a large set of fast simulations of an intensity mapping experiment with characteristics similar to those expected of the Square Kilometre Array (SKA) in order to study the viability and limits of blind foreground subtraction techniques. In particular, we consider different approaches: polynomial fitting, principal component analysis (PCA) and independent component analysis (ICA). We review the motivations and algorithms for the three methods, and show that they can all be described, using the same mathematical framework, as different approaches to the blind source separation pro…
We apply for the first time Gaussian Process Regression (GPR) as a foreground removal technique in the context of single-dish, low redshift HI intensity mapping, and present an open-source Python toolkit for doing so. We use MeerKAT and SKA1-MID-like simulations of 21cm foregrounds (including polarisation leakage), HI cosmological signal and instrumental noise. We find that it is possible to use GPR as a foreground removal technique in this context, and that it is better suited in some cases to recover the HI power spectrum than Principal Component Analysis (PCA), especially on small scales. G…
Generalised Visibility ILC: a new foreground mitigation strategy for interferometric HI intensity mapping in the low-redshift Universe
Instituto de Física de Cantabria (CSIC-Universidad de Cantabria)（坎塔布里亚物理研究所）；Dpto. de Física Moderna, Universidad de Cantabria（坎塔布里亚大学现代物理系）；Jodrell Bank Centre for Astrophysics, Department of Physics and Astronomy, The University of Manchester（曼彻斯特大学天体物理学乔德雷尔银行中心）；State Key Laboratory of Radio Astronomy and Technology, Shanghai Astronomical Observatory, CAS（中国科学院上海天文台射电天文与技术国家重点实验室）
The high dynamic range between contaminating foreground emission and the fluctuating 21cm brightness temperature field is one of the most problematic characteristics of 21cm intensity mapping data. While these components would ordinarily have distinctive frequency spectra, making it relatively easy to separate them, instrumental effects and calibration errors further complicate matters by modulating and mixing them together. A popular class of foreground cleaning method are unsupervised techniques related to Principal Component Analysis (PCA), which exploit the different shapes and amplitudes …
We model a 21 cm intensity mapping survey in the redshift range 0.01<z<1.5 designed to simulate the skies as seen by future radio telescopes such as the Square Kilometre Array (SKA), including instrumental noise and Galactic foregrounds. In our pipeline, we remove the introduced Galactic foregrounds with a fast independent component analysis (fastica) technique. We present the power spectrum of the large-scale matter distribution, C(l), before and after the application of this foreground removal method and calculate the resulting systematic errors. We attempt to reduce systematics in the foreg…
The future of precision cosmology could benefit from cross-correlations between intensity maps of unresolved neutral hydrogen (HI) and more conventional optical galaxy surveys. A major challenge that needs to be overcome is removing the 21cm foreground emission that contaminates the cosmological HI signal. Using N-body simulations we simulate HI intensity maps and optical catalogues which share the same underlying cosmology. Adding simulated foreground contamination and using state-of-the-art reconstruction techniques we investigate the impacts that 21cm foregrounds and other systematics have …
Potential evidence for primordial non-Gaussianity (PNG) is expected to lie in the largest scales mapped by cosmological surveys. Forthcoming 21cm intensity mapping experiments will aim to probe these scales by surveying neutral hydrogen (HI) within galaxies. However, foreground signals dominate the faint 21cm emission, meaning foreground cleaning is required to recover the cosmological signal. The effect this has is to damp the HI power spectrum on the largest scales, especially along the line-of-sight. Whilst there is agreement that this contamination is potentially problematic for probing PN…
Neutral Hydrogen Intensity Mapping (HI IM) surveys will be a powerful new probe of cosmology. However, strong astrophysical foregrounds contaminate the signal and their coupling with instrumental systematics further increases the data cleaning complexity. In this work, we simulate a realistic single-dish HI IM survey of a $5000$~deg$^2$ patch in the $950 - 1400$ MHz range, with both the MID telescope of the SKA Observatory (SKAO) and MeerKAT, its precursor. We include a state-of-the-art HI simulations and explore different foreground models and instrumental effects such as non-homogeneous ther…
We conduct the first case study towards developing optimal foreground mitigation strategies for neutral hydrogen (HI) intensity mapping using radio interferometers at low redshifts. A pipeline for simulation, foreground mitigation and power spectrum estimation is built, which can be used for ongoing and future surveys using MeerKAT and Square Kilometre Array Observatory (SKAO). It simulates realistic sky signals to generate visibility data given instrument and observation specifications, which is subsequently used to perform foreground mitigation and power spectrum estimation. A quadratic esti…
Measurements of the one-point probability distribution function and higher-order moments (variance, skewness, and kurtosis) of the high-redshift 21 cm fluctuations are among the most direct statistical probes of the non-Gaussian nature of structure formation and evolution during reionization. However, contamination from astrophysical foregrounds and instrument systematics pose significant challenges in measuring these statistics in real observations. In this work, we use forward modelling to investigate the feasibility of measuring 21 cm one-point statistics through a foreground avoidance stra…
21 cm intensity mapping has arisen as a powerful probe of the high-redshift universe, but its potential is limited by extremely bright foregrounds and high source confusion. In this Letter, we propose a new analysis which can help solve both problems. From the combination of an intensity map with an overlapping galaxy survey we construct a new one-point statistic which is unbiased by foregrounds and contains information left out of conventional analyses. We show that our method can measure the HI mass function with unprecedented precision using observations similar to recent 21 cm detections.
Intensity mapping experiments survey the spectrum of diffuse line radiation rather than detect individual objects at high signal-to-noise. Spectral maps of unresolved atomic and molecular line radiation contain three-dimensional information about the density and environments of emitting gas, and efficiently probe cosmological volumes out to high redshift. Intensity mapping survey volumes also contain all other sources of radiation at the frequencies of interest. Continuum foregrounds are typically ~10^2-10^3 times brighter than the cosmological signal. The instrumental response to bright foreg…
The Principal Component Analysis (PCA) method and the Singular Value Decomposition (SVD) method are widely used for foreground subtraction in 21 cm intensity mapping experiments. We show their equivalence, and point out that the condition for completely clean separation of foregrounds and cosmic 21 cm signal using the PCA/SVD is unrealistic. We propose a PCA-based foreground subtraction method, dubbed "Singular Vector Projection (SVP)" method, which exploits a priori information of the left and/or right singular vectors of the foregrounds. We demonstrate with simulation tests that this new, se…
We forecast ability of dedicated 21 cm intensity mapping experiments to constraint primordial non-Gaussianity using power spectrum and bispectrum. We model the signal including the non-linear biasing expansion using a generalized halo model approach. We consider the importance of foreground filtering scale and of the foreground wedge. We find that the current generation intensity mapping experiments like CHIME do not posses sufficient sensitivity to be competitive with the existing limits. On the other hand, upcoming experiments like HIRAX can improve the current constraints and the proposed P…
Foreground removal in HI 21 cm intensity mapping under frequency-dependent beam distortions
Foundation for Research and Technology Hellas (FORTH)（希腊研究与技术基金会）；Université Paris-Saclay（巴黎萨克雷大学）；Université Paris Cité（巴黎西岱大学）；CEA（法国原子能和替代能源委员会）；CNRS（法国国家科学研究中心）；Observatoire de la Côte d’Azur（蔚蓝海岸天文台）；University of the Western Cape（西开普敦大学）；University of Crete（克里特大学）
We present a framework and an open-source python toolkit to analyse the 2-point statistics of 3D fluctuations in the context of HI intensity maps using the multipole expansion formalism. We include simulations of the cosmological HI signal using N-body and log-normal methods, foregrounds and their removal, as well as instrumental effects. Using these simulations and analytical modelling, we investigate the impact of foreground cleaning and the instrumental beam on the power spectrum multipoles as well as on the Fourier space clustering wedges. We find that both the instrumental beam and the fo…
\textsc{Hi} intensity mapping will provide maps of the large-scale distribution of neutral hydrogen (\textsc{Hi}) in the universe. These are prime candidates to be used to constrain primordial non-Gaussianity using the Large Scale Structure of the Universe as well as to provide further tests of Einstein's theory of Gravity (GR). But \textsc{Hi} maps are contaminated by foregrounds, which can be several orders of magnitude above the cosmological signal. Here we quantify how degenerated are the large-scale effects ($f_{\rm NL}$ and GR effects) with the residual foregrounds. We conclude that a jo…
Intensity mapping has attracted significant interest as an approach to measure the properties of the interstellar medium in typical galaxies at high redshift. Intensity mapping measures the statistics of surface brightness as a function of frequency, making it sensitive not only to all line emission of interest but also radiation from all other sources. Significant effort has gone into developing approaches that reject foreground contamination. Additionally, the target galaxies have multiple sources of emission that can complicate the interpretation of the line brightness. We describe the prob…
We provide perturbation theory predictions for the HI intensity mapping power spectrum multipoles using the Effective Field Theory of Large Scale Structure (EFTofLSS), which should allow us to constrain cosmological parameters exploiting mildly nonlinear scales. Assuming survey specifications typical of proposed interferometric HI intensity mapping experiments like CHORD and PUMA, and realistic ranges of validity for the perturbation theory modelling, we run mock full shape MCMC analyses at a redshift bin centred at $z=0.5$, and compare with Stage-IV optical galaxy surveys. We include the impa…
A detection of the predicted anticorrelation between 21cm and either Ly-alpha or H-alpha from the Epoch of Reionization (EOR) would be a powerful probe of the first galaxies. While 3D intensity maps isolate foregrounds in low k_\parallel modes, infrared surveys cannot yet match the field of view and redshift resolution of radio intensity mapping experiments. In contrast, 2D (i.e., broad band) infrared intensity maps can be measured with current experiments and are limited by foregrounds instead of photon or thermal noise. We show 2D experiments can measure most of the 3D fluctuation power at k…
Intensity mapping of the HI 21 cm line and the CO 2.61 mm line from the epoch of reionization has emerged as powerful, complementary, probes of the high-redshift Universe. However, both maps and their cross-correlation are dominated by foregrounds. We propose a new analysis by which the signal is unbiased by foregrounds, i.e. it can be measured without foreground mitigation. We construct the antisymmetric part of the HI-CO cross-correlation, arising because the statistical fluctuations of two fields have different evolution in time. We show that the sign of this new signal can distinguish mode…
We present a novel application of partial convolutional neural networks (PCNN) that can inpaint masked images of the cosmic microwave background. The network can reconstruct both the maps and the power spectra to a few percent for circular and irregularly shaped masks covering up to ~10% of the image area. By performing a Kolmogorov-Smirnov test we show that the reconstructed maps and power spectra are indistinguishable from the input maps and power spectra at the 99.9% level. Moreover, we show that PCNNs can inpaint maps with regular and irregular masks to the same accuracy. This should be pa…
Line intensity mapping (LIM) is a promising probe to study star formation, the large-scale structure of the Universe, and the epoch of reionization (EoR). Since carbon monoxide (CO) is the second most abundant molecule in the Universe except for molecular hydrogen ${\rm H}_2$, it is suitable as a tracer for LIM surveys. However, just like other LIM surveys, CO intensity mapping also suffers strong foreground contamination that needs to be eliminated for extracting valuable astrophysical and cosmological information. In this work, we take $^{12}$CO($\it J$=1-0) emission line as an example to in…
We introduce DAYENU, a linear, spectral filter for HI intensity mapping that achieves the desirable foreground mitigation and error minimization properties of inverse co-variance weighting with minimal modeling of the underlying data. Beyond 21 cm power-spectrum estimation, our filter is suitable for any analysis where high dynamic-range removal of spectrally smooth foregrounds in irregularly (or regularly) sampled data is required, something required by many other intensity mapping techniques. Our filtering matrix is diagonalized by Discrete Prolate Spheroidal Sequences which are an optimal b…
We present the first application of a new foreground removal pipeline to the current leading HI intensity mapping dataset, obtained by the Green Bank Telescope (GBT). We study the 15hr and 1hr field data of the GBT observations previously presented in Masui et al (2013) and Switzer et al (2013), covering about 41 square degrees at 0.6<z<1.0, for which cross-correlations may be measured with the galaxy distribution of the WiggleZ Dark Energy Survey. In the presented pipeline, we subtract the Galactic foreground continuum and the point source contamination using an independent component analysis…
We assess the performance of the multipole expansion formalism in the case of single-dish HI intensity mapping, including instrumental and foreground removal effects. This formalism is used to provide MCMC forecasts for a range of HI and cosmological parameters, including redshift space distortions and the Alcock-Paczynski effect. We first determine the range of validity of our power spectrum modelling by fitting to simulation data, concentrating on the monopole, quadrupole, and hexadecapole contributions. We then show that foreground subtraction effects can lead to severe biases in the determ…
HI intensity mapping is an emerging tool to probe dark energy. Observations of the redshifted HI signal will be contaminated by instrumental noise, atmospheric and Galactic foregrounds. The latter is expected to be four orders of magnitude brighter than the HI emission we wish to detect. We present a simulation of single-dish observations including an instrumental noise model with 1/f and white noise, and sky emission with a diffuse Galactic foreground and HI emission. We consider two foreground cleaning methods: spectral parametric fitting and principal component analysis. For a smooth freque…
We present a code to generate mock observations of 21 cm intensity mapping experiments. The emphasis of the code is on reducing the computational cost of running a full-blown simulation, trading computational time for accuracy. The code can be used to generate independent realizations of the cosmological signal and foregrounds, which are necessary, for instance, in order to obtain realistic forecasts for future intensity mapping experiments. The code is able to reproduce the correct angular and radial clustering pattern for the cosmological signal, including redshift-space distortions, lightco…
We use full sky simulations, including the effects of foreground contamination and removal, to explore multi-tracer synergies between a SKA-like 21cm intensity mapping survey and a LSST-like photometric galaxy redshift survey. In particular we study ratios of auto and cross-correlations between the two tracers as estimators of the ratio of their biases, a quantity that should benefit considerably from the cosmic variance cancellation of the multi-tracer approach. We show how well we should be able to measure the bias ratio on very large scales (down to $\ell \sim 3$), which is crucial to measu…
The post-reionization $(z \le 6)$ neutral hydrogen (HI) 21-cm intensity mapping signal holds the potential to probe the large scale structures, study the expansion history and constrain various cosmological parameters. Here we apply the Tapered Gridded Estimator (TGE) to estimate $P(k_{\perp},k_{\parallel})$ the power spectrum of the $z = 2.28$ $(432.8\, {\rm MHz})$ redshifted 21-cm signal using a $24.4\,{\rm MHz}$ sub-band drawn from uGMRT Band 3 observations of European Large-Area ISO Survey-North 1 (ELAIS-N1). The TGE allows us to taper the sky response which suppresses the foreground contr…
We quantitatively investigate the possibility of detecting baryonic acoustic oscillations (BAO) using single-dish 21cm intensity mapping observations in the post-reionization era. We show that the telescope beam smears out the isotropic BAO signature and, in the case of the Square Kilometer Array (SKA) instrument, makes it undetectable at redshifts $z\gtrsim1$. We however demonstrate that the BAO peak can still be detected in the radial 21cm power spectrum and describe a method to make this type of measurements. By means of numerical simulations, containing the 21cm cosmological signal as well…
The cosmic microwave background $B$-mode signal is potentially weaker than the diffuse Galactic foregrounds over most of the sky at any frequency. A common method of separating the CMB from these foregrounds is via pixel-based parametric-model fitting. There are not currently enough all-sky maps to fit anything more than the most simple models of the sky. By simulating the emission in seven representative pixels, we demonstrate that the inclusion of a 5 GHz data point allows for more complex models of low-frequency foregrounds to be fitted than at present. It is shown that the inclusion of the…
CMB foregrounds consist of all radiation between the surface of last scattering and the detectors, which can interfere with the cosmological interpretation of CMB data. Fortunately, in temperature (intensity), even though the foregrounds are complex they can relatively easily be mitigated. However, in polarization, diffuse Galactic radiation (synchrotron and thermal dust) can be polarized at a level of >10 % making it more of a challenge. In particular, CMB B-modes, which are a smoking-gun signature of inflation, will be dominated by foregrounds. Component separation will therefore be critical…
The bispectrum is a 3-point statistic with the potential to provide additional information beyond power spectra analyses of survey datasets. Radio telescopes which broadly survey the 21cm emission from neutral hydrogen (HI) are a promising way to probe LSS and in this work we present an investigation into the HI intensity mapping (IM) bispectrum using simulations. We present a model of the redshift space HI IM bispectrum including observational effects from the radio telescope beam and 21cm foreground contamination. We validate our modelling prescriptions with measurements from robust IM simul…
The success of HI intensity mapping is largely dependent on how well 21cm foreground contamination can be controlled. In order to progress our understanding further, we present a range of simulated foreground data from four different $\sim3000$\,deg$^2$ sky regions, with and without effects from polarization leakage. Combining these with underlying cosmological HI simulations creates a range of single-dish intensity mapping test cases that require different foreground treatments. This allows us to conduct the most generalized study to date into 21cm foregrounds and their cleaning techniques fo…
We discuss the possibility of studying diffuse baryon distributions with kinematic Sunyaev-Zel'dovich (kSZ) effect by correlating cosmic microwave background (CMB) temperature fluctuations with density fluctuations from 21\,cm intensity mapping (IM). The biggest challenge for the cross-correlation is the loss of large-scale information in IM, due to foregrounds and the zero spacing problem of interferometers. We apply the tidal reconstruction algorithm to restore the lost large-scale modes, which increases the correlation by more than a factor of three. With the predicted foreground level, we …
Upcoming and future neutral hydrogen Intensity Mapping surveys offer a great opportunity to constrain cosmology in the post-reionization Universe, provided a good accuracy is achieved in the separation between the strong foregrounds and the cosmological signal. Cleaning methods rely on the frequency smoothness of the foregrounds and are often applied under the assumption of a simplistic Gaussian primary beam. In this work, we test the cleaning in the presence of a realistic primary beam model with a non trivial frequency dependence. We focus on the Square Kilometre Array precursor MeerKAT tele…
We demonstrate how cosmic star-formation history can be measured with one-point statistics of carbon-monoxide intensity maps. Using a P(D) analysis, the luminosity function of CO-emitting sources can be inferred from the measured one-point intensity PDF. The star-formation rate density (SFRD) can then be obtained, at several redshifts, from the CO luminosity density. We study the effects of instrumental noise, line foregrounds, and target redshift, and obtain constraints on the CO luminosity density of order 10%. We show that the SFRD uncertainty is dominated by that of the model connecting CO…
The large-scale distribution of neutral hydrogen in the Universe will be luminous through its 21 cm emission. Here, for the first time, we use the auto-power spectrum of 21 cm intensity fluctuations to constrain neutral hydrogen fluctuations at z~0.8. Our data were acquired with the Green Bank Telescope and span the redshift range 0.6 < z < 1 over two fields totalling ~41 deg. sq. and 190 h of radio integration time. The dominant synchrotron foregrounds exceed the signal by ~10^3, but have fewer degrees of freedom and can be removed efficiently. Even in the presence of residual foregrounds, th…
The 21-cm and Lyman Alpha lines are the dominant line-emission spectral features at opposite ends of the spectrum of hydrogen. Each line can be used to create three dimensional intensity maps of large scale structure. The sky brightness at low redshift due to Lyman Alpha emission is estimated to be 0.4 Jy/Steradian, which is brighter than the zodiacal light foreground.
The cross-correlation of a foreground density field with two different background convergence fields can be used to measure cosmographic distance ratios and constrain dark energy parameters. We investigate the possibility of performing such measurements using a combination of optical galaxy surveys and HI intensity mapping surveys, with emphasis on the performance of the planned Square Kilometre Array (SKA). Using HI intensity mapping to probe the foreground density tracer field and/or the background source fields has the advantage of excellent redshift resolution and a longer lever arm achiev…
The 21-cm intensity mapping (IM) of neutral hydrogen (HI) is a promising tool to probe the large-scale structures. Sky maps of 21-cm intensities can be highly contaminated by different foregrounds, such as Galactic synchrotron radiation, free-free emission, extragalactic point sources, and atmospheric noise. We here present a model of foreground components and a method of removal, especially to quantify the potential of Five-hundred-meter Aperture Spherical radio Telescope (FAST) for measuring HI IM. We consider 1-year observational time with the survey area of $20,000\,{\rm deg}^{2}$ to captu…
Methodological Frontiers in 21-cm Intensity Mapping: the Treatment of Systematics and Foreground Contamination
Observatoire de la Côte d’Azur（蔚蓝海岸天文台）；University of the Western Cape（西开普敦大学）；Indian Institute of Technology Kharagpur（印度理工学院卡拉格普尔分校）；Foundation for Research and Technology Hellas (FORTH)（希腊研究与技术基金会）；The University of Manchester（曼彻斯特大学）；The University of Edinburgh（爱丁堡大学）；Indian Institute of Technology Madras（印度理工学院马德拉斯分校）；ETH Zurich（苏黎世联邦理工学院）；University of Portsmouth（朴茨茅斯大学）；Banwarilal Bhalotia College（班瓦里拉尔·巴洛蒂亚学院）；University of Cambridge（剑桥大学）；Northeastern University（东北大学）；Chinese Academy of Sciences（中国科学院）
Line intensity maps (LIMs) are in principle sensitive to a large amount of information about faint, distant galaxies which are invisible to conventional surveys. However, actually extracting that information from a confused, foreground-contaminated map can be challenging. In this work we present the first application of convolutional neural network (CNN) to directly determine the underlying luminosity function of a LIM, including a treatment of extragalactic foregrounds and instrumental noise. We apply the CNN to simulations of mock Carbon Monoxide (CO) line intensity maps similar to those whi…
HI intensity mapping (IM) is an exciting new probe that could revolutionize the future of cosmology. However, the relative faintness of the HI signal when compared to foregrounds of astrophysical or terrestrial origin will make HI IM extremely challenging. The imprint of these foregrounds may result in systematic errors in the recovered cosmological signal. We discuss an IM simulation pipeline developed at Manchester that can introduce systematic errors at the TOD level in order to help assess their impact. We will present results for two potential sources of systematics for HI IM surveys: 1/f…
Mitigating point-source contamination in CMB polarization: a Generalized Point Spread Function fitting approach
Key Laboratory of Particle Astrophysics, Institute of High Energy Physics, Chinese Academy of Sciences（高能物理研究所粒子天体物理重点实验室）；University of Chinese Academy of Sciences（中国科学院大学）
Understanding large-angular-scale galactic foregrounds is crucial for future CMB experiments aiming to detect $B$-mode polarization from primordial gravitational waves. Traditionally, the dust component has been separated using its different frequency dependence. However, using non-CMB observations has potential to increase fidelity and decrease the reconstruction noise. In this exploratory paper we investigate the capability of galactic 21 cm observations to predict the dust foreground in intensity. We train a neural network to predict the dust foreground as measured by the Planck Satellite f…
JESSNet: Joint dEconvolution and Sparse Separation Network for HI 21 cm foreground cleaning
Foundation for Research and Technology Hellas (FORTH)（希腊基金会）；Université Paris-Saclay（巴黎萨克雷大学）；Université Paris Cité（巴黎西岱大学）；CEA（法国原子能和替代能源委员会）；CNRS（法国国家科学研究中心）；University of Crete（克里特大学）；Observatoire de la Côte d’Azur（蔚蓝海岸天文台）；Laboratoire Lagrange（拉格朗日实验室）；University of the Western Cape（西开普大学）；Institute of Computer Science, Foundation for Research and Technology Hellas (FORTH)（希腊基金会计算机科学研究所）
