-
DanceCam: atmospheric turbulence mitigation in wide-field astronomical images with short-exposure video streams
Authors:
Spencer Bialek,
Emmanuel Bertin,
Sébastien Fabbro,
Hervé Bouy,
Jean-Pierre Rivet,
Olivier Lai,
Jean-Charles Cuillandre
Abstract:
We introduce a novel technique to mitigate the adverse effects of atmospheric turbulence on astronomical imaging. Utilizing a video-to-image neural network trained on simulated data, our method processes a sliding sequence of short-exposure ($\sim$0.2s) stellar field images to reconstruct an image devoid of both turbulence and noise. We demonstrate the method with simulated and observed stellar fi…
▽ More
We introduce a novel technique to mitigate the adverse effects of atmospheric turbulence on astronomical imaging. Utilizing a video-to-image neural network trained on simulated data, our method processes a sliding sequence of short-exposure ($\sim$0.2s) stellar field images to reconstruct an image devoid of both turbulence and noise. We demonstrate the method with simulated and observed stellar fields, and show that the brief exposure sequence allows the network to accurately associate speckles to their originating stars and effectively disentangle light from adjacent sources across a range of seeing conditions, all while preserving flux to a lower signal-to-noise ratio than an average stack. This approach results in a marked improvement in angular resolution without compromising the astrometric stability of the final image.
△ Less
Submitted 8 May, 2024;
originally announced May 2024.
-
StarUnLink: identifying and mitigating signals from communications satellites in stellar spectral surveys
Authors:
Spencer Bialek,
Sara Lucatello,
Sebastien Fabbro,
Kwang Moo Yi,
Kim Venn
Abstract:
A relatively new concern for the forthcoming massive spectroscopic sky surveys is the impact of contamination from low earth orbit satellites. Several hundred thousand of these satellites are licensed for launch in the next few years and it has been estimated that, in some cases, up to a few percent of spectra could be contaminated when using wide field, multi-fiber spectrographs. In this paper, a…
▽ More
A relatively new concern for the forthcoming massive spectroscopic sky surveys is the impact of contamination from low earth orbit satellites. Several hundred thousand of these satellites are licensed for launch in the next few years and it has been estimated that, in some cases, up to a few percent of spectra could be contaminated when using wide field, multi-fiber spectrographs. In this paper, a multi-staged approach is used to assess the practicality and limitations of identifying and minimizing the impact of satellite contamination in a WEAVE-like stellar spectral survey. We develop a series of convolutional-network based architectures to attempt identification, stellar parameter and chemical abundances recovery, and source separation of stellar spectra that we artificially contaminate with satellite (i.e. solar-like) spectra. Our results show that we are able to flag 67% of all contaminated sources at a precision level of 80% for low-resolution spectra and 96% for high-resolution spectra. Additionally, we are able to remove the contamination from the spectra and recover the clean spectra with a $<$1% reconstruction error. The errors in stellar parameter predictions reduce by up to a factor of 2-3 when either including contamination as an augmentation to a training set or by removing the contamination from the spectra, with overall better performance in the former case. The presented methods illustrate several machine learning mitigation strategies that can be implemented to improve stellar parameters for contaminated spectra in the WEAVE stellar spectroscopic survey and others like it.
△ Less
Submitted 28 June, 2023;
originally announced June 2023.
-
The Pristine survey -- XII: Gemini-GRACES chemo-dynamical study of newly discovered extremely metal-poor stars in the Galaxy
Authors:
Collin L. Kielty,
Kim A. Venn,
Federico Sestito,
Else Starkenburg,
Nicolas F. Martin,
David S. Aguado,
Anke Arentsen,
Sébastien Fabbro,
Jonay I. González Hernández,
Vanessa Hill,
Pascale Jablonka,
Carmela Lardo,
Lyudmila I. Mashonkina,
Julio F. Navarro,
Chris Sneden,
Guillaume F. Thomas,
Kris Youakim,
Spencer Bialek,
Rubén Sánchez-Janssen
Abstract:
High-resolution optical spectra of 30 metal-poor stars selected from the \Pristine\ survey are presented, based on observations taken with the Gemini Observatory GRACES spectrograph. Stellar parameters \teff and \logg are determined using a Gaia DR2 colour-temperature calibration and surface gravity from the Stefan-Boltzmann equation. GRACES spectra are used to determine chemical abundances (or up…
▽ More
High-resolution optical spectra of 30 metal-poor stars selected from the \Pristine\ survey are presented, based on observations taken with the Gemini Observatory GRACES spectrograph. Stellar parameters \teff and \logg are determined using a Gaia DR2 colour-temperature calibration and surface gravity from the Stefan-Boltzmann equation. GRACES spectra are used to determine chemical abundances (or upper-limits) for 20 elements (Li, O, Na, Mg, K, Ca, Ti, Sc, Cr, Mn, Fe, Ni, Cu, Zn, Y, Zr, Ba, La, Nd, Eu). These stars are confirmed to be metal-poor ([Fe/H]$<-2.5$), with higher precision than from earlier medium-resolution analyses. The chemistry for most targets is similar to other extremely metal-poor stars in the Galactic halo. Three stars near [Fe/H]$=-3.0$ have unusually low Ca and high Mg, suggestive of contributions from few SN~II where alpha-element formation through hydrostatic nucleosynthesis was more efficient. Three new carbon-enhanced metal-poor stars are also identified (two CEMP-s and one potential CEMP-no star) when our chemical abundances are combined with carbon from previous medium-resolution analyses. The GRACES spectra also provide precision radial velocities ($σ_{\rm RV}\le0.2$km\,s$^{-1}$) for dynamical orbit calculations with the Gaia DR2 proper motions. Most of our targets are dynamically associated with the Galactic halo; however, five stars with [Fe/H]$<-3$ have planar-like orbits, including one retrograde star. Another five stars are dynamically consistent with the Gaia-Sequoia accretion event; three have typical halo [$α$/Fe] ratios for their metallicities, whereas two are [Mg/Fe]-deficient, and one is a new CEMP-s candidate. These results are discussed in terms of the formation and early chemical evolution of the Galaxy.
△ Less
Submitted 12 June, 2021; v1 submitted 7 December, 2020;
originally announced December 2020.
-
Strategic Policymaking for Implementing Renewable Portfolio Standards: A Tri-level Optimization Approach
Authors:
Jip Kim,
Sylwia Bialek,
Burcin Unel,
Yury Dvorkin
Abstract:
Appropriately designed renewable support policies can play a leading role in promoting renewable expansions and contribute to low emission goals. Meanwhile, ill-designed policies may distort electricity markets, put power utilities and generation companies on an unlevel playing field and, in turn, cause inefficiencies. This paper proposes a framework to optimize policymaking for renewable energy s…
▽ More
Appropriately designed renewable support policies can play a leading role in promoting renewable expansions and contribute to low emission goals. Meanwhile, ill-designed policies may distort electricity markets, put power utilities and generation companies on an unlevel playing field and, in turn, cause inefficiencies. This paper proposes a framework to optimize policymaking for renewable energy sources, while incorporating conflicting interests and objectives of different stakeholders. We formulate a tri-level optimization problem where each level represents a different entity: a state regulator, a power utility and a wholesale electricity market. To solve this tri-level problem, we exploit optimality conditions and develop a modification of the Column-and-Cut Generation (C&CG) algorithm that generates cuts for bilinear terms. The case study based on the ISO New England 8-zone test system reveals different policy trade-offs that policymakers face under different decarbonization goals and implementation scenarios.
△ Less
Submitted 11 February, 2021; v1 submitted 2 September, 2020;
originally announced September 2020.
-
Interpreting Stellar Spectra with Unsupervised Domain Adaptation
Authors:
Teaghan O'Briain,
Yuan-Sen Ting,
Sébastien Fabbro,
Kwang M. Yi,
Kim Venn,
Spencer Bialek
Abstract:
We discuss how to achieve map** from large sets of imperfect simulations and observational data with unsupervised domain adaptation. Under the hypothesis that simulated and observed data distributions share a common underlying representation, we show how it is possible to transfer between simulated and observed domains. Driven by an application to interpret stellar spectroscopic sky surveys, we…
▽ More
We discuss how to achieve map** from large sets of imperfect simulations and observational data with unsupervised domain adaptation. Under the hypothesis that simulated and observed data distributions share a common underlying representation, we show how it is possible to transfer between simulated and observed domains. Driven by an application to interpret stellar spectroscopic sky surveys, we construct the domain transfer pipeline from two adversarial autoencoders on each domains with a disentangling latent space, and a cycle-consistency constraint. We then construct a differentiable pipeline from physical stellar parameters to realistic observed spectra, aided by a supplementary generative surrogate physics emulator network. We further exemplify the potential of the method on the reconstructed spectra quality and to discover new spectral features associated to elemental abundances.
△ Less
Submitted 6 July, 2020;
originally announced July 2020.
-
Cycle-StarNet: Bridging the gap between theory and data by leveraging large datasets
Authors:
Teaghan O'Briain,
Yuan-Sen Ting,
Sébastien Fabbro,
Kwang M. Yi,
Kim Venn,
Spencer Bialek
Abstract:
The advancements in stellar spectroscopy data acquisition have made it necessary to accomplish similar improvements in efficient data analysis techniques. Current automated methods for analyzing spectra are either (a) data-driven, which requires prior knowledge of stellar parameters and elemental abundances, or (b) based on theoretical synthetic models that are susceptible to the gap between theor…
▽ More
The advancements in stellar spectroscopy data acquisition have made it necessary to accomplish similar improvements in efficient data analysis techniques. Current automated methods for analyzing spectra are either (a) data-driven, which requires prior knowledge of stellar parameters and elemental abundances, or (b) based on theoretical synthetic models that are susceptible to the gap between theory and practice. In this study, we present a hybrid generative domain adaptation method that turns simulated stellar spectra into realistic spectra by applying unsupervised learning to large spectroscopic surveys. We apply our technique to the APOGEE H-band spectra at R=22,500 and the Kurucz synthetic models. As a proof of concept, two case studies are presented. The first of which is the calibration of synthetic data to become consistent with observations. To accomplish this, synthetic models are morphed into spectra that resemble observations, thereby reducing the gap between theory and observations. Fitting the observed spectra shows an improved average reduced $χ_R^2$ from 1.97 to 1.22, along with a reduced mean residual from 0.16 to -0.01 in normalized flux. The second case study is the identification of the elemental source of missing spectral lines in the synthetic modelling. A mock dataset is used to show that absorption lines can be recovered when they are absent in one of the domains. This method can be applied to other fields, which use large data sets and are currently limited by modelling accuracy. The code used in this study is made publicly available on github.
△ Less
Submitted 13 November, 2020; v1 submitted 6 July, 2020;
originally announced July 2020.
-
The Pristine survey X: a large population of low-metallicity stars permeates the Galactic disk
Authors:
Federico Sestito,
Nicolas F. Martin,
Else Starkenburg,
Anke Arentsen,
Rodrigo A. Ibata,
Nicolas Longeard,
Collin Kielty,
Kristopher Youakim,
Kim A. Venn,
David S. Aguado,
Raymond G. Carlberg,
Jonay I. Gonzalez Hernandez,
Vanessa Hill,
Pascale Jablonka,
Georges Kordopatis,
Khyati Malhan,
Julio F. Navarro,
Ruben Sanchez-Janssen,
Guillame Thomas,
Eline Tolstoy,
Thomas G. Wilson,
Pedro Alonso Palicio,
Spencer Bialek,
Rafael Garcia-Dias,
Romain Lucchesi
, et al. (4 additional authors not shown)
Abstract:
The orbits of the least chemically enriched stars open a window on the formation of our Galaxy when it was still in its infancy. The common picture is that these low-metallicity stars are distributed as an isotropic, pressure-supported component since these stars were either accreted from the early building blocks of the assembling Milky Way, or were later brought by the accretion of faint dwarf g…
▽ More
The orbits of the least chemically enriched stars open a window on the formation of our Galaxy when it was still in its infancy. The common picture is that these low-metallicity stars are distributed as an isotropic, pressure-supported component since these stars were either accreted from the early building blocks of the assembling Milky Way, or were later brought by the accretion of faint dwarf galaxies. Combining the metallicities and radial velocities from the Pristine and LAMOST surveys and Gaia DR2 parallaxes and proper motions for an unprecedented large and unbiased sample of very metal-poor stars at $[Fe/H]\leq-2.5$ we show that this picture is incomplete. This sample shows strong statistical evidence (at the $5.0σ$ level) of asymmetry in their kinematics, favouring prograde motion. Moreover, we find that $31\%$ of the stars that currently reside in the disk do not venture outside of the disk plane throughout their orbit. The discovery of this population implies that a significant fraction of stars with iron abundances $[Fe/H]\leq-2.5$ formed within or concurrently with the Milky Way disk and that the history of the disk was quiet enough to allow them to retain their disk-like orbital properties.
△ Less
Submitted 19 November, 2019;
originally announced November 2019.
-
Assessing the performance of LTE and NLTE synthetic stellar spectra in a machine learning framework
Authors:
Spencer Bialek,
Sébastien Fabbro,
Kim A. Venn,
Nripesh Kumar,
Teaghan O'Briain,
Kwang Moo Yi
Abstract:
In the current era of stellar spectroscopic surveys, synthetic spectral libraries are the basis for the derivation of stellar parameters and chemical abundances. In this paper, we compare the stellar parameters determined using five popular synthetic spectral grids (INTRIGOSS, FERRE, AMBRE, PHOENIX, and MPIA/1DNLTE) with our convolutional neural network (CNN, $\texttt{StarNet}$). The stellar param…
▽ More
In the current era of stellar spectroscopic surveys, synthetic spectral libraries are the basis for the derivation of stellar parameters and chemical abundances. In this paper, we compare the stellar parameters determined using five popular synthetic spectral grids (INTRIGOSS, FERRE, AMBRE, PHOENIX, and MPIA/1DNLTE) with our convolutional neural network (CNN, $\texttt{StarNet}$). The stellar parameters are determined for six physical properties (effective temperature, surface gravity, metallicity, [$α$/Fe], radial velocity, and rotational velocity) given the spectral resolution, signal-to-noise, and wavelength range of optical FLAMES-UVES spectra from the Gaia-ESO Survey. Both CNN modelling and epistemic uncertainties are incorporated through training an ensemble of networks. $\texttt{StarNet}$ training was also adapted to mitigate differences between the synthetic grids and observed spectra by augmenting with realistic observational signatures (i.e. resolution matching, wavelength sampling, Gaussian noise, zeroing flux values, rotational and radial velocities, continuum removal, and masking telluric regions). Using the FLAMES-UVES spectra for FGK type dwarfs and giants as a test set, we quantify the accuracy and precision of the stellar label predictions from $\texttt{StarNet}$. We find excellent results over a wide range of parameters when $\texttt{StarNet}$ is trained on the MPIA/1DNLTE synthetic grid, and acceptable results over smaller parameter ranges when trained on the 1DLTE grids. These tests also show that our CNN pipeline is highly adaptable to multiple simulation grids.
△ Less
Submitted 1 September, 2020; v1 submitted 6 November, 2019;
originally announced November 2019.
-
An Application of Deep Neural Networks in the Analysis of Stellar Spectra
Authors:
Sebastien Fabbro,
Kim Venn,
Teaghan O'Briain,
Spencer Bialek,
Collin Kielty,
Farbod Jahandar,
Stephanie Monty
Abstract:
Spectroscopic surveys require fast and efficient analysis methods to maximize their scientific impact. Here we apply a deep neural network architecture to analyze both SDSS-III APOGEE DR13 and synthetic stellar spectra. When our convolutional neural network model (StarNet) is trained on APOGEE spectra, we show that the stellar parameters (temperature, gravity, and metallicity) are determined with…
▽ More
Spectroscopic surveys require fast and efficient analysis methods to maximize their scientific impact. Here we apply a deep neural network architecture to analyze both SDSS-III APOGEE DR13 and synthetic stellar spectra. When our convolutional neural network model (StarNet) is trained on APOGEE spectra, we show that the stellar parameters (temperature, gravity, and metallicity) are determined with similar precision and accuracy as the APOGEE pipeline. StarNet can also predict stellar parameters when trained on synthetic data, with excellent precision and accuracy for both APOGEE data and synthetic data, over a wide range of signal-to-noise ratios. In addition, the statistical uncertainties in the stellar parameter determinations are comparable to the differences between the APOGEE pipeline results and those determined independently from optical spectra. We compare StarNet to other data-driven methods; for example, StarNet and the Cannon 2 show similar behaviour when trained with the same datasets, however StarNet performs poorly on small training sets like those used by the original Cannon. The influence of the spectral features on the stellar parameters is examined via partial derivatives of the StarNet model results with respect to the input spectra. While StarNet was developed using the APOGEE observed spectra and corresponding ASSET synthetic data, we suggest that this technique is applicable to other wavelength ranges and other spectral surveys.
△ Less
Submitted 24 January, 2018; v1 submitted 26 September, 2017;
originally announced September 2017.
-
Properties and Origins of Protein Secondary Structure
Authors:
Nicholas D. Socci,
William S. Bialek,
Jose' Nelson Onuchic
Abstract:
Proteins contain a large fraction of regular, repeating conformations, called secondary structure. A simple, generic definition of secondary structure is presented which consists of measuring local correlations along the protein chain. Using this definition and a simple model for proteins, the forces driving the formation of secondary structure are explored. The relative role of energy and entro…
▽ More
Proteins contain a large fraction of regular, repeating conformations, called secondary structure. A simple, generic definition of secondary structure is presented which consists of measuring local correlations along the protein chain. Using this definition and a simple model for proteins, the forces driving the formation of secondary structure are explored. The relative role of energy and entropy are examined. Recent work has indicated that compaction is sufficient to create secondary structure. We test this hypothesis, using simple non-lattice protein models.
△ Less
Submitted 2 February, 1994;
originally announced February 1994.
