BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

S. Arnouts

S. Arnouts contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
astro-ph.GAastro-ph.COastro-ph.IMMachine Learning

Trust snapshot

Quick read

Trust 19 - UnverifiedVerification L1Unclaimed author
5works
0followers
4topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

5 published item(s)

preprint2022arXiv

A Cautionary Tale of LyC Escape Fraction Estimates from High Redshift Galaxies

Measuring the escape fraction, $f_{\rm esc}$, of ionizing, Lyman Continuum (LyC) radiation is key to our understanding of the process of cosmic reionization. In this paper we provide a methodology for recovering the posterior probability distribution of the LyC escape fraction, $f_{\rm esc}^{\rm PDF}$, considering both the observational uncertainties and ensembles of simulated transmission functions through the intergalactic medium (IGM). We present an example of this method applied to a VUDS galaxy at $z=3.64$ and find $f_{\rm esc}^{\rm PDF}$ = 0.51$^{+0.33}_{-0.34}$ and compare this to the values computed assuming averaged IGM transmission with and without consideration of detection bias along average sightlines yielding $f_{\rm esc}^{\langle T \rangle}$ = 1.40$^{+0.80}_{-0.42}$, and $f_{\rm esc}^{\rm bias}$ = 0.82$^{+0.33}_{-0.16}$. Our results highlight the limitations of methods assuming average, smooth transmission functions. We also present MOSFIRE data for a sample of seven LyC candidates selected based on photometric redshifts at $z > 3.4$, but find that all seven have overestimated photometric redshifts by $Δz \sim 0.2$ making them unsuitable for LyC measurements. This results likely due to a bias induced by our selection criteria.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Photometric Redshift Estimation with Convolutional Neural Networks and Galaxy Images: A Case Study of Resolving Biases in Data-Driven Methods

Deep Learning models have been increasingly exploited in astrophysical studies, yet such data-driven algorithms are prone to producing biased outputs detrimental for subsequent analyses. In this work, we investigate two major forms of biases, i.e., class-dependent residuals and mode collapse, in a case study of estimating photometric redshifts as a classification problem using Convolutional Neural Networks (CNNs) and galaxy images with spectroscopic redshifts. We focus on point estimates and propose a set of consecutive steps for resolving the two biases based on CNN models, involving representation learning with multi-channel outputs, balancing the training data and leveraging soft labels. The residuals can be viewed as a function of spectroscopic redshifts or photometric redshifts, and the biases with respect to these two definitions are incompatible and should be treated in a split way. We suggest that resolving biases in the spectroscopic space is a prerequisite for resolving biases in the photometric space. Experiments show that our methods possess a better capability in controlling biases compared to benchmark methods, and exhibit robustness under varying implementing and training conditions provided with high-quality data. Our methods have promises for future cosmological surveys that require a good constraint of biases, and may be applied to regression problems and other studies that make use of data-driven models. Nonetheless, the bias-variance trade-off and the demand on sufficient statistics suggest the need for developing better methodologies and optimizing data usage strategies.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

COSMOS2020: A panchromatic view of the Universe to $z\sim10$ from two complementary catalogs

The Cosmic Evolution Survey (COSMOS) has become a cornerstone of extragalactic astronomy. Since the last public catalog in 2015, a wealth of new imaging and spectroscopic data has been collected in the COSMOS field. This paper describes the collection, processing, and analysis of this new imaging data to produce a new reference photometric redshift catalog. Source detection and multi-wavelength photometry is performed for 1.7 million sources across the $2\,\mathrm{deg}^{2}$ of the COSMOS field, $\sim$966,000 of which are measured with all available broad-band data using both traditional aperture photometric methods and a new profile-fitting photometric extraction tool, The Farmer, which we have developed. A detailed comparison of the two resulting photometric catalogs is presented. Photometric redshifts are computed for all sources in each catalog utilizing two independent photometric redshift codes. Finally, a comparison is made between the performance of the photometric methodologies and of the redshift codes to demonstrate an exceptional degree of self-consistency in the resulting photometric redshifts. The $i<21$ sources have sub-percent photometric redshift accuracy and even the faintest sources at $25<i<27$ reach a precision of $5\,\%$. Finally, these results are discussed in the context of previous, current, and future surveys in the COSMOS field. Compared to COSMOS2015, reaches the same photometric redshift precision at almost one magnitude deeper. Both photometric catalogs and their photometric redshift solutions and physical parameters will be made available through the usual astronomical archive systems (ESO Phase 3, IPAC IRSA, and CDS).

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Outside the Lyman-break box: detecting Lyman continuum emitters at $3.5<z<5.1$ with CLAUDS

Identifying non-contaminated sample of high-redshift galaxies with escaping Lyman continuum (LyC) flux is important for understanding the sources and evolution of cosmic reionization. We present CLAUDS $u$-band photometry of the COSMOS field to probe LyC radiation from spectroscopically confirmed galaxies at $z\geq3.5$ and outside the standard Lyman-break galaxy colour selection expectations. Complementary to the CLAUDS data, we use Subaru multi-filter photometry, \textit {Hubble Space Telescope} ($HST$) multi-filter imaging, and the spectroscopic surveys D10K, VUDS and 3D-HST. We present a sample of Lyman continuum galaxy (LCG) candidates in the redshift range $3.5\lesssim z\lesssim5.1$. Here, we introduce 5 LCG candidates, where two are flagged quality 1 and three quality 2. The estimated $f_{\rm esc}^{\rm abs}$ for quality 1 candidates are in the range $\sim5\% - 73\%$ and $\sim30\% - 93\%$. These estimates are based on our derived parameters from individual galaxies as inputs to a range of BPASS models as well as mean intergalactic medium (IGM) and maximal intergalactic and circumgalactic media (IGM+CGM) transmission. We conclude that our search for LCGs is most likely biased to lines of sight with low HI densities or free from Lyman limit systems. Our two best LCG candidates have EW (Ly$α)\leq50$Å and we find no correlation or anti-correlation between EW (Ly$α$), $f_{\rm esc}^{\rm abs}$, and $R_{\rm obs}$, the ratio of ionizing to non-ionizing observed flux in the measured passbands. Stacking candidates without solid LyC detections ($S/N<3$) results in an estimated $f_{\rm esc}^{\rm abs}$ from galaxies not greater than $1\%$.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

PhotoWeb redshift: boosting photometric redshift accuracy with large spectroscopic surveys

Improving distance measurements in large imaging surveys is a major challenge to better reveal the distribution of galaxies on a large scale and to link galaxy properties with their environments. Photometric redshifts can be efficiently combined with the cosmic web (CW) extracted from overlapping spectroscopic surveys to improve their accuracy. We apply a similar method using a new generation of photometric redshifts based on a convolution neural network (CNN). The CNN is trained on the SDSS images with the main galaxy sample (SDSS-MGS, $r \leq 17.8$) and the GAMA spectroscopic redshifts up tor $\sim 19.8$. The mapping of the CW is obtained with 680,000 spectroscopic redshifts from the MGS and BOSS surveys. The redshift probability distribution functions (PDF), which are well calibrated (unbiased and narrow, $\leq 120$ Mpc), intercept a few CW structure along the line of sight. Combining these PDFs with the density field distribution provides new photometric redshifts, $z_{web}$, whose accuracy is improved by a factor of two (i.e.,$σ \sim 0.004(1+z)$) for galaxies with $r \leq 17.8$. For half of them, the distance accuracy is better than 10 cMpc. The narrower the original PDF, the larger the boost in accuracy. No gain is observed for original PDFs wider than 0.03. The final $z_{web}$ PDFs also appear well calibrated. The method performs slightly better for passive galaxies than star-forming ones, and for galaxies in massive groups since these populations better trace the underlying large-scale structure. Reducing the spectroscopic sampling by a factor of 8 still improves the photometric redshift accuracy by 25%. Extending the method to galaxies fainter than the MGS limit still improves the redshift estimates for 70% of the galaxies, with a gain in accuracy of 20% at low $z$ where the resolution of the CW is the highest.

0 citations0 reviewsNo code linkedOpen access