(Abridged) The Euclid mission is expected to discover thousands of z>6 galaxies in three Deep Fields, which together will cover a ~40 deg2 area. However, the limited number of Euclid bands and availability of ancillary data could make the identification of z>6 galaxies challenging. In this work, we assess the degree of contamination by intermediate-redshift galaxies (z=1-5.8) expected for z>6 galaxies within the Euclid Deep Survey. This study is based on ~176,000 real galaxies at z=1-8 in a ~0.7 deg2 area selected from the UltraVISTA ultra-deep survey, and ~96,000 mock galaxies with 25.3$\leq$H<27.0, which altogether cover the range of magnitudes to be probed in the Euclid Deep Survey. We simulate Euclid and ancillary photometry from the fiducial, 28-band photometry, and fit spectral energy distributions (SEDs) to various combinations of these simulated data. Our study demonstrates that identifying z>6 with Euclid data alone will be very effective, with a z>6 recovery of 91(88)% for bright (faint) galaxies. For the UltraVISTA-like bright sample, the percentage of z=1-5.8 contaminants amongst apparent z>6 galaxies as observed with Euclid alone is 18%, which is reduced to 4(13)% by including ultra-deep Rubin (Spitzer) photometry. Conversely, for the faint mock sample, the contamination fraction with Euclid alone is considerably higher at 39%, and minimized to 7% when including ultra-deep Rubin data. For UltraVISTA-like bright galaxies, we find that Euclid (I-Y)>2.8 and (Y-J)<1.4 colour criteria can separate contaminants from true z>6 galaxies, although these are applicable to only 54% of the contaminants, as many have unconstrained (I-Y) colours. In the most optimistic scenario, these cuts reduce the contamination fraction to 1% whilst preserving 81% of the fiducial z>6 sample. For the faint mock sample, colour cuts are infeasible.
The various Euclid imaging surveys will become a reference for studies of galaxy morphology by delivering imaging over an unprecedented area of 15 000 square degrees with high spatial resolution. In order to understand the capabilities of measuring morphologies from Euclid-detected galaxies and to help implement measurements in the pipeline, we have conducted the Euclid Morphology Challenge, which we present in two papers. While the companion paper by Merlin et al. focuses on the analysis of photometry, this paper assesses the accuracy of the parametric galaxy morphology measurements in imaging predicted from within the Euclid Wide Survey. We evaluate the performance of five state-of-the-art surface-brightness-fitting codes DeepLeGATo, Galapagos-2, Morfometryka, Profit and SourceXtractor++ on a sample of about 1.5 million simulated galaxies resembling reduced observations with the Euclid VIS and NIR instruments. The simulations include analytic S\'ersic profiles with one and two components, as well as more realistic galaxies generated with neural networks. We find that, despite some code-specific differences, all methods tend to achieve reliable structural measurements (10% scatter on ideal S\'ersic simulations) down to an apparent magnitude of about 23 in one component and 21 in two components, which correspond to a signal-to-noise ratio of approximately 1 and 5 respectively. We also show that when tested on non-analytic profiles, the results are typically degraded by a factor of 3, driven by systematics. We conclude that the Euclid official Data Releases will deliver robust structural parameters for at least 400 million galaxies in the Euclid Wide Survey by the end of the mission. We find that a key factor for explaining the different behaviour of the codes at the faint end is the set of adopted priors for the various structural parameters.
Sub-millimeter observations reveal the star-formation activity obscured by dust in the young Universe. It still remains unclear how galaxies detected at sub-millimeter wavelengths are related to ultraviolet/optical-selected galaxies in terms of their observed quantities, physical properties, and evolutionary stages. Deep near- and mid-infrared observational data are crucial to characterize the stellar properties of galaxies detected with sub-millimeter emission. In this study, we make use of a galaxy catalog from the $Spitzer$ Matching Survey of the UltraVISTA ultra-deep Stripes. By cross-matching with a sub-millimeter source catalog constructed with the archival data of the Atacama Large Millimeter/submillimeter Array (ALMA), we search for galaxies at $z>$ 2 with a sub-millimeter detection in our galaxy catalog. We find that the ALMA-detected galaxies at $z>$ 2 are systematically massive and have redder $K_s$-[4.5] colors than the non-detected galaxies. The redder colors are consistent with the larger dust reddening values of the ALMA-detected galaxies obtained from SED fitting. We also find that the ALMA-detected galaxies tend to have brighter 4.5 $\mu$m magnitudes. This may suggest that they tend to have smaller mass-to-light ratios, and thus, to be younger than star-forming galaxies fainter at sub-millimeter wavelengths with similar stellar masses. We identify starburst galaxies with high specific star-formation rates among both ALMA-detected and non-detected SMUVS sources. Irrespective of their brightness at sub-millimeter wavelengths, these populations have similar dust reddening values, which may suggest a variety of dust SED shapes among the starburst galaxies at $z>2$.
We analyze a sample of 3300 galaxies between redshifts z~3.5 and z~8.5 selected from JWST images in the Hubble Ultra Deep Field (HUDF) and UKIDSS Ultra Deep Survey field, including objects with stellar masses as low as ~ 10^8 Msun up to z~8. The depth and wavelength coverage of the JWST data allow us, for the first time, to derive robust stellar masses for such high-z, low stellar-mass galaxies on an individual basis. We compute the galaxy stellar mass function (GSMF), after complementing our sample with ancillary data from CANDELS to constrain the GMSF at high stellar masses (M > M*). Our results show a steepening of the low stellar-mass end slope (a) with redshift, with a = -1.61 (+/-0.05) at z~4 and a = -1.98 (+/-0.14) at z~7. We also observe an evolution of the normalization phi* from z~7 to z~4, with phi*(z~4)/phi*(z~7)= 130 (+210/-50). Our study incorporates a novel method for the estimation of the Eddington bias that takes into account its possible dependence both on stellar mass and redshift, while allowing for skewness in the error distribution. We finally compute the resulting cosmic stellar mass density and find a flatter evolution with redshift than previous studies.
(Abridged) The Euclid mission is expected to discover thousands of z>6 galaxies in three Deep Fields, which together will cover a ~40 deg2 area. However, the limited number of Euclid bands and availability of ancillary data could make the identification of z>6 galaxies challenging. In this work, we assess the degree of contamination by intermediate-redshift galaxies (z=1-5.8) expected for z>6 galaxies within the Euclid Deep Survey. This study is based on ~176,000 real galaxies at z=1-8 in a ~0.7 deg2 area selected from the UltraVISTA ultra-deep survey, and ~96,000 mock galaxies with 25.3$\leq$H<27.0, which altogether cover the range of magnitudes to be probed in the Euclid Deep Survey. We simulate Euclid and ancillary photometry from the fiducial, 28-band photometry, and fit spectral energy distributions (SEDs) to various combinations of these simulated data. Our study demonstrates that identifying z>6 with Euclid data alone will be very effective, with a z>6 recovery of 91(88)% for bright (faint) galaxies. For the UltraVISTA-like bright sample, the percentage of z=1-5.8 contaminants amongst apparent z>6 galaxies as observed with Euclid alone is 18%, which is reduced to 4(13)% by including ultra-deep Rubin (Spitzer) photometry. Conversely, for the faint mock sample, the contamination fraction with Euclid alone is considerably higher at 39%, and minimized to 7% when including ultra-deep Rubin data. For UltraVISTA-like bright galaxies, we find that Euclid (I-Y)>2.8 and (Y-J)<1.4 colour criteria can separate contaminants from true z>6 galaxies, although these are applicable to only 54% of the contaminants, as many have unconstrained (I-Y) colours. In the most optimistic scenario, these cuts reduce the contamination fraction to 1% whilst preserving 81% of the fiducial z>6 sample. For the faint mock sample, colour cuts are infeasible.
The various Euclid imaging surveys will become a reference for studies of galaxy morphology by delivering imaging over an unprecedented area of 15 000 square degrees with high spatial resolution. In order to understand the capabilities of measuring morphologies from Euclid-detected galaxies and to help implement measurements in the pipeline, we have conducted the Euclid Morphology Challenge, which we present in two papers. While the companion paper by Merlin et al. focuses on the analysis of photometry, this paper assesses the accuracy of the parametric galaxy morphology measurements in imaging predicted from within the Euclid Wide Survey. We evaluate the performance of five state-of-the-art surface-brightness-fitting codes DeepLeGATo, Galapagos-2, Morfometryka, Profit and SourceXtractor++ on a sample of about 1.5 million simulated galaxies resembling reduced observations with the Euclid VIS and NIR instruments. The simulations include analytic S\'ersic profiles with one and two components, as well as more realistic galaxies generated with neural networks. We find that, despite some code-specific differences, all methods tend to achieve reliable structural measurements (10% scatter on ideal S\'ersic simulations) down to an apparent magnitude of about 23 in one component and 21 in two components, which correspond to a signal-to-noise ratio of approximately 1 and 5 respectively. We also show that when tested on non-analytic profiles, the results are typically degraded by a factor of 3, driven by systematics. We conclude that the Euclid official Data Releases will deliver robust structural parameters for at least 400 million galaxies in the Euclid Wide Survey by the end of the mission. We find that a key factor for explaining the different behaviour of the codes at the faint end is the set of adopted priors for the various structural parameters.
What do galaxies that formed right after the Big Bang look like? Are they as massive as the Milky Way is today? Do they already contain old stellar populations and galactic dust? In this PhD thesis, I used wide-area infrared galaxy surveys to investigate the presence and physical properties of galaxies billions of light years away. In Chapter 2, I simulated data for the newly-operating Euclid space telescope to assess its ability to identify early Universe galaxies, and how sample contamination from less distant galaxies with similar observational properties can be mitigated. In Chapter 3, I studied one particularly massive galaxy in the early Universe with various model fitting codes to quantify the effects of modelling assumptions on the stellar mass estimates. In Chapter 4, I studied a sample of infrared sources that are undetected at optical wavelengths with deep, high-resolution imaging from the James Webb Space Telescope to uncover their physical properties and study their Halpha emission. In Chapter 5, I placed the stellar properties of galaxies detected at sub-millimeter wavelengths in the context of comparable infrared galaxies that are undetected at such long wavelengths. In the final chapter, I summarize my findings on early Universe galaxies and discuss how synergies between existing and future infrared facilities will greatly increase statistics on these interesting sources.
We present an analysis of Euclid observations of a 0.5 deg$^2$ field in the central region of the Fornax galaxy cluster that were acquired during the performance verification phase. With these data, we investigate the potential of Euclid for identifying GCs at 20 Mpc, and validate the search methods using artificial GCs and known GCs within the field from the literature. Our analysis of artificial GCs injected into the data shows that Euclid's data in $I_{\rm E}$ band is 80% complete at about $I_{\rm E} \sim 26.0$ mag ($M_{V\rm } \sim -5.0$ mag), and resolves GCs as small as $r_{\rm h} = 2.5$ pc. In the $I_{\rm E}$ band, we detect more than 95% of the known GCs from previous spectroscopic surveys and GC candidates of the ACS Fornax Cluster Survey, of which more than 80% are resolved. We identify more than 5000 new GC candidates within the field of view down to $I_{\rm E}$ mag, about 1.5 mag fainter than the typical GC luminosity function turn-over magnitude, and investigate their spatial distribution within the intracluster field. We then focus on the GC candidates around dwarf galaxies and investigate their numbers, stacked luminosity distribution and stacked radial distribution. While the overall GC properties are consistent with those in the literature, an interesting over-representation of relatively bright candidates is found within a small number of relatively GC-rich dwarf galaxies. Our work confirms the capabilities of Euclid data in detecting GCs and separating them from foreground and background contaminants at a distance of 20 Mpc, particularly for low-GC count systems such as dwarf galaxies.
ABSTRACT Next-generation telescopes, like Euclid, Rubin/LSST, and Roman, will open new windows on the Universe, allowing us to infer physical properties for tens of millions of galaxies. Machine-learning methods are increasingly becoming the most efficient tools to handle this enormous amount of data, because they are often faster and more accurate than traditional methods. We investigate how well redshifts, stellar masses, and star-formation rates (SFRs) can be measured with deep-learning algorithms for observed galaxies within data mimicking the Euclid and Rubin/LSST surveys. We find that deep-learning neural networks and convolutional neural networks (CNNs), which are dependent on the parameter space of the training sample, perform well in measuring the properties of these galaxies and have a better accuracy than methods based on spectral energy distribution fitting. CNNs allow the processing of multiband magnitudes together with $H_{\scriptscriptstyle \rm E}$-band images. We find that the estimates of stellar masses improve with the use of an image, but those of redshift and SFR do not. Our best results are deriving (i) the redshift within a normalized error of <0.15 for 99.9 ${{\ \rm per\ cent}}$ of the galaxies with signal-to-noise ratio >3 in the $H_{\scriptscriptstyle \rm E}$ band; (ii) the stellar mass within a factor of two ($\sim\!0.3 \rm \ dex$) for 99.5 ${{\ \rm per\ cent}}$ of the considered galaxies; and (iii) the SFR within a factor of two ($\sim\!0.3 \rm \ dex$) for $\sim\!70{{\ \rm per\ cent}}$ of the sample. We discuss the implications of our work for application to surveys as well as how measurements of these galaxy parameters can be improved with deep learning.
The technique of photometric redshifts has become essential for the exploitation of multi-band extragalactic surveys. While the requirements on photo-zs for the study of galaxy evolution mostly pertain to the precision and to the fraction of outliers, the most stringent requirement in their use in cosmology is on the accuracy, with a level of bias at the sub-percent level for the Euclid cosmology mission. A separate, and challenging, calibration process is needed to control the bias at this level of accuracy. The bias in photo-zs has several distinct origins that may not always be easily overcome. We identify here one source of bias linked to the spatial or time variability of the passbands used to determine the photometric colours of galaxies. We first quantified the effect as observed on several well-known photometric cameras, and found in particular that, due to the properties of optical filters, the redshifts of off-axis sources are usually overestimated. We show using simple simulations that the detailed and complex changes in the shape can be mostly ignored and that it is sufficient to know the mean wavelength of the passbands of each photometric observation to correct almost exactly for this bias; the key point is that this mean wavelength is independent of the spectral energy distribution of the source}. We use this property to propose a correction that can be computationally efficiently implemented in some photo-z algorithms, in particular template-fitting. We verified that our algorithm, implemented in the new photo-z code Phosphoros, can effectively reduce the bias in photo-zs on real data using the CFHTLS T007 survey, with an average measured bias Delta z over the redshift range 0.4
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
