EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 678 (October 2023) A&A, 678 (2023) A25 Full HTML
Open Access
Issue
A&A
Volume 678, October 2023
Article Number A25
Number of page(s) 12
Section Catalogs and data
DOI https://doi.org/10.1051/0004-6361/202346293
Published online 29 September 2023

© The Authors 2023

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1 Introduction

A major theme in extragalactic astronomy is understanding when and how galaxies formed and evolved. Spectroscopic surveys play a fundamental role in this respect, not only because they provide robust redshifts, but especially because the analysis of emission and absorption lines and spectral breaks grants access to intrinsic physical properties of galaxies such as the chemical composition of their gas and stellar populations, the ionising radiation field, and the gas and star kinematics.

Over the past two decades, several multi-slit and multi-fibre surveys have been carried out, targeting increasingly distant galaxies: from the Sloan Digital Sky Survey (SDSS) in the local Universe (Abazajian et al. 2003; Abdurro'uf et al. 2022), passing through the VIMOS VLT1 Deep Survey (VVDS; Le Fèvre et al. 2013; Garilli et al. 2008), zCOSMOS (Lilly et al. 2007), VIMOS Public Extragalactic Redshift Survey (VIPERS; Guzzo et al. 2014; Scodeggio et al. 2018), and the Large Early Galaxy Census (LEGA-C; van der Wel et al. 2016) at < z >~ 0.7, the Galaxy Mass Assembly ultradeep Spectroscopic Survey (GMASS; Cimatti et al. 2008; Kurk et al. 2013) at Cosmic Noon, and up to z ~ 4–6 with KBSS-MOSFIRE2 (Steidel et al. 2014) and the VIMOS Ultra Deep Survey (VUDS; Le Fèvre et al. 2015), along with a number of smaller samples targeting the reionisation epoch (e.g. Pentericci et al. 2018b). All these surveys have improved our understanding of galaxy evolution, mainly by drawing a detailed 3D map of the Universe with thousands of redshifts.

VANDELS is an ESO public VIMOS survey of the Chandra Deep-Field South (CDFS) and Ultra-Deep Survey (UDS) fields that was designed to complement and extend the work of the CANDELS (Grogin et al. 2011; Koekemoer et al. 2011) imaging campaigns. The strategy of VANDELS was not to limit itself to finding a redshift, but to focus on ultra-long exposures of a relatively small number of galaxies that provide high signal-to-noise ratio (S/N) spectra to study in detail the physical characteristics of the high-redshift galaxies (McLure et al. 2018). Since the first data release of VANDELS (Pentericci et al. 2018a), a number of papers have been published studying several properties, ranging from dust attenuation, interstellar medium properties, and stellar metallicities of star-forming (Cullen et al. 2018, 2019; Calabrò et al. 2021, 2022a,b; Fontanot et al. 2021) and quiescent galaxies (Carnall et al. 2019, Carnall et al. 2020, 2022; Hamadouche et al. 2022, 2023; Tomasetti et al. 2023), to intergalactic medium properties (Thomas et al. 2020, 2021), the ionising photon production efficiency (Castellano et al. 2023), the LyC escape fraction (Begley et al. 2022; Saldana-Lopez et al. 2023), Lya, HeIIλ1640, CIVλ1550, and CIII]λ1908 emitters (Marchi et al. 2019; Hoag et al. 2019; Cullen et al. 2020; Saxena et al. 2020a,b, 2022; Guaita et al. 2020; Llerena et al. 2022; Mascia et al. 2023), AGN (Magliocchetti et al. 2020), and high-mass X-ray binaries (Saxena et al. 2021).

This paper represents the official release of the VANDELS spectroscopic measurements (i.e. lines, indices, and breaks), which are herewith made available to the whole astrophysical community. The catalogues include all spectra from the VAN-DELS final data release (DR4) presented in Garilli et al. (2021) with a robust spectroscopic redshift.

The paper is organised as follows: Section 2 briefly describes the VANDELS survey; Sect. 3 summarises the methods of measurements; Sect. 4 discusses the measurement code validation tests, including a description of the creation of ad hoc mock 1D spectra; Sect. 5 discusses an issue with the error spectra and its resolution; Sect. 6 describes the released catalogues and the comparison with independent measurements from previously published works inside the VANDELS collaboration; and Sect. 7 provides a brief summary. In this paper, we provide magnitudes in the AB photometric system (Oke & Gunn 1983).

2 The VANDELS survey

VANDELS is a spectroscopic survey performed with the ESO-VLT VIMOS spectrograph in two CANDELS fields, over a total area of ~0.2 square degree. For all the details on the survey design, target selection, observations, data reduction, and spectroscopic redshift measurements, we refer the reader to McLure et al. (2018); Pentericci et al. (2018a); Garilli et al. (2021).

The VANDELS spectra cover a wavelength range of 4800 Å < λobs < 9800 Å, with a dispersion of 2.5 Å pixel−1 and a spectral resolution of R ~ 650, corresponding to a FWHMres ~ 460 km s−1 (or FWHMres ~ 11.2 Å at 7300 Å). The main targets of the survey were massive passive galaxies at 1 < z < 2.5, bright star-forming galaxies (SFGs) at 2.4 < z < 5.5, and fainter SFGs at 3 < z < 7 Lyman-break galaxies, plus a small sample of AGN, pre-selected using various multi-wavelength criteria. The VAN-DELS spectroscopic targets were pre-selected using high-quality photometric redshifts and were observed for a minimum of 20h and up to 80 h, depending on their brightness, in order to ensure an approximately homogeneous S/N on the continuum within each class of galaxy. The data reduction was carried out using the recipes provided by the VIMOS Interactive Pipeline and Graphical Interface (VIPGI) package (Scodeggio et al. 2005) and the EASYLIFE environment (Garilli et al. 2012). The measured S/N per resolution element is higher than ten for all passive and star-forming galaxies, and higher than five for 85% of Lyman-break galaxies and AGN (Garilli et al. 2021). Spectroscopic redshifts were determined for all objects using the Easy redshift (EZ) software package within the PANDORA environment (Garilli et al. 2010).

A redshift confidence flag was also assigned to each target, according to the following scheme, already applied to previous VIMOS surveys (e.g. VVDS, Le Fèvre et al. 2005; zCOSMOS, Lilly et al. 2007; VUDS, Le Fèvre et al. 2015).

  • Flag 4: a highly reliable redshift (estimated to have a > 99% probability of being correct), based on a high S/N spectrum and supported by obvious and consistent spectral features.

  • Flag 3: also a very reliable redshift, comparable in confidence with Flag 4, supported by clear spectral features in the spectrum, but not necessarily with a high S/N.

  • Flag 2: a fairly reliable redshift measurement, although not as straightforward to confirm as those for Flags 3 and 4, supported by cross-correlation results, continuum shape, and some spectral features.

  • Flag 1: a reasonable redshift measurement, based on weak spectral features and/or continuum shape.

    An a posteriori analysis of the redshift reliability showed that the reliability of Flag 2 redshifts is ~79%, while that of Flag 1 redshifts is 41% (Garilli et al. 2021).

  • Flag 0: no reliable spectroscopic redshift measurement was possible.

  • Flag 9: a redshift based on only one single clear spectral emission feature. An a posteriori analysis confirmed a redshift reliability of ~95% for spectra with this flag.

  • Flag -10: spectrum with clear problems in the observation or data-processing phases.

  • Flag 10+any of the above: broad line AGN (BLAGN). This preliminary classification has been subsequently revised by Bongiorno et al. (in prep.).

  • Serendipitous (also called secondary) objects appearing by chance within the slit of the main target were identified by adding a ‘2’ in front of the main flag.

The redshift accuracy, estimated by internal comparison between different observations, is σΔz/(1+z) = 0.0007 (Garilli et al. 2021). The redshift distribution of the entire VANDELS sample is shown in Fig. 1.

In the official catalogues, we include only the measurements for the 1811 objects with a reliable redshift confidence flag (2, 3, 4, and 9 and the equivalent for BLAGN and secondary objects), whose redshift distribution is also shown in Fig. 1.

We measured spectroscopic features using two methods: Gaussian fit and direct integration.

Table 1

slinefit parameters.

thumbnail Fig. 1

Redshift distribution of the final VANDELS sample: the grey histogram includes all measurements, and the green histogram includes only reliable redshifts (confidence flag > 1 ; see the text for more details).

3 Gaussian fit measurements

Gaussian fit measurements were performed using slinefit3 (Schreiber et al. 2018), an automated code that models the observed spectrum of a galaxy as a combination of a stellar continuum model and a set of emission and absorption lines.

3.1 slinefit parameters

A set of templates from EAzY (Brammer et al. 2008), based on the Bruzual & Chariot (2003) stellar population models, is linearly combined to best fit the continuum. Table 1 summarises the parameters that were set to produce the official VANDELS catalogue. The code searches for lines around their expected locations given by the input redshift: lines with a S/N lower than offset_snr_min are fixed at their expected position, while a velocity offset with respect to the measured redshift is allowed for lines with a higher S/N, with a maximum value set by the offset_max parameter. We stress that in the catalogue the tr of each line is provided, not the FWHM.

We measured 40 individual lines, including 7 resolved doublets, which are listed in Table 2. Unresolved doublets (e.g. CIV/11550 and CIII]λ11908) were treated as a single line. For the NIIλλ6548,6583 and SIIλλ7I6,6730 doublets, we fixed the line flux ratios to 0.33:1 and 1:0.75, respectively, while no constraints were imposed for the other doublets. All lines were modelled as single symmetric Gaussians, either in emission or in absorption. This might not have been the best choice for the Lya line, which typically is asymmetric and sometimes even split into a blue and a red component. Therefore, after visual inspection of the spectra by four members of the team, we added a flag indicating whether the fit was good (1) or not (0) and recommend using with caution the Lya parameters from the catalogue in the latter cases. In general, in the case of multi-component lines (e.g. P-Cygni profiles), only the strongest feature is fitted. We stress that slinefit, in our chosen configuration, always provides a solution. Therefore, we recommend caution when using spectral parameters when the lines are narrower than the spectral resolution (i.e. FWHM ~ 460 km s−1, corresponding to tr ~ 195 km s−1 ), because it might be a noise spike. S /N < 1, the line should be considered as undetected and the error on the flux can be used as a 1er upper limit. Errors were evaluated through a Monte Carlo technique: the galaxy spectrum was randomly perturbed according to its re-scaled error spectrum (see the next Section) and trie uncertainties on the spectroscopic parameters were then computed from the standard deviation of num_mc realisations of the fit.

Table 2

Measured spectral lines (Gaussian fit).

thumbnail Fig. 2

Construction of the 1D mock spectra for the slinefit code validation. Top: synthetic rest-frame template (from Talia et al. 2012), normalised to unity at 1750 Å. Bottom: comparison between three examples of mock 1D spectra and real VANDELS spectra. Mock spectra are shown in black. VANDELS spectra are colour-coded with respect: to their depth: 20 h (red), 40 h (green), and 80 Is (blue).

3.2 Mock 1D spectra and slinefit code validation

In order to validate the slinefit code performance, we built a set of 1D mock spectra to mimic; the characteristics of the observed VANDELS ones. We started from a rest-frame template, normalised to unity at 1750 Å and created using the stacked spectrum of SFGs at z ~ 2 from Talia et al. (2012) as reference (Fig. 2, top). The continuum was modelled as a cubic spline, with a slope of β ~ − 1.1 and a dispersion of 1Å pixel−1. Emission and absorption spectral lines that are common in the UV range of SFG spectra (Table 2) were added as symmefric Gaussians. The lines were not all added at their vacuum rest-frame wavelength: some shifts were introduced in order to mimic the effects of outflows. We (created three templates with the same continuum and varying the lines’ peak S/N in the range 0.3–7. Then, each reSt-frame templare was used to create 30 redshifted templates, with redshitts evenly ttstributed in the range 2.2–5.

The redshifted templates were normalised to the F814W observed magnitude, following the magnitude versus redshift relation ot the VANDELS survey. They were then re-sampled and cut to the VANDELS dispersion and observed wavelength range.

In order to add realistic noise, we extracted 113 spectra from empty regions in observed 2D spectra from the VANDELS survey at different exposure times and added them to the redshifted templates. The final validation sampie counts 270 mock spectra. In Fig. 2 (bottom), we show the comparison between three example of mock 1D spectra and real VANDELS spectra at different redshifts and with different quality flags.

Finally, we ran the slinefit on the sample of mock 1D spectra with different sets of input parameters and checked the relative change in the measured spectral quantities with respect to their input values, and the pull distributions. In Fig. 3, we show the resulte from the run with the best set of parameters, which is summarised in Table 1. AH the measured lines are included in the plots, but we stress that separating emission and absorption lines does not change the results. Alf distributions are consistent with a Gaussian with a null mean and unity sigma.

4 Direct integration

The direct integration measurements were performed using pylick4, a flexible Python tool to measure spectral indices and associated uncertainties. The code is described in Borghi et al. (2022) and was extensively tested usino spectra and results from the LEGA-C survey (van der Wel et al. 2016; Straatman et al. 2018). Following the approach of the Lick groupe (Worthey & Ottaviani 1997), the code compufes the strengths of a set of atomic and moleoular indices and continuum breaks such as the D4000 (Bruzual A. 1983). Errors are evaluated following the S/N method by Cardiel et al. (1998).

In our catalogue, we have included 55 indices and breaks defined in previous work) to which we added three UV emission tine indices (see Table 3). "The new indices were defined on the basis of a high-S/N composite spectrum of all VAN-DELS sources with a high-redshift euality flag (i.e. 3 and 4). It was built by median stacking the de-redshifted, tcaled (by the median flux in the wavelength range 1410–1510 Å), and rebinned (0.6 Å pixel−1 ) spectra. In Fig. 4, we show the zoomed-in regions around the HeIIλλ1640+OfIII]/t1666 and CIII]λ1909 lines, with the central bandpass and pseudo-continuum ranges marked in different colours. It should be noted that tor the direct integration catalogue, no offset of the bandpass es is allowed with respect to the expected wavelength, given the redshifs.

For she Lyα line, we opted for a different approach. Following Cullen et al. (2020), we applied the method by Kornei et al. (2010) to meagre the EW of the line, which takes into account the line’s morphology to optimise the wavelength range over which the flux is integrated. The Lyα line of the 1218 individual galaxies at z ≳ 2.95 (i.e. the redshift limit for the Lyα to be in the VIMOS wavelength range) was visually classified as either emission, absorption, combination, or noise; examples are shown in Frg. 5. The emission spectra are clearly dominated by a Lyα emission feature. The absorption spectra are dominated by an extended through around the Lya position. In the combination case, the spectrum contain superimposed emission and absorption features. Finally, the noise category include spectra where no clear feature could be identified at the Lyα position (see Kornei et al. 2010, for a detailed description). In the firet three cases, after the peak of the emission and absorption, the integration window is defined by the wavelength values on either side of the peak where the flux intersects the average continuum level. The blue and red continua are defined as the median flux values in the range λ = [1120–1180] Å and [1228–1255] Å, respectively. In the case of absorption and combination sources, the spectra were first smoothed with a boxcar function of six pixels in width to minimise the possibility of noise spikes affecting the determination of the boundaries of the integration range. For noise sources, the Lyα flux iv simply defined as the integrated flux in the range λ = [1200–12228] Å. In all cases the line flux was divided by the red continuum value to obtain she EW.

Table 3

pylick spectral indices.

thumbnail Fig. 3

Comparison between slinefit results and input values for the sample of 270 mock spectra. In the top panels, we plotted the pull distributions. As a reference, we marked with a grey curve a Gaussian with a null mean and unity sigma. In the bottom panels, we plotted the relative change of the measured spectral quantities, with respect to their input values, as a function of the peak S/N of the lines. Black squares represent the median values of the relative change in bins of the S/N; error bars are the semi-interquartile range (SIQR). The line parameters are, starting clockwise from the top left figure: line centroid, EW, FWHM, and flux.
thumbnail Fig. 4

Median composite spectrum of VANDELS sources (grey). The upper and lower panels show zoomed-in regions around the HeIIλ1640+OIII]λ1666 and CIII]λ1909 lines, respectively. The central bandpasses, as indicated in Table 3, are marked in black, while the two local continuum windows are marked in blue and red. The green points and dashed lines indicate the mean flux in the continuum bandpasses and the linear pseudo-continuum.
thumbnail Fig. 5

Examples illustrating the four Lyα categories from Kornei et al. (2010). Clockwise, from the top left: emission, combination, absorption, and noise.

5 Scaling of the error spectra

The spectra distributed as part of the VANDELS public data release include the ID noise estimate5 in erg cm−2 s−1 Å−1. The error spectrum is a direct product of the data reduction procedures (Garilli et al. 2021), and should reflect the noise level of the corresponding object spectrum. However, the comparison between the error spectra and the noise r.m.s. of the object spectra, measured in line-free regions, shows a discrepancy, with the error spectra underestimating the noise level by a factor of ~2, on average. We performed several tests on 2D and 1D spectra: our hypothesis is that the discrepancy is caused by the fact that the data reduction pipeline does not take the full covariance matrix into account. We opted for an a posteriori statistical correction of the error spectra (e.g. van der Wei et al. 2021). In particular, for each object, we computed a scaling factor to be applied to the error spectrum. The scaling factor is defined as the standard deviation of the fit residuals, divided by the error spectrum:

1.482×MAD [ (objectspectrum−modelspectrum)/errorspectrum ],$1.482 \times {\rm{MAD}}\,\left[ {{{\left( {{\rm{objec}}{{\rm{t}}_{{\rm{spectrum}}}} - {\rm{mode}}{{\rm{l}}_{{\rm{spectrum}}}}} \right)} \mathord{\left/ {\vphantom {{\left( {{\rm{objec}}{{\rm{t}}_{{\rm{spectrum}}}} - {\rm{mode}}{{\rm{l}}_{{\rm{spectrum}}}}} \right)} {{\rm{erro}}{{\rm{r}}_{{\rm{spectrum}}}}}}} \right. \kern-\nulldelimiterspace} {{\rm{erro}}{{\rm{r}}_{{\rm{spectrum}}}}}}} \right],$

where modelspectrum is the output of slinefit and MAD is the median absolute deviation. If the error spectrum is an accurate representation of the noise in the object spectrum, the above quantity should be close to 1; if the error spectrum underestimates the noise, then the above quantity can be used as a scaling factor. We computed it in five wavelength windows, free of strong sky lines, and then defined the scaling factor as the mean of the five values. The associated uncertainty is the error on the mean, which takes into account a slight wavelength dependence of the ratio between the noise r.m.s. of the object spectrum and the error spectrum (i.e. the ratio is on average ~10% lower close to the spectral edges than in the central region). Figure 6 shows the distribution of the scaling factor.

The slinefit code can actually perform the scaling of the error spectrum internally. If the appropriate keyword (residual_rescale) is switched on, the previously defined scaling factor is computed locally for each line; then, the whole error spectrum is normalised by interpolating between the scaling factors of the chosen lines, and the whole fit is performed a second time. The measurements in the official catalogue were instead performed by applying a single scaling factor to each error spectrum before running slinefit with the residual_rescale keyword switched off. This choice allowed us to provide a set of measurements that could be easily reproduced by other codes that do not include a scaling feature. The direct integration measurements were also performed after scaling the error spectra.

The scaling factors and their uncertainties are included in both catalogues. We stress that the error spectra in the VANDELS data release (i.e. NOISE extension or ERR column) are not scaled: they have to be multiplied by the scaling factor in order to obtain reliable errors on the spectroscopic measurements.

thumbnail Fig. 6

Distribution of the multiplicative scaling factors to correct the mismatch between the error spectra and the noise of the object spectra.
thumbnail Fig. 7

Distributions of the D4000 break and the EW of some notable lines. In each panel we show the distributions at S/N ≥ 1 (light grey) and S/N ≥ 3 (green). For the lines from the 'Gaussian fit' catalogues, the cut is in S/N flux. Top left: Lya EW (direct integration). Top right: D4000 (direct integration). Middle left: CIII]λ909 Å EW (Gaussian fit). Middle right: [OII]λ3727 Å EW (Gaussian fit). Bottom left: OI+SiIIλ303 Å EW (direct integration). Bottom right: OI+SiIIλ1303 A EW (Gaussian fit). In the last two panels, we also show the distribution at S/N > 2 (dark grey).

6 The catalogues

We have produced a total of four catalogues: two (one for each field) for the Gaussian fit measurements performed with slinefit and two (again, one for each field) for the direct integration measurements performed with pylick plus Lya following the Kornei et al. (2010) method. The contents of the catalogues are summarised in Table 4, while in Fig. 7, we show the distributions of the EW of some notable lines and, for the passive galaxies' sample at z < 2, the D4000 break. As already mentioned, we computed spectral properties only for galaxies with a reliable redshift, namely those with a quality flag = 2, 3,4, 9. In the slinefit catalogue, we have not included the measurements for the three BLAGN whose emission line fits require two components6. Dedicated spectral measurements for these objects will be presented in Bongiorno et al. (in prep.).

The EW of some lines was measured using both the Gaussian fit and direct integration methods. The agreement between the two measurements is very good in the case of single lines, as shown in Fig. 8 for trie SiIIλ1526Å, as an example: on average, the linear correlation coefficient ii rxy > 0.9 and the root-mean-square error (RMSE) is ~0.5–0.6 Å. A systematic small offsett of <0.5 Å is attributable to the different ways of determining the continuum level in the two methods. In lhe case of unresolved groups of lines, where a single-Gaussian model was assumed (e.g. OI+SiIIλ1303 Å), the correlation coeffictent between the two methods ts still high (rxy ≳ 0.8, on average, with an RMSE of ~0.7–0.8Å), but the Gaussian fit; tends to systematically underestimate the flux, more than `wli^t would be expected by accounting only for the differences in the continuum. The EW ratio between the two methods is on average between 0.6 and 0.8, depending on the group of lines.

Finally, as an additional validation, in Fig. 9 we compare subsets of measurements from our catalogues to independent and previously published measurements performed with different codes and methods by VANDELS team members. For the Gaussian fit catalogue, we compared our measurements of the CIII]λ1909 Å flux and of the centroids of four absorption lines to those by Calabrò et al. (2022b), which were obtained by fitting each line profile with a Gaussian function using the Python version of the MPFIT routine (Markwardt 2009). The continuum was parameterised as a straight line and fitted simultaneously with the lines. We find a good agreement between the two sets of measurements (no S/N cut applied): for the CIII]λ1909 Å fluxes, the linear correlation coefficient is rxy ~ 0.9 and the RMSE ~ 0.8 × 10−18 erg s−1 cm2, with no significant offset with respect to the 1-to-1 relation. On the other hand, for the absorption lines’ centroids rxy ~ 1.0 and RMSE ~ 5.0 × 10−4 Å.

We checked the Lya flux measurements from Guaita et al. (2020), which were obtained with a custom code based on the optimize.leastsq Python function (see also Guaita et al. 2017), by fitting the lines with a Gaussian profile and assuming a linear continuum. Guaita et al. (2020) provide two sets of measurements: one assuming a symmetric Gaussian profile, the other using a skewed Gaussian function. For our exercise, we took the former set (i.e. symmetric Gaussian) and limited the comparison to the galaxies with a goodness-of-fit flag for Lya equal to 1 (see Sect. 3). We also find in this case a good agreement between the two measurements, with an rxy ~ 0.9 and the RMSE ~ 0.3 × 10−17erg s−1 cm2.

We compared our flux measurements for different emission lines in the VANDELS AGN sample (excluding BLAGN) to the ones obtained with a custom Python code from Bongiorno et al. (in prep.): the line fluxes are the mean of a Gaussian and a Lorentzian profile fit, plus a polynomial continuum. The rxy and RMSE range from 0.7 to 0.9 and from 2.0 × 10−17 to 4.0 × 10−17erg s−1 cm2, respectively, depending on the line. Finally, we checked the Dn4000 break in the VANDELS subsample of quiescent galaxies from our direct integration catalogue against the independent measurements by Hamadouche et al. (2022) and we found an excellent agreement (rxy ~ 1.0 and RMSE ~ 3.0 × 10−2).

Table 4

Legend of catalogue content.

thumbnail Fig. 8

Examples of EW comparison between Gaussian fit and direct integration methods for a single line (SiIIλ/1526 Å; red triangles) and an unresolved group of lines (OI+SiIIλ/1303 Å; green squares). Only measurements at S/N > 3 are shown. The 1-to-1 relation is also indicated in black.
thumbnail Fig. 9

Comparison between trie measurements presented in this work and previously published VANDELS results. In all plots we also show the one-to-one relation (dashed red line). Top left: CIII]λ1909Å flux from Calabrò et al. (2022b; Gaussian fit). Top right: interstellar medium absorption line centroids from Calabrò et al. (2022b; Gaussian fit; the points for the different ions have been shifted by 0.01 for visualisation purposes). Middle left: Lyα flux from Guaita et al. (2020; Gaussian fit; no error was available for these measurements). Middle right: AGN emission line flux trom Bongiorno et al. (in prep.; Gaussian tit). Bottom: Dn4000 from Hamadouche et al. (2022; direct integration).

7 Summary

In this paper, we present the public release of the spectroscopic measurements of the VANDELS survey (Pentericci et al. 2018a; McLure et al. 2018; Garilli et al. 2021). We built two catalogues: one containing line properties from Gaussian fit measurements performed with the slinefit code, the other including line indices and continuum breaks measured with the pylick code, plus Lya EWs following the Kornei et al. (2010) method. We created a set of mock spectra to mimic observed VANDELS sources in order to validate the slinefit code, while the pylick code was already tested in a previous work (Borghi et al. 2022). As a further check of the accuracy of our catalogues, we compared subsets of measurements to previous results obtained with different codes and methods. We have also found that the error spectra included in the VANDELS data release underestimate the noise level when compared to the r.m.s. of the object spectra and computed a correction that we provide in the catalogues. The full spectroscopic catalogues, together with the spectra, redshift catalogues, complementary photometric information, and SED fitting derived quantities, are publicly available from the VANDELS survey database7 and at the CDS.

Aknowledgements

This paper is dedicated to the memory of Olivier Le Fèvre. We would like to thank the anonymous referee for their constructive comments. The VANDELS Data Release 4 (DR4), including the catalogues presented in this papers, is publicly available and can be accessed using the VANDELS database at http://vandels.inaf.it/dr4.html, or through the ESO archives. The data published in this paper have been obtained using the pandora.ez software developed by INAF IASF-Milano. MT, LPoz and ACim acknowledge the support from grant PRIN MIUR 2017 20173ML3WW_001. MT acknowledges the use of computational resources from the parallel computing cluster of the Open Physics Hub (https://site.unibo.it/openphysicshub/en) at the Physics and Astronomy Department in Bologna. ACim and MMor acknowledge the grants ASI n.I/023/12/0 and ASI no. 2018-23-HH.0. Mmor acknowledges support from MIUR, PRIN 2017 (grant 20179ZF5KS) LPoz acknowledges the support from "fondi premiali" MITiC (MIning The Cosmos Big Data and Innovative Italian Technology for Frontier Astrophysics and Cosmology). LPen, ACal and MC acknowledge support from the Mainstream Grant VANDELS. The Cosmic Dawn Center (DAWN) is funded by the Danish National Research Foundation under grant no.140. JPUF acknowledges support from the Carlsberg Foundation. RA acknowledges support from ANID Fondecyt Regular 1202007. ACC thanks the Leverhulme Trust for their support via a Leverhulme Early Career Fellowship. MLH acknowledges the support of the UK Science and Technology Facilities Council. ASL acknowledges support from Swiss National Science Foundation.

References

  1. Abazajian, K., Adelman-McCarthy, J. K., Agüeros, M. A., et al. 2003, AJ, 126, 2081 [Google Scholar]
  2. Abdurro'uf, Accetta, K., Aerts, C., et al. 2022, ApJS, 259, 35 [NASA ADS] [CrossRef] [Google Scholar]
  3. Balogh, M. L., Morris, S. L., Yee, H. K. C., Carlberg, R. G., & Ellingson, E. 1999, ApJ, 527, 54 [Google Scholar]
  4. Begley, R., Cullen, F., McLure, R. J., et al. 2022, MNRAS, 513, 3510 [NASA ADS] [CrossRef] [Google Scholar]
  5. Borghi, N., Moresco, M., Cimatti, A., et al. 2022, ApJ, 927, 164 [NASA ADS] [CrossRef] [Google Scholar]
  6. Brammer, G. B., van Dokkum, P. G., & Coppi, P. 2008, ApJ, 686, 1503 [Google Scholar]
  7. Bruzual, A. G. 1983, ApJ, 273, 105 [CrossRef] [Google Scholar]
  8. Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 [NASA ADS] [CrossRef] [Google Scholar]
  9. Calabrò, A., Castellano, M., Pentericci, L., et al. 2021, A&A, 646, A39 [EDP Sciences] [Google Scholar]
  10. Calabrò, A., Guaita, L., Pentericci, L., et al. 2022a, A&A, 664, A75 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  11. Calabrò, A., Pentericci, L., Talia, M., et al. 2022b, A&A, 667, A117 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  12. Cardiel, N., Gorgas, J., Cenarro, J., & Gonzalez, J. J. 1998, A&AS, 127, 597 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  13. Carnall, A. C., McLure, R. J., Dunlop, J. S., et al. 2019, MNRAS, 490, 417 [Google Scholar]
  14. Carnall, A. C., Walker, S., McLure, R. J., et al. 2020, MNRAS, 496, 695 [Google Scholar]
  15. Carnall, A. C., McLure, R. J., Dunlop, J. S., et al. 2022, ApJ, 929, 131 [NASA ADS] [CrossRef] [Google Scholar]
  16. Castellano, M., Belfiori, D., Pentericci, L., et al. 2023, A&A, 675, A121 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  17. Cimatti, A., Cassata, P., Pozzetti, L., et al. 2008, A&A, 482, 21 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  18. Cullen, F., McLure, R. J., Khochfar, S., et al. 2018, MNRAS, 476, 3218 [Google Scholar]
  19. Cullen, F., McLure, R. J., Dunlop, J. S., et al. 2019, MNRAS, 487, 2038 [Google Scholar]
  20. Cullen, F., McLure, R. J., Dunlop, J. S., et al. 2020, MNRAS, 495, 1501 [Google Scholar]
  21. Daddi, E., Renzini, A., Pirzkal, N., et al. 2005, ApJ, 626, 680 [NASA ADS] [CrossRef] [Google Scholar]
  22. Fanfani, V. 2019, PhD thesis, University of Bologna, Italy [Google Scholar]
  23. Fontanot, F., Calabrò, A., Talia, M., et al. 2021, MNRAS, 504, 4481 [CrossRef] [Google Scholar]
  24. Garilli, B., Le Fèvre, O., Guzzo, L., et al. 2008, A&A, 486, 683 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  25. Garilli, B., Fumana, M., Franzetti, P., et al. 2010, PASP, 122, 827 [Google Scholar]
  26. Garilli, B., Paioro, L., Scodeggio, M., et al. 2012, PASP, 124, 1232 [Google Scholar]
  27. Garilli, B., McLure, R., Pentericci, L., et al. 2021, A&A, 647, A150 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  28. Grogin, N. A., Kocevski, D. D., Faber, S. M., et al. 2011, ApJS, 197, 35 [NASA ADS] [CrossRef] [Google Scholar]
  29. Guaita, L., Talia, M., Pentericci, L., et al. 2017, A&A, 606, A19 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  30. Guaita, L., Pompei, E., Castellano, M., et al. 2020, A&A, 640, A107 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Guzzo, L., Scodeggio, M., Garilli, B., et al. 2014, A&A, 566, A108 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  32. Hamadouche, M. L., Carnall, A. C., McLure, R. J., et al. 2022, MNRAS, 512, 1262 [NASA ADS] [CrossRef] [Google Scholar]
  33. Hamadouche, M. L., Carnall, A. C., McLure, R. J., et al. 2023, MNRAS, 521, 5400 [NASA ADS] [CrossRef] [Google Scholar]
  34. Hoag, A., Treu, T., Pentericci, L., et al. 2019, MNRAS, 488, 706 [Google Scholar]
  35. Koekemoer, A. M., Faber, S. M., Ferguson, H. C., et al. 2011, ApJS, 197, 36 [NASA ADS] [CrossRef] [Google Scholar]
  36. Kornei, K. A., Shapley, A. E., Erb, D. K., et al. 2010, ApJ, 711, 693 [NASA ADS] [CrossRef] [Google Scholar]
  37. Kurk, J., Cimatti, A., Daddi, E., et al. 2013, A&A, 549, A63 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. Le Fèvre, O., Vettolani, G., Garilli, B., et al. 2005, A&A, 439, 845 [Google Scholar]
  39. Le Fèvre, O., Cassata, P., Cucciati, O., et al. 2013, A&A, 559, A14 [Google Scholar]
  40. Le Fèvre, O., Tasca, L. A. M., Cassata, P., et al. 2015, A&A, 576, A79 [Google Scholar]
  41. Leitherer, C., Tremonti, C. A., Heckman, T. M., & Calzetti, D. 2011, AJ, 141, 37 [NASA ADS] [CrossRef] [Google Scholar]
  42. Lilly, S. J., Le Fèvre, O., Renzini, A., et al. 2007, ApJS, 172, 70 [Google Scholar]
  43. Llerena, M., Amorín, R., Cullen, F., et al. 2022, A&A, 659, A16 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  44. Magliocchetti, M., Pentericci, L., Cirasuolo, M., et al. 2020, MNRAS, 493, 3838 [Google Scholar]
  45. Maraston, C., Nieves Colmenárez, L., Bender, R., & Thomas, D. 2009, A&A, 493, 425 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  46. Marchi, F., Pentericci, L., Guaita, L., et al. 2019, A&A, 631, A19 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  47. Markwardt, C. B. 2009, ASP Conf. Ser., 411, 251 [Google Scholar]
  48. Mascia, S., Pentericci, L., Saxena, A., et al. 2023, A&A, 674, A221 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  49. McLure, R. J., Pentericci, L., Cimatti, A., et al. 2018, MNRAS, 479, 25 [NASA ADS] [Google Scholar]
  50. Oke, J. B., & Gunn, J. E. 1983, ApJ, 266, 713 [NASA ADS] [CrossRef] [Google Scholar]
  51. Pentericci, L., McLure, R. J., Garilli, B., et al. 2018a, A&A, 616, A174 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  52. Pentericci, L., Vanzella, E., Castellano, M., et al. 2018b, A&A, 619, A147 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  53. Saldana-Lopez, A., Schaerer, D., Chisholm, J., et al. 2023, MNRAS, 522, 6295 [NASA ADS] [CrossRef] [Google Scholar]
  54. Saxena, A., Pentericci, L., Mirabelli, M., et al. 2020a, A&A, 636, A47 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  55. Saxena, A., Pentericci, L., Schaerer, D., et al. 2020b, MNRAS, 496, 3796 [Google Scholar]
  56. Saxena, A., Ellis, R. S., Förster, P. U., et al. 2021, MNRAS, 505, 4798 [NASA ADS] [CrossRef] [Google Scholar]
  57. Saxena, A., Cryer, E., Ellis, R. S., et al. 2022, MNRAS, 517, 1098 [Google Scholar]
  58. Schreiber, C., Glazebrook, K., Nanayakkara, T., et al. 2018, A&A, 618, A85 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  59. Scodeggio, M., Franzetti, P., Garilli, B., et al. 2005, PASP, 117, 1284 [NASA ADS] [CrossRef] [Google Scholar]
  60. Scodeggio, M., Guzzo, L., Garilli, B., et al. 2018, A&A, 609, A84 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  61. Spinrad, H., Dey, A., Stern, D., et al. 1997, ApJ, 484, 581 [NASA ADS] [CrossRef] [Google Scholar]
  62. Steidel, C. C., Rudie, G. C., Strom, A. L., et al. 2014, ApJ, 795, 165 [Google Scholar]
  63. Straatman, C. M. S., van der Wel, A., Bezanson, R., et al. 2018, ApJS, 239, 27 [NASA ADS] [CrossRef] [Google Scholar]
  64. Talia, M., Mignoli, M., Cimatti, A., et al. 2012, A&A, 539, A61 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  65. Thomas, R., Pentericci, L., Le Fevre, O., et al. 2020, A&A, 634, A110 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  66. Thomas, R., Pentericci, L., Le Fèvre, O., et al. 2021, A&A, 650, A63 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  67. Tomasetti, E., Moresco, M., Borghi, N., et al. 2023, A&A, in press, https://doi.org/18.1851/8884-6361/282346992 [Google Scholar]
  68. Trager, S. C., Worthey, G., Faber, S. M., Burstein, D., & González, J. J. 1998, ApJS, 116, 1 [Google Scholar]
  69. van der Wel, A., Noeske, K., Bezanson, R., et al. 2016, ApJS, 223, 29 [Google Scholar]
  70. van der Wel, A., Bezanson, R., D'Eugenio, F., et al. 2021, ApJS, 256, 44 [NASA ADS] [CrossRef] [Google Scholar]
  71. Worthey, G., & Ottaviani, D. L. 1997, ApJS, 111, 377 [CrossRef] [Google Scholar]
1

Visible Multi Object Spectrograph (VIMOS); Very Large Telescope (VLT).

2

KBSS = Keck Baryonic Structure Survey.

5

This is the extension NOISE in the multi-extension FITS flies distributed through the VANDELS collaboration website and the column ERR in the FITS binary tables downloadable from the ESO archive.

6

There are entries in She catalogue for these objects, but all cells in the table were let to –99.0.

All Tables

Table 1

slinefit parameters.

In the text
Table 2

Measured spectral lines (Gaussian fit).

In the text
Table 3

pylick spectral indices.

In the text
Table 4

Legend of catalogue content.

In the text

All Figures

thumbnail Fig. 1

Redshift distribution of the final VANDELS sample: the grey histogram includes all measurements, and the green histogram includes only reliable redshifts (confidence flag > 1 ; see the text for more details).
In the text
thumbnail Fig. 2

Construction of the 1D mock spectra for the slinefit code validation. Top: synthetic rest-frame template (from Talia et al. 2012), normalised to unity at 1750 Å. Bottom: comparison between three examples of mock 1D spectra and real VANDELS spectra. Mock spectra are shown in black. VANDELS spectra are colour-coded with respect: to their depth: 20 h (red), 40 h (green), and 80 Is (blue).
In the text
thumbnail Fig. 3

Comparison between slinefit results and input values for the sample of 270 mock spectra. In the top panels, we plotted the pull distributions. As a reference, we marked with a grey curve a Gaussian with a null mean and unity sigma. In the bottom panels, we plotted the relative change of the measured spectral quantities, with respect to their input values, as a function of the peak S/N of the lines. Black squares represent the median values of the relative change in bins of the S/N; error bars are the semi-interquartile range (SIQR). The line parameters are, starting clockwise from the top left figure: line centroid, EW, FWHM, and flux.
In the text
thumbnail Fig. 4

Median composite spectrum of VANDELS sources (grey). The upper and lower panels show zoomed-in regions around the HeIIλ1640+OIII]λ1666 and CIII]λ1909 lines, respectively. The central bandpasses, as indicated in Table 3, are marked in black, while the two local continuum windows are marked in blue and red. The green points and dashed lines indicate the mean flux in the continuum bandpasses and the linear pseudo-continuum.
In the text
thumbnail Fig. 5

Examples illustrating the four Lyα categories from Kornei et al. (2010). Clockwise, from the top left: emission, combination, absorption, and noise.
In the text
thumbnail Fig. 6

Distribution of the multiplicative scaling factors to correct the mismatch between the error spectra and the noise of the object spectra.
In the text
thumbnail Fig. 7

Distributions of the D4000 break and the EW of some notable lines. In each panel we show the distributions at S/N ≥ 1 (light grey) and S/N ≥ 3 (green). For the lines from the 'Gaussian fit' catalogues, the cut is in S/N flux. Top left: Lya EW (direct integration). Top right: D4000 (direct integration). Middle left: CIII]λ909 Å EW (Gaussian fit). Middle right: [OII]λ3727 Å EW (Gaussian fit). Bottom left: OI+SiIIλ303 Å EW (direct integration). Bottom right: OI+SiIIλ1303 A EW (Gaussian fit). In the last two panels, we also show the distribution at S/N > 2 (dark grey).
In the text
thumbnail Fig. 8

Examples of EW comparison between Gaussian fit and direct integration methods for a single line (SiIIλ/1526 Å; red triangles) and an unresolved group of lines (OI+SiIIλ/1303 Å; green squares). Only measurements at S/N > 3 are shown. The 1-to-1 relation is also indicated in black.
In the text
thumbnail Fig. 9

Comparison between trie measurements presented in this work and previously published VANDELS results. In all plots we also show the one-to-one relation (dashed red line). Top left: CIII]λ1909Å flux from Calabrò et al. (2022b; Gaussian fit). Top right: interstellar medium absorption line centroids from Calabrò et al. (2022b; Gaussian fit; the points for the different ions have been shifted by 0.01 for visualisation purposes). Middle left: Lyα flux from Guaita et al. (2020; Gaussian fit; no error was available for these measurements). Middle right: AGN emission line flux trom Bongiorno et al. (in prep.; Gaussian tit). Bottom: Dn4000 from Hamadouche et al. (2022; direct integration).
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.