J-PLUS: Systematic impact of metallicity on photometric calibration with the stellar locus
C. López-Sanjuan, H. Yuan, H. Vázquez Ramió, J. Varela, D. Cristóbal-Hornillos, P.-E. Tremblay, A. Marín-Franch, A. J. Cenarro, A. Ederoclite, E. J. Alfaro, A. Alvarez-Candal, S. Daflon, A. Hernán-Caballero, C. Hernández-Monteagudo, F. M. Jiménez-Esteban, V. M. Placco, E. Tempel, J. Alcaniz, R. E. Angulo, R. A. Dupke, M. Moles, L. Sodré Jr
AAstronomy & Astrophysics manuscript no. main © ESO 2021February 1, 2021
J-PLUS: Systematic impact of metallicity on photometriccalibration with the stellar locus
C. López-Sanjuan , H. Yuan , H. Vázquez Ramió , J. Varela , D. Cristóbal-Hornillos , P. -E. Tremblay ,A. Marín-Franch , A. J. Cenarro , A. Ederoclite , E. J. Alfaro , A. Alvarez-Candal , , , S. Daflon ,A. Hernán-Caballero , C. Hernández-Monteagudo , , , F. M. Jiménez-Esteban , , V. M. Placco , E. Tempel ,J. Alcaniz , R. E. Angulo , , R. A. Dupke , , , M. Moles , and L. Sodré Jr. Centro de Estudios de Física del Cosmos de Aragón (CEFCA), Unidad Asociada al CSIC, Plaza San Juan 1, 44001 Teruel, Spaine-mail: [email protected] Department of Astronomy, Beijing Normal University, Beijing 100875, People’s Republic of China Centro de Estudios de Física del Cosmos de Aragón (CEFCA), Plaza San Juan 1, 44001 Teruel, Spain Department of Physics, University of Warwick, Coventry, CV4 7AL, UK Instituto de Astronomia, Geofísica e Ciências Atmosféricas, Universidade de São Paulo, 05508-090 São Paulo, Brazil Instituto de Astrofísica de Andalucía (IAA-CSIC), Glorieta de la astronomía s / n, 18008 Granada, Spain IUFACyT, Universidad de Alicante, San Vicent del Raspeig, 03080 Alicante, Spain Observatório Nacional, Rua General José Cristino, 77 - Bairro Imperial de São Cristóvão, 20921-400 Rio de Janeiro, Brazil Instituto de Astrofísica de Canarias (IAC), 38205 La Laguna, Spain Departamento de Astrofísica, Universidad de La Laguna (ULL), 38200 La Laguna, Spain Centro de Astrobiología (CSIC-INTA), ESAC Campus, Camino Bajo del Castillo s / n, 28692 Villanueva de la Cañada, Spain Spanish Virtual Observatory, 28692 Villanueva de la Cañada, Spain Community Science and Data Center / NSF’s NOIRLab, 950 N. Cherry Ave., Tucson, AZ 85719, USA Tartu Observatory, University of Tartu, Observatooriumi 1, 61602 Tõravere, Estonia Donostia International Physics Centre (DIPC), Paseo Manuel de Lardizabal 4, 20018 Donostia-San Sebastián, Spain IKERBASQUE, Basque Foundation for Science, 48013, Bilbao, Spain University of Michigan, Department of Astronomy, 1085 South University Ave., Ann Arbor, MI 48109, USA University of Alabama, Department of Physics and Astronomy, Gallalee Hall, Tuscaloosa, AL 35401, USAReceived 29 January 2021
ABSTRACT
Aims.
We present the updated photometric calibration of the twelve optical passbands for the Javalambre Photometric Local UniverseSurvey (J-PLUS) second data release (DR2), comprising 1 088 pointings of two square degrees, and study the systematic impact ofmetallicity in the stellar locus technique.
Methods.
The [Fe / H] metallicity from LAMOST DR5 for 146 184 high-quality calibration stars, defined with S / N >
10 in J-PLUSpassbands and S / N > Gaia parallax, was used to define the metallicity-dependent stellar locus (ZSL). The initial homogenizationof J-PLUS photometry, performed with a unique stellar locus, was refined by including the metallicity e ff ect in colours via the ZSL. Results.
The variation of the average metallicity along the Milky Way produces a systematic o ff set in J-PLUS calibration. This e ff ectis well above 1% for the bluer passbands and amounts 0 .
07, 0 .
07, 0 .
05, 0 .
03, and 0 .
02 mag in u , J J J J ff ect with the Milky Way location of the J-PLUS pointing, providing also an updated calibration forthose observations without LAMOST information. The estimated accuracy in the calibration after including the metallicity e ff ect isat 1% level for the bluer J-PLUS passbands and below for the rest. Conclusions.
Photometric calibration with the stellar locus technique is prone to significant systematic bias along the Milky Waylocation for passbands bluer than λ = Key words. methods:statistical, techniques:photometric, surveys
1. Introduction
One fundamental step in the data processing of any imaging sur-vey is the photometric calibration of the observations. The cal-ibration process aims to translate the observed counts in astro-nomical images to a physical flux scale referred to the top ofthe atmosphere. Because accurate colours are needed to derivephotometric redshifts for galaxies, atmospheric parameters forMilky Way (MW) stars, and surface characteristics for minorbodies; and reliable absolute fluxes are involved in the estima- tion of the luminosity and the stellar mass of galaxies, currentand future photometric surveys target a calibration uncertainty atthe 1% level and below to reach their ambitious scientific goals.One particular approach to perform the photometric calibra-tion is the use of the stellar locus (Covey et al. 2007; High et al.2009; Kelly et al. 2014). This procedure takes advantage ofthe way stars with di ff erent stellar parameters populate colour-colour diagrams, defining a well-constrained region (stellar lo-cus) whose shape depends on the specific colours used. The Article number, page 1 of 14 a r X i v : . [ a s t r o - ph . GA ] J a n able 1. J-PLUS photometric system, extinction coe ffi cients, and limiting magnitudes (5 σ , 3 (cid:48)(cid:48) aperture) of J-PLUS DR2. Passband ( X ) Central Wavelength FWHM m DR2lim k X = A X E ( B − V ) Comments[nm] [nm] [AB] u J J + K; similar to the CaHK filter from
PristineJ δ J g J b Triplet r J α ; in common with J-PAS i J z ff ected by the amount of interstellar matter thatstar-light passes through until reaching the observer and by pos-sible local variations in the extinction law. This leaves two solu-tions to define the reference locus: de-reddening the photometry,which implies knowing (or assuming) in each case the local ex-tinction law, or choosing a set of dust-free objects. It would be away of saying that we need photometry "outside" the Galaxy toset the stellar locus as reference (High et al. 2009).In addition to the interstellar extinction, the average proper-ties of the stars also varies with their position in the MW. Thestellar locus location for main sequence (MS) stars is dominatedby temperature variations, so the measured correlation in colour-colour diagrams is roughly a temperature sequence. However,the metallicity is also a relevant parameter that a ff ects apprecia-bly the stellar locus location (e.g. Lenz et al. 1998; Ivezi´c et al.2008; Yuan et al. 2015a; Kesseli et al. 2017), specially at thebluer optical passbands. With the average metallicity of the ob-served MW stars decreasing as we move from disk-dominated([Fe / H] ∼ − . / H] ∼ − . ff ectshall be the leading source of systematic in the calibration withthe stellar locus technique.Several large-area photometric surveys covering the blueedge ( λ < ugriz broad bands), the Pristine survey (Starkenburg et al. 2017; a unique CaHK filter of 98 Åwidth centered at 3952 Å), and the Javalambre Photometric Lo-cal Universe Survey (J-PLUS, Cenarro et al. 2019; 5 broad + u band with the Sloan Digital Sky Survey (SDSS, Ai-hara et al. 2011) photometry reveals a systematic variation withgalactic latitude, that the authors link to the change in metallicity(Kuijken et al. 2019). Furthermore, the photometric metallicity derived from Pristine presents a systematic variation with thesky position when the stellar locus calibration is performed with0 . < ( g − i ) < . . < ( g − i ) < . , was made public inNovember 2020, and we describe here its photometric calibra-tion. It is based on the stellar and white dwarf loci proceduredetailed in López-Sanjuan et al. (2019) and that was applied toJ-PLUS first data release (DR1). In the present paper, we tookadvantage of the [Fe / H] information provided by the Large SkyArea Multi-Object Fibre Spectroscopic Telescope (LAMOST,Cui et al. 2012) surveys to implement the metallicity-dependentstellar locus for calibration. That improved the accuracy ofthe J-PLUS DR2 photometry, specially at passbands bluer than λ = ff ects are neglected.In addition to a metallicity-dependent stellar locus, the ac-cess to massive spectroscopic information also permits the appli-cation of the Stellar Color Regression (SCR, Yuan et al. 2015b;Huang et al. 2020) method. Using T e ff , log g , and [Fe / H] fromspectroscopy, the SCR matches stars of the same properties, i.e.intrinsic colours, and assigns observed colour di ff erences to thee ff ect of interstellar extinction. This permits to homogenize thephotometric solution by naturally accounting for temperature,gravity, and metallicity e ff ects. The application of the SCR toJ-PLUS data is beyond the scope of the present paper, and it isexplored in a forthcoming work.This paper is organized as follows. The J-PLUS DR2 andthe ancillary data used on its calibration are presented in Sect. 2.The calibration methodology is summarised in Sect. 3, high-lighting the addition of the metallicity-dependent stellar locusin the process. The precision, accuracy, and the systematic im-pact of metallicity in the J-PLUS DR2 calibration are discussedin Sect. 4. We devoted Sect. 5 to summarise the work and presentour conclusions. Magnitudes are given in the AB system (Oke& Gunn 1983). Article number, page 2 of 14ópez-Sanjuan et al.: J-PLUS. Systematic impact of metallicity on photometric calibration with the stellar locus
2. J-PLUS photometric data
J-PLUS is being conducted from the Observatorio Astrofísicode Javalambre (OAJ, Teruel, Spain; Cenarro et al. 2014) usingthe 83 cm Javalambre Auxiliary Survey Telescope (JAST / T80)and T80Cam, a panoramic camera of 9.2k × field of view (FoV) with a pixel scale of0.55 (cid:48)(cid:48) pix − (Marín-Franch et al. 2015). The J-PLUS filter sys-tem, composed of twelve bands, is summarized in Table 1.These filters were designed to optimise the characterization ofMW stars. The J-PLUS observational strategy, image reduction,and main scientific goals are presented in Cenarro et al. (2019).In addition to its scientific potential, J-PLUS was defined withthe technical goal of ensure the photometric calibration of theJavalambre Physics of the Accelerating Universe AstrophysicalSurvey (J-PAS; Benítez et al. 2014; Bonoli et al. 2020), that willscan thousands of square degrees with 56 narrow bands of ∼ m ∼ . ) ob-served and reduced in all survey bands (Varela & J-PLUS collab-oration 2021). The limiting magnitudes (5 σ , 3 (cid:48)(cid:48) aperture) of theDR2 are presented in Table 1 for reference. The median pointspread function (PSF) full width at half maximum (FWHM) inthe DR2 r -band images is 1.1 (cid:48)(cid:48) . Source detection was done inthe r band using SExtractor (Bertin & Arnouts 1996), and theflux measurement in the twelve J-PLUS bands was performed atthe position of the detected sources using the aperture defined inthe r -band image. Objects near the borders of the images, closeto bright stars, or a ff ected by optical artefacts were masked. TheDR2 is publicly available at the J-PLUS website .The calibration process presented in Sect. 3 uses J-PLUSDR2 in combination with ancillary data from Gaia , thePanoramic Survey Telescope and Rapid Response System (Pan-STARRS), and LAMOST. We describe these datasets in the fol-lowing.
The Pan-STARRS1 is a 1.8 m optical and near-infrared telescopelocated on Mount Haleakala, Hawaii. The telescope is equippedwith the Gigapixel Camera 1 (GPC1), consisting of an array of60 CCD detectors, each 4 800 pixels by side (Chambers et al.2016).The 3 π Steradian Survey (hereafter PS1; Chambers et al.2016) covers the sky north of declination δ = − ◦ in fourSDSS-like passbands, griz , with an additional passband in thenear-infrared, y . The entire filter set spans the range 400 − σ depths of 22.0, 21.8, 21.5, 20.9,and 19.7 in grizy , respectively (Chambers et al. 2016). The PS1DR1 occurred in December 2016, and provided a static-sky cat-alogue, stacked images from the 3 π Steradian Survey, and otherdata products (Flewelling et al. 2016).Because of its large footprint, homogeneous depth, and ex-cellent internal calibration, PS1 photometry provides an idealreference for the calibration of the gri
J-PLUS broad bands. The z − band photometry from PS1 was reserved to test the calibrationprocedure. The
Gaia spacecraft is mapping the 3D positions and kinemat-ics of a representative fraction of MW stars (Gaia Collabora-tion et al. 2016). The mission will eventually provide astrometry(positions, proper motions, and parallaxes) and optical spectro-photometry for over a billion stars, as well as radial velocitymeasurements of more than 100 million stars.In the present paper, we used the
Gaia
DR2 (Gaia Collabo-ration et al. 2018), that is based on 22 months of observations. Itcontains five-parameter astrometric determinations and providesintegrated photometry in three broad bands, G (330 − G BP (330 −
680 nm), and G RP (630 − G <
21. The typical uncertainties in
Gaia
DR2measurements at G =
17 are ∼ . ∼ G -band photometry, and ∼
10 mmag in G BP and G RP magnitudes (Gaia Collaboration et al. 2018). LAMOST is a 4-metre quasi-meridian reflecting Schmidt tele-scope equipped with thousands of fibers distributed in a FoV ofabout 20 deg . It can simultaneously collect spectra per exposureof up to 4 000 objects, covering the wavelength range 380 − R ∼ A, F, G, and K type starcatalog , that includes the basic stellar parameters T e ff , log g ,and [Fe / H] derived with the LAMOST stellar parameter pipeline(LASP; Wu et al. 2011; Luo et al. 2015).
3. Photometric calibration of J-PLUS DR2
The photometric calibration of the J-PLUS DR2 data followsthe main steps presented in López-Sanjuan et al. (2019) for thecalibration of J-PLUS DR1. We provide a brief summary of theprocess in Sect. 3.1. The main improvement with respect to DR1procedure is the inclusion of the metallicity e ff ect in the stellarlocus location, as detailed in Sect. 3.2.The goal of any calibration strategy is to obtain the zero point(ZP) of the observation, that relates the magnitude of the sourcesin passband X on top of the atmosphere with the magnitudesobtained from the analogue to digital unit (ADU) counts of thereduced images. We simplify the notation in the following usingthe passband name as the magnitude in such filter. Thus, X = − . (ADU X ) + ZP X . (1)In the estimation of the J-PLUS DR2 raw catalogues, the reducedimages were normalized to a one-second exposure and an arbi- http://dr5.lamost.org/v3/doc/data-production-description Article number, page 3 of 14 rary instrumental zero point ZP X =
25 was set. This defined theinstrumental magnitudes X ins .The calibration process applied in J-PLUS DR2 have di ff er-ent steps, as described in Sect. 3.1. At the end, we estimated thezero point of the passband X in the pointing p id asZP X ( p id , X , Y ) =∆ X atm ( p id ) + P X ( p id , X , Y ) + ∆ X FeH ( p id ) + ∆ X WD + , (2)where ∆ X atm is the term that accounts for the atmospheric extinc-tion at the moment of the observation, P X defines a plane thataccounts for the 2D variation of the calibration with the ( X , Y )position of the sources on the CCD, ∆ X FeH includes the e ff ect ofthe metallicity in the stellar locus homogenization process, and ∆ X WD is the global o ff set provided by the white dwarf (WD) lo-cus that translates homogenized magnitudes to calibrated mag-nitudes outside the atmosphere.The J-PLUS instrumental magnitudes used for calibrationwere measured on a 6 (cid:48)(cid:48) diameter aperture. This aperture is notdominated by background noise and limits the flux contamina-tion from neighbouring sources, although it is not large enoughto capture the total flux of the stars. Thus, we applied an aper-ture correction C aper that depends on the pointing and the pass-band. The aperture correction was computed from the growthcurves of bright, non-saturated stars in the pointing. For eachstar, increasingly larger circular apertures were measured untilconvergence within errors. This defined the aperture size thatprovides the total magnitude of the sources in the pointing, thatis then compared with the magnitude at 6 (cid:48)(cid:48) aperture to provide C aper . The typical number of stars used is 50 and the medianaperture correction varies from C aper = − .
09 mag in the u bandto C aper = − .
11 mag in the z band, with a median value of C aper = − .
09 mag for all the filters. The typical uncertaintyin the correction aperture, estimated from the dispersion in themeasurements, is ∼ (cid:48)(cid:48) magnitudes corrected for aperture e ff ects provided the total fluxof stars.We worked with dust de-reddened magnitudes and coloursin several stages of the calibration process. We empirically com-puted the extinction coe ffi cients k X of each J-PLUS passband,presented in Table 1, by applying the star-pair technique de-scribed in Yuan et al. (2013) to J-PLUS DR1. The de-reddenedJ-PLUS photometry, either instrumental or calibrated, is notedwith the subscript 0 and was obtained as X = X − k X E ( B − V ) . (3)We estimated the colour excess at infinite distance of each J-PLUS source from the Schlegel et al. (1998) extinction map.The stars used in the calibration process have distance informa-tion from Gaia
DR2 parallaxes (Sect. 3.1), and we included the3D information using the MW dust model presented in Li et al.(2018). We integrated the dust model to infinity and to the dis-tance provided by
Gaia at star location, scaling accordingly thecolour excess from Schlegel et al. (1998) map to obtain E ( B − V ).The uncertainty in E ( B − V ) was fixed to 0 .
012 mag. This errorwas estimated by comparing the colour excess directly measuredfrom the star-pair method (Yuan et al. 2013) with the assumed E ( B − V ). The dispersion in this comparison was set as the uncer-tainty in the used 3D extinction. We test the assumed extinctioncorrection in Sect. 4.3. We provide in this section a brief summary of the steps involvedin the photometric calibration of J-PLUS DR2 images. The up- u, J0378, J0395, J0410, J0430, J0515, J0660, J0861, z homogenizationMetallicity-dependentstellar locus * Absolute magnitude vs. colour diagram *Direct comparison DA and DB+DCWD locus *TheoreticalWD locus (Tremblay+13,Cukanovaite+18) gri zero points2D : Schlegel+98 : Li+18 DR2 DR2DR1
E(B-V)MS stars(main sequence) WDs(white dwarfs)
Varela+21 Gaia Collaboration+18bChambers+16 u, J0378, J0395, J0410, J0430, g, J0515, r, J0660, i, J0861, z
AB zero points
DR5 dr5.lamost.org
LAMOST
Instrumentalstellar locus *
Fig. 1.
Updated flowchart of the calibration method used in thiswork. Arrows that originate in small dots indicate that the precedingdata product is an input to the subsequent analysis. Datasets are shownwith their project logo, and external codes or models with grey boxes.The rounded-shape boxes show the calibration steps. The asterisks in-dicate those steps based on dust de-reddened magnitudes. The whiteboxes show intermediate data products, and ovals highlight data prod-ucts of the calibration process. The changes with respect to J-PLUSDR1 calibration are the modification in the assumed dust extinction andthe addition of the metallicity-dependent stellar locus step in the ho-mogenization (Sect. 3.2). dated flowchart of the calibration process is presented in Fig. 1.We refer the reader to López-Sanjuan et al. (2019) for an exten-sive description of the calibration procedure but the metallicity-dependent stellar locus step, added for J-PLUS DR2 and de-scribed in Sect. 3.2. The calibration steps were: • Definition of a high-quality sample of MS stars for cali-bration. We selected those sources in common between J-PLUS DR2 and
Gaia
DR2 with signal-to-noise (S / N) largerthan ten in all the photometric bands and with S / N > Gaia parallax. We constructed the dust de-reddened abso-lute G magnitude versus G BP − G RP diagram and selectedthose sources belonging to the main sequence. This provided Article number, page 4 of 14ópez-Sanjuan et al.: J-PLUS. Systematic impact of metallicity on photometric calibration with the stellar locus • Calibration of the gri broad-band filters with PS1 photome-try. The J-PLUS instrumental magnitudes were comparedwith the PSF magnitudes in PS1 after accounting for thecolour terms between both photometric systems. This stepprovides the ∆ X atm and the 2D variation along the CCD ofthe gri broad-band filters. Because we used PS1 calibratedmagnitudes as reference, ∆ X FeH = ∆ X WD ∼
0. Thelatter term is not zero because residual di ff erences betweenJ-PLUS and PS1 photometric systems can exist, as discussedin Sect. 4.3. • Initial homogenization of the narrow bands with the instru-mental stellar locus (ISL). For each remaining passband, wecomputed the dust de-reddened ( X ins − r ) versus ( g − i ) colour-colour diagrams of the MS calibration stars. Fromthese, we computed the o ff sets that lead to a consistent ISLamong all the J-PLUS DR2 pointings. This provides ∆ X atm and the 2D variation along the CCD for the other nine J-PLUS passbands. After this step, we defined the ISL magni-tudes as X ISL = X ins + ∆ X atm + P X . (4) • Final homogenization with the metallicity-dependent stel-lar locus (ZSL). We refined the methodology presented inLópez-Sanjuan et al. (2019) by including the e ff ect of metal-licity in the stellar locus location. We used the metallic-ity measurements from LAMOST DR5 and the procedure isfully detailed in Sect. 3.2. This step provided ∆ X FeH , defin-ing the ISL + ZSL magnitudes X ISL + ZSL = X ISL + ∆ X FeH . (5) • Absolute colour calibration with the white dwarf locus.From the
Gaia absolute magnitude versus colour diagramin the first step, we also selected 639 high-quality WDs. Wecompared the observed colour-colour locus in ( X ISL + ZSL − r ) versus ( g − i ) with the theoretical expectations from pure hy-drogen (DA; Tremblay et al. 2013) and pure helium (DB andDC; Cukanovaite et al. 2018) models. The Bayesian model-ing of the WD locus provided the ∆ X WD for all the passbandsbut r , that was used as the reference band in the process.The performance of J-PLUS DR2 calibration is presented inSect. 4. The median zero points obtained after the complete cal-ibration process are presented in Table 2 for reference. The calibration process presented in López-Sanjuan et al. (2019)and summarised in the previous section was updated for J-PLUSDR2 by including the impact of metallicity in the stellar locuslocation. We use the u band as reference to illustrate the process,and the methodology was similar for other J-PLUS passbandsbut gri , anchored to PS1 photometry. The improvement in theaccuracy of J-PLUS calibration along the surveyed area fromthis step is presented in Sect. 4.2. We started by gathering the [Fe / H] (dubbed metallicity hereafter)information of the MS calibration stars. We cross-matched thecalibration sample with the LAMOST catalogue using a 1 arcsecradius. A total of 146 184 sources in common were retrieved. ( g − i ) ( u I S L − r ) ZSL o [Fe / H] = − . − . − . − . − . . . [ F e / H ] − . − . . . . ( u ISL − r ) − ZSL o . . . . . P r o b a b li t y [Fe / H] : - . - . - . . . Fig. 2.
Top panel : Binned ( u ISL − r ) versus ( g − i ) colour-colour dia-gram. The colour scale shows the median [Fe / H] in each bin estimatedfrom LAMOST spectra. The black solid line marks the stellar locus for − . < [Fe / H] < − .
20 stars in the range 0 . < ( g − i ) < .
5, notedZSL o . Bottom panel : Normalized histogram of the ( u ISL − r ) colourdi ff erence with respect to ZSL o for samples of di ff erent metallicities,defined with a central [Fe / H] ± . The median uncertainty in [Fe / H] is 0.1 dex, providing a high-quality data set to derive the metallicity-dependent stellar locus.Despite the large sky coverage of LAMOST, not all the J-PLUS pointings have metallicity information. We have 178(16%) pointings with less than ten calibration stars in commonwith LAMOST. This implies that the metallicity-dependent stel-lar locus procedure detailed in Sect. 3.2.3 cannot be applied toall J-PLUS DR2 observations. We circumvented this limitationby modelling the o ff set in the stellar locus due to metallicity withMW location (Sect. 3.2.4). The stellar locus is known to vary with metallicity (e.g. Yuanet al. 2015a). Such variation is more prominent at blue opti-cal wavelengths, with the e ff ect in the u band being an order ofmagnitude larger than in the g band (Yuan et al. 2015a). To il-lustrate this e ff ect with J-PLUS photometry, the median [Fe / H]from LAMOST in the ( u ISL − r ) versus ( g − i ) colour-colourspace is presented in the top panel of Fig. 2. At a given ( g − i ) colour, redder stars in ( u ISL − r ) have larger metallicities.As starting point, we defined the reference stellar locus,noted ZSL o , from those stars with − . < [Fe / H] < − .
20 inthe colour range 0 . < ( g − i ) < .
5. This metallicity range waschosen to cover the density peak in the distribution of LAMOST
Article number, page 5 of 14 . . . ( g − i ) − − [ F e / H ] ( u I S L − r ) . . . ( g − i ) − − [ F e / H ] − . − . . . . ( u I S L − r ) − Z S L o Fig. 3.
Binned metallicity versus ( g − i ) colour diagram of the MScalibration stars with measurements from LAMOST. Top panel : Mean( u ISL − r ) colour in each bin, defining the metallicity-dependent stellarlocus (ZSL). Bottom panel : ( u ISL − r ) colour di ff erence with respect toZSL o . The median metallicity of the reference locus is marked with theblack dashed line. metallicities. From the ZSL o reference, the colour di ff erence forstars of di ff erent metallicities was computed, as shown in the bottom panel of Fig. 2. We find a ( u ISL − r ) colour di ff erence of − .
20 mag for [Fe / H] ∼ − . + .
23 mag for[Fe / H] ∼ . σ = .
092 mag to σ = .
047 mag after accounting for the metallicity dependence.As shown by Yuan et al. (2015a), the metallicity-dependentstellar locus, noted ZSL, is not just a shift from the reference, andthe simple modelling described above must be refined. Insteadof performing an analytic fit to the data, we mapped the mean( u ISL − r ) colour as a function of [Fe / H] and ( g − i ) . The mappingwas done with a two-dimensional histogram. The used rangeswere ( g − i ) ∈ [0 . , .
5] and [Fe / H] ∈ [ − . , . ff ective bin width was 0 .
009 magin colour and 0 .
017 dex in metallicity. The ZSL and its di ff erencewith respect to the reference locus ZLS o are shown in Fig. 3,highlighting the shift and the change in curvature of the stellarlocus with metallicity. We compare the J-PLUS ZSL with theresults from Yuan et al. (2015a) in Sect. 4.6. The ZSL estimated in the previous section can be used to com-pute the calibration o ff set due to metallicity in each J-PLUS DR2pointing, named ∆ u FeH . The star-by-star expected colour is esti- . . ( g − i ) − . . . ( u I S L − r ) − Z S L o − . − . − . − . − . . [ F e / H ] . . ( g − i ) − . . . ( u I S L − r ) − Z S L − . − . − . − . − . . [ F e / H ] ∆ u FeH = − . Fig. 4. ( u ISL − r ) colour di ff erence with respect to the referencelocus ZSL o ( top panel ) and the ZSL ( bottom panel ) as a function of ( g − i ) for the MS calibration stars with LAMOST information in pointing p id = / H] from LAMOST. The dashed lines mark zero o ff set. The dottedline marks the median di ff erence with respect to the ZSL. The derivedmetallicity o ff set is labelled in the bottom panel . mated from the ZSL and subtracted to the observed colour, δ u FeH = ( u ISL − r ) − ZSL . (6)The distribution of these di ff erences in each pointing was fittedto a Gaussian with median − ∆ u FeH , the targeted metallicity o ff setfor the pointing. We assumed that the measured o ff set is due tothe di ff erent metallicity, i.e. stellar locus location, of the stars inthe pointing with respect to the J-PLUS ISL. In this process onlyZSL bins with more than ten sources and pointings with morethan 50 LAMOST stars with δ u FeH computed were kept.We illustrate the process using the J-PLUS pointing p id = σ = .
09 mag and a clear dependence with [Fe / H] ispresent ( top panel in Fig. 4). After accounting for metallicitye ff ects with the ZSL, the dispersion reduces to σ = .
04 and the[Fe / H] gradient has disappeared ( bottom panel in Fig. 4). Themedian of the measured δ u FeH is 0 .
038 mag, and the estimatedmetallicity o ff set is then ∆ u FeH = − .
038 mag.After applying the above procedure to all J-PLUS pointings,we obtained a valid ∆ u FeH for 746 of them, 69% of the total tar-gets. We study the trends in the derived o ff sets in the next sec-tion, and also detailed how we assigned a value to those orphanpointings without a measurement of the metallicity o ff set. Article number, page 6 of 14ópez-Sanjuan et al.: J-PLUS. Systematic impact of metallicity on photometric calibration with the stellar locus l [deg] − b [ d e g ] − ∆ u F e H [ mm ag ] l [deg] − b [ d e g ] − F u [ mm ag ] l [deg] − b [ d e g ] − ∆ u F e H [ mm ag ] −
50 0 50 ∆ u FeH × P r o b a b ili t y σ ISL+ZSL = 7 mmag σ ISL = 24 mmag
Fig. 5.
Photometric o ff set for each J-PLUS DR2 pointing estimated from the metallicity-dependent stellar locus. Top left panel : Initial o ff setin galactic coordinates with the homogenization from ISL, ∆ u FeH . Top right panel : Modelled metallicity o ff set F u in galactic coordinates. Thosepointings without o ff set estimation and not used in the modelling procedure are highlighted with a black edge. Bottom left panel : Final metallicityo ff set in galactic coordinates after the homogenization from ISL + ZSL, ∆ u FeH = ∆ u FeH − F u . Bottom right panel : Distribution of the metallicityo ff sets ∆ u FeH (gray) and ∆ u FeH (coloured). The Gaussian distributions that better describe the data are also shown, with their dispersion labelled inthe panel. The dotted line marks zero o ff set. The metallicity o ff sets for each J-PLUS pointing with a validmeasurement are shown as a function of galactic coordinates inthe top left panel of Fig. 5 (2D representation) and in the top pan-els of Fig. 6 (1D representation). We find a systematic trend inthe o ff sets, changing from ∆ u FeH ∼ + .
05 mag to ∆ u FeH ∼ − . / H] of the MS calibration stars with LAMOSTinformation, noted (cid:104) [Fe / H] (cid:105) LAMOST , that changes from − . − . ff sets is σ ISL = bottom right panel in Fig. 5), that translates to the ob-served edge-to-edge di ff erence of ∼ . ff sets is their systematic variation, thattranslates to a systematic shift in the calibration and poses a limi-tation to the scientific cases that depends on the information fromthe bluer J-PLUS passbands. As an example, we explore the im-pact in the estimation of photometric metallicity in Sect. 4.6.To correct for the metallicity o ff sets, we modelled their vari-ation with Galactic coordinates using a fourth degree polynomialfit, F u ( l , b ) = (cid:88) m , n = C mn × l m × b n , (7) where ( l , b ) are the galactic longitude and latitude of the J-PLUSpointings, and C mn are the coe ffi cients of the polynomial. Thismodelling assumes a smooth variation of the metallicity, i.e. ofthe calibration o ff sets along the Galaxy. In addition, it permitsto assign a metallicity o ff set to those orphan pointings without avalid measurement. As a drawback, local metallicity variationscan still a ff ect the calibration and in several cases the o ff sets areextrapolated from the area with available information.The model F u was applied as a proxy for the metallicity o ff -set in Eq. (2). We note that this action changes the photometryof the J-PLUS stars used to compute the ZSL. To ensure self-consistency, we computed an updated version of the ZSL afterobtaining the new calibration and iterate the process until con-vergence. It took four iterations to reach variations lower than 1mmag in the measured metallicity o ff sets.The final estimated model F u for all the J-PLUS DR2 point-ings is presented in the top right panel of Fig. 5. The final residu-als, noted ∆ u FeH = ∆ u FeH − F u , have a dispersion of σ ISL + ZLS = bottom rightpanel in Fig. 5). The improvement is also clear in the lower pan-els of Figs. 5 and 6, where the initial structures are suppressedand no systematic variations with the pointing location remain.This implies that the original systematic error is now statistical,greatly improving the accuracy of the J-PLUS calibration alongthe surveyed area (Sect. 4.2). Article number, page 7 of 14
50 0 50 b [deg] − ∆ u F e H [ mm ag ] − . − . − . − . − . h [ F e / H ] i L A M O S T
200 100 0 100 200 300 l [deg] − ∆ u F e H [ mm ag ] − . − . − . − . − . h [ F e / H ] i L A M O S T b < b > −
50 0 50 b [deg] − ∆ u F e H [ mm ag ] − . − . − . − . − . h [ F e / H ] i L A M O S T
200 100 0 100 200 300 l [deg] − ∆ u F e H [ mm ag ] − . − . − . − . − . h [ F e / H ] i L A M O S T b < b > Fig. 6.
Photometric o ff set estimated from the metallicity-dependent stellar locus. Top panels : Initial o ff set with the homogenization from ISL, ∆ u FeH . Bottom panels : Final o ff set with the homogenization from ISL + ZS, ∆ u FeH = ∆ u FeH − F u . The left panels show the dependence on galacticlatitude b and the right panels on galactic longitude l , showing pointings with positive and negative latitudes separately. In all the panels the colourscale shows the median metallicity in the pointing estimated from LAMOST spectra. As a summary of this section, we have estimated and cor-rected the systematic impact of the varying MW metallicity inthe stellar locus calibration. We have used the u passband asillustrative example, and the results for the other J-PLUS pass-bands are presented in Sect. 4.2.
4. Error budget and the impact of metallicity onphotometric calibration
This section is devoted to the error budget analysis and the im-pact of the metallicity in the J-PLUS DR2 calibration. We studythe precision in the photometry in Sect. 4.1, and the accuracyalong the surveyed area in Sect. 4.2. The uncertainty in the ab-solute calibration is discussed in Sect. 4.3.
J-PLUS pointings slightly overlap with each other. We measuredthe precision of the calibration by comparing the photometry ofthose MS calibration stars observed independently in the over-lapping areas between adjacent pointings. We computed the dif-ferences in the calibrated magnitudes and estimated the medianof the sources shared by every pair of overlapping pointings. Weobtained 2 449 unique pair pointings in J-PLUS DR2. The dis-tribution of these median di ff erences was then used to estimatethe precision in the calibration. The distributions are describedby Gaussian functions and the desired precision is obtained as σ/ √
2, where σ is the measured dispersion. We find that the precision obtained in X ISL + ZSL magnitudes issimilar and replicates the results from J-PLUS DR1 at one mmaglevel. The results are summarised in Table 2. The measured pre-cision is ∼
18 mmag in u , J J ∼ J J ∼ g , J r , J i , J z . We also find that the results with X ISL magnitudes mimicthose in Table 2. The negligible change with respect to DR1 andafter applying the ZSL reflects that metallicity variations alongthe MW impacts the calibration at scales larger than a few squaredegrees. This limited local impact is exacerbated when distantpointings are compared, as analysed in the next section.
The comparison of the photometry in adjacent pointings is notable to provide a measurement of the accuracy of the calibrationalong the surveyed area. This was a drawback of the analysisperformed with J-PLUS DR1 by López-Sanjuan et al. (2019).As shown in Sect. 3.2.3, the systematic variation of the metal-licity along the MW accordingly produces a systematic o ff set inthe photometric solution. The metallicity o ff sets ∆ X FeH providestherefore a measurement of the accuracy in the calibration alongthe J-PLUS DR2 surveyed area (Fig. 7).The dispersion in the metallicity o ff sets when X ISL magni-tudes were used, noted σ accISL , are gathered in Table 2. However,the systematic nature of the o ff sets, with a clear smooth variationwith galactic latitude (Fig. 6), implies that the relevant measure- Article number, page 8 of 14ópez-Sanjuan et al.: J-PLUS. Systematic impact of metallicity on photometric calibration with the stellar locus
Table 2.
Estimated error budget of the J-PLUS DR2 photometric calibration and final median zero points.
Precision AccuracyPassband σ preISL + ZSL σ WD σ cal σ accISL s ISL σ accISL + ZSL σ accSCR (cid:104) ZP X (cid:105) [mmag] a [mmag] b [mmag] c [mmag] d [mmag] e [mmag] f [mmag] g [mag] u
17 4 18 24 65 7 11 21.16 J J J J g · · · · · · · · · J r · · · · · · · · · · · · J i · · · · · · · · · J z Notes. (a)
Instrumental stellar locus (ISL), the plane correction to account for 2D variations along the CCD, and the metallicity-dependent stellarlocus (ZSL) were used to homogenize the photometry. The calibration was anchored to PS1 photometry for gri passbands. Precision estimatedfrom duplicated MS stars in overlapping pointings. (b)
Uncertainty in the colour calibration from the Bayesian analysis of the white dwarf locus (Sect. 4.3). (c)
Final precision in the J-PLUS DR2 flux calibration, σ = σ + ZSL + σ + σ r , where σ r = (d) Dispersion in the metallicity o ff sets ∆ X FeH when the ISL was used to homogenize the photometry. (e)
Accuracy along the surveyed area estimated from the di ff erence between the percentile 95 and the percentile 5 in ∆ X FeH distribution when theISL magnitudes were used. (f)
Accuracy along the surveyed area estimated from the dispersion in ∆ X FeH when the ISL + ZSL were used to homogenize the photometry. (g)
Accuracy estimated from the comparison of the final ISL + ZSL calibration with results from the Stellar Color Regression method (Sect. 4.5). ment of the accuracy is not the dispersion, but the edge-to-edge( ± σ ) variation. In this context, we estimated the accuracy asthe o ff set di ff erence between the 5th percentile and the 95th per-centile in ∆ X FeH distribution. This measurement is expressedas s ISL in Table 2. We find that the calibration accuracy whenmetallicity e ff ects are neglected is well above 1% for the pass-bands at λ < s ISL ∼ .
07, 0 .
07, 0 . .
03, and 0 .
02 mag in u , J J J J ∼ .
01 mag in J z , and negligible in the J J ff sets by a factor of two-three (Fig. 7), but also removes themain systematic errors. This is, the dispersion is now a propermeasurement of the accuracy in the calibration. The final uncer-tainty estimated for J-PLUS DR2 is summarised in Table 2 and itis at 1% level or below. The improvement in the bluer passbandsis roughly a factor ten, decreasing from s ISL ∼ −
20 mmag to σ accISL + ZSL ∼ − The stellar locus steps, both ISL and ZSL, are devoted to the ho-mogenization of the J-PLUS photometry in those passbands notanchored to PS1. The absolute colour calibration was performedwith the white dwarf locus. Thanks to the large area observed(2 176 deg ) and the already homogenized photometry, a set of639 high-quality white dwarfs were retrieved from the Gaia ab-solute magnitude versus colour diagram. We performed a jointBayesian analysis of the eleven ( X ISL + ZSL − r ) versus ( g − i ) colour-colour diagrams to estimate the o ff sets ∆ X WD that trans-late instrumental magnitudes to calibrated magnitudes on top ofthe atmosphere. We summarise the obtained values in Table 3 forreference. The typical uncertainty in these o ff sets is at 4 mmaglevel, as presented also in Table 2.In addition to the o ff sets, the Bayesian modeling providesthe intrinsic dispersion in the WD locus (Table 3) and two phys-ical parameters of the WD population, the fraction of DA andthe median gravity. We find a DA fraction of f DA = . ± . g = . ± .
04. Both values are consistentwith J-PLUS DR1 results in López-Sanjuan et al. (2019) and themedian surface gravity agrees with the literature (e.g. Jiménez-Esteban et al. 2018; Gentile Fusillo et al. 2019; Tremblay et al.2019; Bergeron et al. 2019, and references therein). We referthe reader to López-Sanjuan et al. (2019) for a detailed descrip-tion of the Bayesian modelling and the assumptions in the whitedwarf locus step.A relevant change with respect to J-PLUS DR1 is on the in-ferred o ff sets in the g and i passbands. We obtained ∆ g WD = ∆ i WD = − ∆ g WD = − ∆ i WD = ffi cients, as detailed in the next section. Article number, page 9 of 14 −
50 0 50 100 ∆ J FeH × P r o b a b ili t y σ ISL+ZSL = 8 mmag σ ISL = 26 mmag − −
25 0 25 50 ∆ J FeH × P r o b a b ili t y σ ISL+ZSL = 6 mmag σ ISL = 17 mmag −
20 0 20 ∆ J FeH × P r o b a b ili t y σ ISL+ZSL = 4 mmag σ ISL = 10 mmag − −
10 0 10 20 ∆ J FeH × P r o b a b ili t y σ ISL+ZSL = 3 mmag σ ISL = 6 mmag − . − . . . . ∆ J FeH × P r o b a b ili t y σ ISL+ZSL = 1 mmag σ ISL = 2 mmag − ∆ J FeH × P r o b a b ili t y σ ISL+ZSL = 1 mmag σ ISL = 1 mmag − − ∆ J FeH × P r o b a b ili t y σ ISL+ZSL = 2 mmag σ ISL = 4 mmag − − ∆ z FeH × P r o b a b ili t y σ ISL+ZSL = 2 mmag σ ISL = 4 mmag
Fig. 7.
Distribution of the metallicity o ff sets for J-PLUS DR2. In all the panels, the initial o ff sets ∆ X FeH from ISL magnitudes are presentedin gray and the final o ff sets ∆ X FeH computed with ISL + ZSL magnitudes in coloured. From top to bottom and left to right, passbands J J J J J J J z are shown. The Gaussian distributions that better describe the data are also presented, with theirdispersion labelled in each panel. Table 3.
Estimated o ff sets to transport the ISL + ZSL photometry outsidethe atmosphere, ∆ X WD , and intrinsic dispersion of the WD locus, σ int .The r band was used as reference in the estimation of the colour o ff sets. Passband ( X ) ∆ X WD σ int [mag] [mag] u − . ± .
004 0 . ± . J − . ± .
004 0 . ± . J − . ± .
004 0 . ± . J − . ± .
004 0 . ± . J − . ± .
003 0 . ± . g . ± .
002 0 . ± . J − . ± .
002 0 . ± . r · · · · · · J − . ± .
003 0 . ± . i − . ± .
002 0 . ± . J − . ± .
004 0 . ± . z − . ± .
003 0 . ± . r band to σ r = ff sets to provide the absolute flux uncertaintyin J-PLUS DR2 (Table 2). The final precision is comparable toDR1 and the new calibration considerably improves the accuracyof our photometry. We compared the final zero points obtained with the stellar andwhite dwarf loci against the zero points obtained by direct com-parison with the PS1 z passband. The di ff erence is well de-scribed by a Gaussian with median µ = − σ = Bayestar17 (Green et al.2018), based on Pan-STARRS stellar colours. We found that thebest consistency with the PS1 z -band photometry is reached withthe estimation based on Schlegel et al. (1998). In all the cases,the systematic o ff sets due to metallicity are present.Interestingly, the application of the metallicity-dependentstellar locus to the X ISL magnitudes worsen the comparison be-tween J-PLUS and PS1 in the
Bayestar17 case, going for σ = σ = σ = σ = ff erences are subtle, but measurable. Wesuggest that the Bayestar17 extinction, derived from the varia-tion of the PS1 stellar locus, is containing part of the colour vari-ation due to metallicity. The extinction maps from Schlegel et al.(1998) and Planck Collaboration et al. (2014) are not related withthe photometry used in the calibration, providing therefore an in-dependent frame for the homogenization process. http://argonaut.skymaps.info/ Article number, page 10 of 14ópez-Sanjuan et al.: J-PLUS. Systematic impact of metallicity on photometric calibration with the stellar locus . . . ( g − i ) − − [ F e / H ] − . − . . . . ( u − r ) − Z S L o J − PLUS 0 . . . ( g − i ) − − [ F e / H ] − . − . . . . ( u − r ) − Z S L o SDSSYuan + Fig. 8.
Modelled ( u − r ) colour di ff erence with respect to the reference locus, ZSL o , as a function of ( g − i ) and [Fe / H]. The median metallicityof the reference locus is marked with the black dashed line.
Left panel : Estimation from J-PLUS DR2 final photometry.
Right panel : Estimationfrom Yuan et al. (2015a) using SDSS photometry.
As already pointed out in Sect. 1, the stellar color regression(SCR; Yuan et al. 2015b; Huang et al. 2020) method deals withthe di ff erent stellar properties in a consistent way, providing analternative homogenization process for the calibration. UsingLAMOST DR5 as reference, the SCR method has been appliedto J-PLUS DR2.We found that the comparison between the ISL + ZSL and theSCR zero points follows a Gaussian distribution with dispersion σ accSCR , as summarised in Table 2. The dispersion is ∼
12 mmag inthe u , J J ∼ J J ∼ ff erent treatment of theinterstellar extinction, our functional approach to the impact ofthe metallicity o ff set, and the inherent statistical dispersion ofeach method.A detailed application and analysis of the SCR calibrationfor J-PLUS DR2 is beyond the scope of the present paper andwill be presented in a forthcoming work. The comparison withthe independent SCR method provided an extra measurement forthe accuracy in the photometry, that we set at percent level forpassbands bluer than λ ∼ In this section, we highlight the impact of the improved cali-bration in the estimation of the photometric metallicity from J-PLUS DR2 data. As in other sections, we use the u band as ex-ample, but similar results are obtained with J-PLUS passbands J J u − r ) versus( g − i ) space as in Sect. 3.2, but using the final J-PLUS DR2 cal-ibrated magnitudes. Following Yuan et al. (2015a), we modelledthe ( u − r ) locus with a fourth degree polynomial in ( g − i ) and[Fe / H]. The resulting model in those bins with data was normal-ized to the expected locus at [Fe / H] = -0.225 dex, as shown inthe left panel of Fig. 8. The curvature in the locus is evident.We compared the J-PLUS ZSL with the results from Yuanet al. (2015a) using SDSS photometry. They provide themetallicity-dependent stellar locus ( u − g ) and ( g − r ) as a func-tion of ( g − i ) and [Fe / H]. We combined both loci to obtain the( u − r ) variation and normalized again to the expected locus at[Fe / H] = − .
225 dex. The result is presented in the right panel of Fig. 8. We find a close agreement between both studies, thatobtain similar structures and general variations for the ZSL. Thediscrepancies, at 0.04 mag level, are expected because of the dif-ferent photometric systems used (J-PLUS versus SDSS).After checking our final ZSL with the results in Yuan et al.(2015a), we aim to test the impact of the calibration in the pho-tometric metallicities estimated from J-PLUS DR2. We decidedto compute the J-PLUS photometric metallicities using the sim-plest o ff set model, relating the ( u − r ) colour distance to thereference locus at [Fe / H] = − .
225 dex with a [Fe / H] measure-ment. We used 144 375 stars in common with LAMOST andwith 0 . ≤ ( g − i ) ≤ . σ = .
14 dex. We stress thatthe goal of this section is just to illustrate the net improvementof the photometric calibration. We expect to get better metallic-ity estimates from the whole twelve-band J-PLUS photometry(e.g. Whitten et al. 2019).Because LAMOST metallicities were used in both the cal-ibration and the estimation of the photometric metallicity, weensured an independent test by comparing J-PLUS metallici-ties with the spectroscopic values from the Apache Point Ob-servatory Galactic Evolution Experiment (APOGEE, Jönssonet al. 2020) latest data release (SDSS DR16 ). The avail-able data contains high-resolution ( R ∼
22 500), near-infrared(15 140 −
16 940 Å) spectra for about 430 000 stars coveringboth the Northern and Southern sky, from which radial veloci-ties, stellar parameters, and chemical abundances of 20 speciesare determined.We cross-matched the MS calibration stars with theAPOGEE sample using 1 arcsec radius. A total flag equal to zeroin APOGEE information and a J-PLUS colour 0 . ≤ ( g − i ) ≤ . ff er-ence between the J-PLUS and APOGEE values was defined as ∆ [Fe / H] = [Fe / H] J − PLUS − [Fe / H] APOGEE . (8)The star-by-star di ff erence defines a Gaussian with median µ = .
03 dex and dispersion σ = .
13 dex.To explore the possible systematic trend of ∆ [Fe / H] withgalactic latitude, we computed the median metallicity di ff erencewith respect to APOGEE using 25 bins of variable size to en-sure ∼
90 sources per bin. The uncertainties where estimated by Article number, page 11 of 14 − −
30 0 30 60 90 b [deg] − . . . ∆ [ F e / H ] ISL − − −
30 0 30 60 90 b [deg] − . . . ∆ [ F e / H ] ISL + ZSL
Fig. 9.
Metallicity di ff erence between J-PLUS photometric values and APOGEE spectroscopic values, ∆ [Fe / H], as a function of galactic latitude b . Left panel : Using X ISL photometry.
Right panel : Using X ISL + ZSL photometry. The solid line in both panels shows the best linear fitting to thedata, with the gray areas depicting the 68% and 95% confidence intervals. The dashed lines marks zero di ff erence. The dotted line in the rightpanel shows a di ff erence of 0.02 dex. − . . . . × ∇ [Fe / H] [deg − ] P r o b a b ili t y Fig. 10.
Distribution in the best linear-fitting slope of the metal-licity di ff erence versus galactic latitude estimated from X ISL (red) and X ISL + ZSL (purple) photometry. The dotted line marks a zero slope. bootstrapping. The results using X ISL magnitudes and the finalcalibration are presented in Fig. 9. We found that the metallic-ity di ff erences are roughly flat with the final J-PLUS DR2 cal-ibration, as desired, presenting a slight bias of 0.02 dex. How-ever, neglecting the ZSL step in the calibration produces a cleartrend with galactic latitude: the estimated ∆ [Fe / H] changes from − .
02 dex at | b | ∼
30 deg to + .
10 dex at | b | ∼
80 deg. Weperformed a linear fit to the data, using | b | as independent vari-able, and present the distribution of the slope ∇ [Fe / H], with[deg − ] units, in Fig. 10. The slope for the final calibration is100 × ∇ [Fe / H] = − . ± .
03, while neglecting the ZSL stepprovides 100 × ∇ [Fe / H] = . ± .
03. The slope is compatiblewith zero, as desired, by including the impact of metallicity inthe stellar locus position, while the slope is positive at 10 σ levelwhen the metallicity e ff ects are not accounted for.We conclude that the improved photometric calibration of J-PLUS DR2 yields a reliable twelve-bands photometric catalogfor an important fraction of the Northern sky.
5. Summary and conclusions
We have explored the impact of metallicity on the photometriccalibration of J-PLUS DR2, based on the stellar locus technique,and update the error budget in the calibration. Using the metallicity information from LAMOST, we findthat the J-PLUS passbands bluer than 4 500 Å are strongly af-fected by the Milky Way metallicity gradient in Galactic latitude,that breaks the assumption of an homogeneous dust de-reddenedstellar locus across the sky. The peak-to-peak variation amounts0 .
07, 0 .
07, 0 .
05, 0 .
03, and 0 .
02 mag in u , J J J J ∼ .
01 mag in J z , while negligible in J J ff ect is sys-tematic and smooth along the surveyed area. We modelled themetallicity-dependent o ff set in the stellar locus in those areas incommon with LAMOST to improve the photometric calibrationin the complete J-PLUS DR2 data set. The accuracy of the cali-bration in the surveyed area is expected to be at percent level forthe bluer J-PLUS passbands and sub-percent in the rest of thefilters after including the metallicity information in the process.The precision in the calibration, measured from repeatedsources in the overlapping areas between pointings and includ-ing absolute colour and flux scale uncertainties, is ∼
18 mmagin u , J J ∼
11 mmag in J J ∼ g , J r , J i , J z . These val-ues are similar to those derived in López-Sanjuan et al. (2019)with J-PLUS DR1 data, reflecting that the metallicity impactsthe calibration at scales larger than a few square degrees.Our analysis highlights the expected impact of metallicity onthe stellar locus technique at λ (cid:46) ff sets at a few degreescale and impacting the physical properties derived for stars andgalaxies. Large-area surveys with blue optical passbands mustevaluate the impact of metallicity in the use of the stellar lo-cus to homogenize their photometry, and techniques based onlarge overlapping areas or methods that accounts for the varietyof stars’ physical properties (e.g., SCR or ISL + ZSL) should befavoured.Regarding the technical goal of J-PLUS, i.e. ensure the pho-tometric calibration of J-PAS, the lessons learnt have been ofgreat importance to define the J-PAS calibration strategy. Thecurrent roadmap for J-PAS calibration has three steps: (1) ho-mogenization using half-CCD overlapping areas thanks to alarge dithering pattern between the four exposures per filter. Thiswill permit to derive a consistent photometric solution alongthe surveyed area by comparing four measurements of the samesource, and to trace 2D variations along the focal plane. (2) Ab-solute colour calibration using the white dwarf locus. The prop-
Article number, page 12 of 14ópez-Sanjuan et al.: J-PLUS. Systematic impact of metallicity on photometric calibration with the stellar locus erties of the locus, with two populations and curved profiles, willpermit the colour calibration without using external photometricdata. (3) Absolute calibration by anchoring the J-PAS referencebroad-band to Pan-STARRS. In this case only one o ff set will beneeded to translate the already homogeneous photometry out-side the atmosphere. The calibration against Gaia is also a pos-sibility, but with J-PAS photometry being independent of
Gaia spectro-photometry it will be possible to test systematic e ff ectsin both surveys. Acknowledgements.
We dedicate this paper to the memory of our six IAC col-leagues and friends who met with a fatal accident in Piedra de los Cochinos,Tenerife, in February 2007, with special thanks to Maurizio Panniello, whoseteachings of python were so important for this paper. We thank the relevantdiscussions and suggestions from the J-PLUS collaboration members. Based onobservations made with the JAST / T80 telescope at the Observatorio Astrofísicode Javalambre (OAJ), in Teruel, owned, managed, and operated by the Centro deEstudios de Física del Cosmos de Aragón. We acknowledge the OAJ Data Pro-cessing and Archiving Unit (UPAD) for reducing the OAJ data used in this work.Funding for the J-PLUS Project has been provided by the Governments of Spainand Aragón through the Fondo de Inversiones de Teruel; the Aragón Governmentthrough the Reseach Groups E96, E103, and E16_17R; the Spanish Ministryof Science, Innovation and Universities (MCIU / AEI / FEDER, UE) with grantsPGC2018-097585-B-C21 and PGC2018-097585-B-C22; the Spanish Ministryof Economy and Competitiveness (MINECO) under AYA2015-66211-C2-1-P, AYA2015-66211-C2-2, AYA2012-30789, and ICTS-2009-14; and EuropeanFEDER funding (FCDD10-4E-867, FCDD13-4E-2685). The Brazilian agenciesFINEP, FAPESP, and the National Observatory of Brazil have also contributedto this project. E. J. A acknowledges financial support from PGC2018-095049-B-C21 (MCIU / AEI / FEDER, UE) and SEV-2017-0709. A. A. C. acknowledgessupport from the Universidad de Alicante (contract UATALENTO18-02). Thework of V. M. P. is supported by NOIRLab, which is managed by the Asso-ciation of Universities for Research in Astronomy (AURA) under a coopera-tive agreement with the National Science Foundation. E. T. acknowledges sup-port by ETAg grant PRG1006 and by EU through the ERDF CoE grant TK133.L. S. J. acknowledges support from Brazilian agencies FAPESP (2019 / / ffi ce,the Max-Planck Society and its participating institutes, the Max Planck Insti-tute for Astronomy, Heidelberg, and the Max Planck Institute for ExtraterrestrialPhysics, Garching, The Johns Hopkins University, Durham University, the Uni-versity of Edinburgh, the Queen’s University Belfast, the Harvard-SmithsonianCenter for Astrophysics, the Las Cumbres Observatory Global Telescope Net-work Incorporated, the National Central University of Taiwan, the Space Tele-scope Science Institute, the National Aeronautics and Space Administration un-der Grant No. NNX08AR22G issued through the Planetary Science Divisionof the NASA Science Mission Directorate, the National Science FoundationGrant No. AST-1238877, the University of Maryland, Eotvos Lorand Univer-sity (ELTE), the Los Alamos National Laboratory, and the Gordon and BettyMoore Foundation. This work has made use of data from the European SpaceAgency (ESA) mission Gaia ( ), pro-cessed by the Gaia
Data Processing and Analysis Consortium (DPAC, ). Funding for theDPAC has been provided by national institutions, in particular the institutionsparticipating in the
Gaia
Multilateral Agreement. Funding for SDSS-III hasbeen provided by the Alfred P. Sloan Foundation, the Participating Institutions,the National Science Foundation, and the U.S. Department of Energy O ffi ceof Science. The SDSS-III web site is . SDSS-IIIis managed by the Astrophysical Research Consortium for the Participating In-stitutions of the SDSS-III Collaboration including the University of Arizona,the Brazilian Participation Group, Brookhaven National Laboratory, CarnegieMellon University, University of Florida, the French Participation Group, theGerman Participation Group, Harvard University, the Instituto de Astrofisica deCanarias, the Michigan State / Notre Dame / JINA Participation Group, Johns Hop-kins University, Lawrence Berkeley National Laboratory, Max Planck Institutefor Astrophysics, Max Planck Institute for Extraterrestrial Physics, New Mex-ico State University, New York University, Ohio State University, PennsylvaniaState University, University of Portsmouth, Princeton University, the SpanishParticipation Group, University of Tokyo, University of Utah, Vanderbilt Uni-versity, University of Virginia, University of Washington, and Yale University.Funding for the Sloan Digital Sky Survey IV has been provided by the Alfred P. Sloan Foundation, the U.S. Department of Energy O ffi ce of Science, andthe Participating Institutions. SDSS-IV acknowledges support and resourcesfrom the Center for High Performance Computing at the University of Utah.The SDSS website is . SDSS-IV is managed by the Astrophys-ical Research Consortium for the Participating Institutions of the SDSS Col-laboration including the Brazilian Participation Group, the Carnegie Institutionfor Science, Carnegie Mellon University, Center for Astrophysics | Harvard &Smithsonian, the Chilean Participation Group, the French Participation Group,Instituto de Astrofísica de Canarias, The Johns Hopkins University, Kavli In-stitute for the Physics and Mathematics of the Universe (IPMU) / Universityof Tokyo, the Korean Participation Group, Lawrence Berkeley National Lab-oratory, Leibniz Institut für Astrophysik Potsdam (AIP), Max-Planck-Institutfür Astronomie (MPIA Heidelberg), Max-Planck-Institut für Astrophysik (MPAGarching), Max-Planck-Institut für Extraterrestrische Physik (MPE), NationalAstronomical Observatories of China, New Mexico State University, New YorkUniversity, University of Notre Dame, Observatário Nacional / MCTI, The OhioState University, Pennsylvania State University, Shanghai Astronomical Obser-vatory, United Kingdom Participation Group, Universidad Nacional Autónomade México, University of Arizona, University of Colorado Boulder, Universityof Oxford, University of Portsmouth, University of Utah, University of Virginia,University of Washington, University of Wisconsin, Vanderbilt University, andYale University. This research made use of
Astropy , a community-developedcore
Python package for Astronomy (Astropy Collaboration et al. 2013), and
Matplotlib , a 2D graphics package used for
Python for publication-qualityimage generation across user interfaces and operating systems (Hunter 2007).
References
Aihara, H., Allende Prieto, C., An, D., et al. 2011, ApJS, 193, 29Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558,A33Benítez, N., Dupke, R., Moles, M., et al. 2014, [ArXiv:1403.5237]Bergeron, P., Dufour, P., Fontaine, G., et al. 2019, ApJ, 876, 67Bertin, E. & Arnouts, S. 1996, A&AS, 117, 393Bonoli, S., Marín-Franch, A., Varela, J., et al. 2020, A&A, submitted[arXiv:2007:01910]Cenarro, A. J., Moles, M., Cristóbal-Hornillos, D., et al. 2019, A&A, 622, A176Cenarro, A. J., Moles, M., Marín-Franch, A., et al. 2014, in Proc. SPIE, Vol.9149, Observatory Operations: Strategies, Processes, and Systems V, 91491IChambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, [arXiv:1612.05560]Covey, K. R., Ivezi´c, Ž., Schlegel, D., et al. 2007, AJ, 134, 2398Cui, X.-Q., Zhao, Y.-H., Chu, Y.-Q., et al. 2012, Research in Astronomy andAstrophysics, 12, 1197Cukanovaite, E., Tremblay, P. E., Freytag, B., Ludwig, H. G., & Bergeron, P.2018, MNRAS, 481, 1522Deng, L.-C., Newberg, H. J., Liu, C., et al. 2012, Research in Astronomy andAstrophysics, 12, 735Flewelling, H. A., Magnier, E. A., Chambers, K. C., et al. 2016,[arXiv:1612.05243]Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2018, A&A, 616, A1Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al. 2016, A&A, 595, A1Gentile Fusillo, N. P., Tremblay, P.-E., Gänsicke, B. T., et al. 2019, MNRAS,482, 4570Green, G. M., Schlafly, E. F., Finkbeiner, D., et al. 2018, MNRAS, 478, 651High, F. W., Stubbs, C. W., Rest, A., Stalder, B., & Challis, P. 2009, AJ, 138,110Huang, Y., Yuan, H., Li, C., et al. 2020, ApJ, in press [arXiv:2011.07172]Hunter, J. D. 2007, Computing In Science & Engineering, 9, 90Ivezi´c, Ž., Sesar, B., Juri´c, M., et al. 2008, ApJ, 684, 287Jiménez-Esteban, F. M., Torres, S., Rebassa-Mansergas, A., et al. 2018, MN-RAS, 480, 4505Jönsson, H., Holtzman, J. A., Prieto, C. A., et al. 2020, AJ, 160, 120Kelly, P. L., von der Linden, A., Applegate, D. E., et al. 2014, MNRAS, 439, 28Kesseli, A. Y., West, A. A., Veyette, M., et al. 2017, ApJS, 230, 16Kuijken, K., Heymans, C., Dvornik, A., et al. 2019, A&A, 625, A2Lenz, D. D., Newberg, J., Rosner, R., Richards, G. T., & Stoughton, C. 1998,ApJS, 119, 121Li, L., Shen, S., Hou, J., et al. 2018, ApJ, 858, 75López-Sanjuan, C., Varela, J., Cristóbal-Hornillos, D., et al. 2019, A&A, 631,A119Luo, A. L., Zhao, Y.-H., Zhao, G., et al. 2015, Research in Astronomy and As-trophysics, 15, 1095Magnier, E. A., Chambers, K. C., Flewelling, H. A., et al. 2016a,[arXiv:1612.05240]Magnier, E. A., Schlafly, E. F., Finkbeiner, D. P., et al. 2016b,[arXiv:1612.05242]
Article number, page 13 of 14 agnier, E. A., Sweeney, W. E., Chambers, K. C., et al. 2016c,[arXiv:1612.05244]Marín-Franch, A., Taylor, K., Cenarro, J., Cristobal-Hornillos, D., & Moles, M.2015, in IAU General Assembly, Vol. 29, 2257381Oke, J. B. & Gunn, J. E. 1983, ApJ, 266, 713Planck Collaboration, Abergel, A., Ade, P. A. R., et al. 2014, A&A, 571, A11Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525Starkenburg, E., Martin, N., Youakim, K., et al. 2017, MNRAS, 471, 2587Tonry, J. L., Stubbs, C. W., Lykke, K. R., et al. 2012, ApJ, 750, 99Tremblay, P. E., Cukanovaite, E., Gentile Fusillo, N. P., Cunningham, T., &Hollands, M. A. 2019, MNRAS, 482, 5222Tremblay, P. E., Ludwig, H. G., Ste ff en, M., & Freytag, B. 2013, A&A, 559,A104Varela, J. & J-PLUS collaboration. 2021, in preparationWaters, C. Z., Magnier, E. A., Price, P. A., et al. 2016, [arXiv:1612.05245]Whitten, D. D., Placco, V. M., Beers, T. C., et al. 2019, A&A, 622, A182Wu, Y., Luo, A. L., Li, H.-N., et al. 2011, Research in Astronomy and Astro-physics, 11, 924Yuan, H., Liu, X., Xiang, M., Huang, Y., & Chen, B. 2015a, ApJ, 799, 134Yuan, H., Liu, X., Xiang, M., et al. 2015b, ApJ, 799, 133Yuan, H. B., Liu, X. W., Huo, Z. Y., et al. 2015c, MNRAS, 448, 855Yuan, H. B., Liu, X. W., & Xiang, M. S. 2013, MNRAS, 430, 2188Zhao, G., Zhao, Y.-H., Chu, Y.-Q., Jing, Y.-P., & Deng, L.-C. 2012, Research inAstronomy and Astrophysics, 12, 723en, M., & Freytag, B. 2013, A&A, 559,A104Varela, J. & J-PLUS collaboration. 2021, in preparationWaters, C. Z., Magnier, E. A., Price, P. A., et al. 2016, [arXiv:1612.05245]Whitten, D. D., Placco, V. M., Beers, T. C., et al. 2019, A&A, 622, A182Wu, Y., Luo, A. L., Li, H.-N., et al. 2011, Research in Astronomy and Astro-physics, 11, 924Yuan, H., Liu, X., Xiang, M., Huang, Y., & Chen, B. 2015a, ApJ, 799, 134Yuan, H., Liu, X., Xiang, M., et al. 2015b, ApJ, 799, 133Yuan, H. B., Liu, X. W., Huo, Z. Y., et al. 2015c, MNRAS, 448, 855Yuan, H. B., Liu, X. W., & Xiang, M. S. 2013, MNRAS, 430, 2188Zhao, G., Zhao, Y.-H., Chu, Y.-Q., Jing, Y.-P., & Deng, L.-C. 2012, Research inAstronomy and Astrophysics, 12, 723