The Survey of Centaurus A's Baryonic Structures (SCABS). I. Survey Description and Initial Source Catalogues
Matthew A. Taylor, Roberto P. Muñoz, Thomas H. Puzia, Steffen Mieske, Paul Eigenthaler, Mia Sauda Bovill
MMNRAS , 1–17 (2016) Preprint 22 February 2018 Compiled using MNRAS L A TEX style file v3.0
The Survey of Centaurus A’s Baryonic Structures(SCABS). I. Survey Description and Initial SourceCatalogues
Matthew A. Taylor, , (cid:63) Roberto P. Mu˜noz, Thomas H. Puzia, Steffen Mieske, Paul Eigenthaler, and Mia Sauda Bovill Institute of Astrophysics, Pontificia Universidad Cat´olica de Chile, Av. Vicu˜na Mackenna 4860, 7820436 Macul, Santiago, Chile European Southern Observatory, Alonso de Cordova 3107, Vitacura, Santiago, Chile Space Telescope Science Institute, 3700 San Martin Drive, 21218, Baltimore, Maryland, USA
Accepted XXX. Received YYY; in original form 22 February 2018
ABSTRACT
We present new, wide-field, optical ( u (cid:48) g (cid:48) r (cid:48) i (cid:48) z (cid:48) ) Dark Energy Camera observations cov-ering ∼
21 deg centred on the nearby giant elliptical galaxy NGC 5128 called “TheSurvey of Centaurus A’s Baryonic Structures” (SCABS). The data reduction and anal-ysis procedures are described including initial source detection, photometric and as-trometric calibration, image stacking, and point-spread function modelling. We esti-mate 50 and 90 percent, field-dependent, point-source completeness limits of at least u (cid:48) = 24 .
08 and 23 .
62 mag (AB), g (cid:48) = 22 .
67 and 22 .
27 mag, r (cid:48) = 22 .
46 and 22 .
00 mag, i (cid:48) = 22 .
05 and 21 .
63 mag, and z (cid:48) = 21 .
71 and 21 .
34 mag. Deeper imaging in the u (cid:48) -, i (cid:48) -and z (cid:48) -bands provide the fainter limits for the inner ∼ of the survey, and we findvery stable photometric sensitivity across the entire field of view. Source cataloguesare released in all filters including spatial, photometric, and morphological informa-tion for a total of ∼ × − . × detected sources (filter-dependent). We finishwith a brief discussion of potential science applications for the data including, but notlimited to, upcoming works by the SCABS team. Key words:
Astronomical Data bases: miscellaneous – catalogues – surveys – galax-ies: individual: NGC 5128
Since the turn of the century, the Sloan Digital Sky Sur-vey (SDSS; York et al. 2000), with its thousands of deg ofsky coverage, has demonstrated the utility of modern CCD-based large-scale imaging campaigns. As a result, a new gen-eration of wide-field imaging cameras has enabled relativelysmall research groups to conduct intermediate-scale surveyscapable of deeply imaging areas of sky ranging from dozensto 100s of deg that were previously accessible to much largerconsortia of researchers. These imagers are particularly use-ful to conduct deep studies of nearby galaxy groups andclusters, which are rapidly revealing the never before acces-sible faint properties of these systems (e.g. Chiboucas et al. (cid:63) Visiting astronomer, Cerro Tololo Inter-American Observa-tory, National Optical Astronomy Observatory, which is oper-ated by the Association of Universities for Research in Astron-omy (AURA) under a cooperative agreement with the NationalScience Foundation. E-mail: [email protected] (MAT) . ± . ∼
11 deg around NGC 5128 in the optical g (cid:48) - and r (cid:48) -bands using the optical Megacam imager at the 6.5-mMagellan II Clay telescope (McLeod et al. 2015). This sur-vey has recently revealed nine new low-surface brightness(25 . (cid:46) µ r, (cid:46) . − (cid:46) M V (cid:46) − . ∼
150 kpc of NGC 5128. A yet more am-bitious program uses the
Dark Energy Camera (DECam;Flaugher et al. 2015) to cover ∼
550 deg encompassingNGC 5128, and the nearby M 83 complex (M¨uller et al. 2015, c (cid:13) a r X i v : . [ a s t r o - ph . GA ] A ug M. A. Taylor et al. µ r, ≈
29 mag arcsec − . While these candidatesawait confirmation via spectroscopy and/or resolved stellarpopulation studies, the utility of wide-field imagers like DE-Cam in revealing the secrets of these iconic neighbours isclear.In this contribution we present a new imaging campaignof ∼
72 deg centred on NGC 5128 using DECam at the 4-mBaade telescope at the Cerro Tololo Interamerican Observa-tory (CTIO) in Chile. Our program, The Survey of Centau-rus A’s Baryonic Structures (SCABS) is complementary tothe other two recent large-scale projects. Among the mostpowerful features of the PISCeS campaign is the ability toconduct resolved stellar population studies, combined withdepth that samples the red giant branch at the distance ofNGC 5128. On the other hand, while the MJB15/16 pro-gram reaches similar depths as PISCeS, they do not benefitfrom their resolution capabilities, but probe an area ∼ × greater. Noting the strengths and shortcomings of these twosurveys, SCABS addresses complementary science goals byhomogeneously imaging NGC 5128 to ∼ × the galactocen-tric radius (out to R gc ≈
300 kpc) compared to PISCeS,which has so-far been focussed mainly on the NE quadrantof the galaxy. More importantly, while we only sample afraction of the area covered by MJB15/16, we do so in thefive optical u (cid:48) , g (cid:48) , r (cid:48) , i (cid:48) , and z (cid:48) filters, providing significantspectral energy distribution (SED) leverage and opening thedoor to a suite of ancillary science goals not possible with g (cid:48) - and r (cid:48) -band imaging alone.The paper is organized as follows. § § ∼
21 deg region of SCABS upon whichthis work is based. We also describe artificial star experi-ments which were used to asses the photometric quality anddepth of our observations. § m − M = 27 . ± .
05 mag, corresponding to a distance of3 . ± . The u (cid:48) , g (cid:48) , r (cid:48) , i (cid:48) , and z (cid:48) -band SCABS data were taken dur-ing the nights of 2014 April 4–5 (CNTAC ID: 2014A-0610;PI: Matthew Taylor), and 2014 August 25–27 (CNTAC ID:2014B-0609; PI: Roberto Mu˜noz) using DECam mounted atthe prime focus on the 4 m Victor Blanco telescope at theCerro Tololo Inter-American Observatory (CTIO) in Chile.DECam is a wide field imager equipped with the optical u’, g’, r’, i’, z’ and Y filters. Working at a focal ratio of f/ .
7, it is comprised by 62 2 048 × ∼
53 and40 (cid:48)(cid:48) ) gaps along the long and short edges respectively, giv-ing a spatial coverage of ∼ . . One chip (N30) showspoor charge transfer efficiency rendering it scientifically use-less, and since 30 November 2013 chip S30 stopped working. Additionally the Southern half of S7 malfunctions, and thusDECam is effectively comprised of 59.5 CCDs with a total of ∼
500 Mpix at pixel scales between 0.2626 (cid:48)(cid:48) /pix at the edgeof the field-of-view, to 0.2637 (cid:48)(cid:48) /pix at the centre. Luckily,the positions of the bad chips make it easy to recover thefull spatial extent of DECam in image mosaics that employan appropriate dithering pattern. For more details on theDECam technical specifications we refer the reader to theonline documentation .SCABS uses the large field-of-view of DECam to imageNGC 5128 out to its approximate virial radius of ∼
300 kpc,shown in Fig. 1 as the red-dashed ellipse. A five-point dither-ing strategy per pointing covers the DECam chip gaps, andresults in the flower-like mosaic of DECam footprints shownby the blue shading in Fig. 1. To account for the missingN30 and S30 chips, each pointing overlaps the Northern-and Southern-most CCD rows, and is shifted either E orW by a single chip width, which gives rise to the slightlydenser zig-zag regions shown by the darker N–S blue bandson Fig. 1. The first night of the 2014A-0610 program fo-cussed on the i (cid:48) , z (cid:48) , and u (cid:48) filters, split into 5 ×
20 = 100 s,5 ×
40 = 200 s, and 5 ×
240 = 1 200 s exposures, respec-tively. The second night was used to finish the u (cid:48) -band ex-posures (Tiles 13–23; see Fig. 1) and to conduct the short5 ×
12 = 60 s and 5 ×
20 = 100 s r (cid:48) - and g (cid:48) -band SCABSimaging. These exposure times were selected in order toreach targeted apparent point-source magnitudes of m u (cid:48) (cid:39) . m g (cid:48) (cid:39) . m r (cid:48) (cid:39) . m i (cid:48) (cid:39) . m z (cid:48) (cid:39) . σ past the GCLFturnover magnitude, assuming a dispersion of σ g (cid:48) = 1 . § g (cid:48) - and r (cid:48) -band observations, as well as a sin-gle dither position of the u (cid:48) -band for Tiles 13–23 were notconducted until the 2014B-0609 run. These data are stillundergoing reduction before combination with the previousdata can be made and thus are not considered further here.Table 1 gives a summary of the data collected, reduced, andanalyzed in this work. The observing conditions varied overthe course of the two nights with light cirrus giving wayto clear and stable conditions during the first night, witha similar trend during the first half of the second followedby deteriorating conditions in the latter half. Seeing as re-ported by the CTIO dimm averaged 0 . (cid:48)(cid:48) with a dispersionof 0 . (cid:48)(cid:48) during the first night, while the second night hadpoorer conditions with mean dimm seeing of 1 . (cid:48)(cid:48) . The firsthalf of the second night was also less stable, with a disper-sion of 0 . (cid:48)(cid:48) , largely due to a short-term spike in the seeingthat particularly affected the photometric calibration of the r (cid:48) -band images (see § MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures α [J2000] -49.52-48.24-46.96-45.69-44.41-43.13-41.85-40.57-39.30-38.02 δ [ J ]
12 3 456789 10 11 12 13141516171819202122 23 24
NGC 5128Confirmed GCsKnown Dwarfs ω Centauri
Figure 1.
The spatial coverage of the SCABS observations. The position of NGC 5128 is shown by the green star, while the surroundingcloud of black dots indicates the population of radial velocity confirmed GCs. Orange triangles denote the positions of the previouslyknown dwarf galaxy population within the SCABS footprint, and the position of the galactic GC ω -Centauri is indicated by the marooncircle, which serendipitously falls within the overall SCABS footprint. Different tiles are indicated by the numbers shown, with Tile 1centred on NGC 5128 itself. This section describes the multi-stage processing appliedto the raw data taken at the Blanco telescope, which isschematically represented in Fig. 2. Briefly, the process in-cludes initial pipeline-based calibrations, from which we takelow-level calibrated products and further process them usinga custom reduction cascade that derives the final astrometricand photometric solutions, image stacking, and the produc-tion of final photometric source catalogues.
Preliminary reduction steps for all DECam images are per-formed by the CTIO DECam Community Pipeline (CDCP;Valdes et al. 2014, v.3.1.1) and are shown in more detail by the left-hand flowchart in Fig. 2 (adapted from the DE-Cam User’s Guide v.2.0.5). While in principle the CDCPcan provide fully calibrated, sky subtracted, and stackedimages, the photometric calibration can be inaccurate byan unknown amount, possibly reaching as high as 0.5 mag,and thus are not appropriate for our science goals (see the
NOAO Data Handbook for details). For this reason we startwith InstCal products from the CDCP, and perform furtherreduction/calibration steps using custom routines describedbelow. The
InstCal images have basic pre-processing stepsapplied to them, correcting for electronic bias, crosstalk be-tween DECam chip amplifiers, and fringing effects. Addi-tionally, the
InstCal frames have been flat-fielded before http://ast.noao.edu/sites/default/files/NOAO_DHB_v2.2.pdf MNRAS000
InstCal frames have been flat-fielded before http://ast.noao.edu/sites/default/files/NOAO_DHB_v2.2.pdf MNRAS000 , 1–17 (2016)
M. A. Taylor et al.
Table 1.
SCABS observational log. Cols. (1) and (2) list the Tile ID (see Fig. 1) and date of observation, followed by the right ascensionand declination of the central dither pointings in Cols. (3) and (4), respectively. The remaining five columns list total exposure times for,in order, the u (cid:48) , g (cid:48) , r (cid:48) , i (cid:48) , and z (cid:48) filters.Tile Date α δ u (cid:48) g (cid:48) r (cid:48) i (cid:48) z (cid:48) ( hh : mm : ss ) ( ◦ : (cid:48) : (cid:48)(cid:48) ) (s) (s) (s) (s) (s)1 2014 Apr. 4 13:25:27.62 -43:01:08.81 2 400 ... ... 300 4002014 Apr. 5 ... 100 60 ... ...2 2014 Apr. 4 13:32:40.12 -41:52:05.25 1 200 ... ... 100 2002014 Apr. 5 ... 100 60 ... ...3 2014 Apr. 4 13:23:46.73 -41:12:37.50 1 200 ... ... 100 2002014 Apr. 5 ... 100 60 ... ...4 2014 Apr. 4 13:16:30.83 -42:21:41.06 1 200 ... ... 100 2002014 Apr. 5 ... 100 60 ... ...5 2014 Apr. 4 13:18:13.56 -44:10:12.36 1 200 ... ... 100 2002014 Apr. 5 ... 100 60 ... ...6 2014 Apr. 4 13:27:12.74 -44:49:40.11 1 200 ... ... 100 2002014 Apr. 5 ... 100 60 ... ...7 2014 Apr. 4 13:34:21.17 -43:40:36.55 1 200 ... ... 100 2002014 Apr. 5 ... 100 60 ... ... Input Output
Raw, Master DQMCross-talk CoefficientsBiasesPupil TemplateFringe TemplateDomeFlat, SkyFlatAstrometric CatalogPhotometric Catalog
Artifact FlaggingCross-talk CorrectionBias CorrectionPupil Ghost CorrectionFringe CorrectionFlat-field CorrectionWCS CalibrationPhotometric CalibrationRe-projectStack Resampled, Resample MaskInstCal,InstCal MaskStack, Mask, ExpMap
Input Output
LSE_44 InstCal,image InstCal,wtmap
Photometric Calibration(estimate zp)
InstCal,image InstCal,wtmap
Source Detection(SExtractor)Astrometric Solution(SCAMP)Re-projection and Image Stacking(SWARP)Aperture Photometry(SExtractor)Point-spread function modelling(PSFEx)PSF Photometry(SExtractor)Stack, Weight Map Stack, Weight MapPSF modelsPSF models Source CataloguePoint source vignettesPoint source vignettes Initial Source CatalogueInitial Source Catalogue
CDCP SCABS
Figure 2.
The SCABS data reduction cascade. The preliminary CDCP reduction steps are shown on the left (adapted from the DECamUser’s Guide v.2.0.5) where the
InstCal products are taken as a starting point for our custom reduction steps depicted by the right-handflowchart, and described in more detail in the text. rudimentary astrometric and photometric calibrations areapplied. The end result are frames that are clean of all ma-jor cosmetic defects, along with the data quality and back-ground weight maps which are used for further data reduc-tion steps and subsequent analysis as described in the fol-lowing.
Main image processing was carried out by a custom
IDL -and
Python -based data reduction pipeline , which itselfcalls standard image processing packages from the Astro-matic software (Bertin & Arnouts 1996; Bertin et al. 2002; https://github.com/rpmunoz/DECam/tree/master/data_reduction MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures Bertin 2006, 2011). The reduction cascade, as shown bythe right-hand schematic in Fig. 2, consists of initial sourcedetections made on the individual frames provided by theCDCP, which are then used in the subsequent astromet-ric and photometric calibrations. Using the astrometric so-lutions, frames are aligned and combined for each tile–filter combination to produce stacked images suitable forpoint-source photometry. The photometric measurementsare made by performing a second round of source detec-tions on the image stacks, from which bright, unsaturatedpoint sources are identified for point spread function (PSF)modelling. Final photometric catalogues are generated byintegrating over the resulting PSF models. These data-reduction cascade steps are listed in detail in the followingsub-sections.
The first step in constructing the final image stacks is a“first-pass” source detection. These sources are used for derivingthe astrometric and photometric calibrations described be-low, and are detected in the individual CDCP
InstCal frameson a chip-by-chip basis using
SExtractor ( SE ; v. 2.19.5).For the purpose of calibration, relatively bright, well definedsources are preferred, and so we only took sources detectedat a relatively conservative 1.8 σ above the background ( de-tect thresh=1.8 ). To maximize the accuracy of the astro-metric solution described below, all CDCP-processed framesare considered at this stage, resulting in 455 individualsource catalogues. The software package
Scamp (v. 2.0.4) was used to derivethe relative astrometric solution across the SCABS field.
Scamp reads in the output source catalogues provided by SE in the previous step, and matches them to sourcesfrom the 2MASS astrometric reference catalogue (Skrut-skie et al. 2006). For this procedure, we set a reference starsearch radius of 1 . (cid:48)(cid:48) ( position maxerr=1.2 ) and only usematches with signal-to-noise ratios (S/N) between 40 and 80( sn threshholds =40,80), which typically results in severalhundred reference star matches per CCD in a given filter.We apply lanczos2 resampling to the images, and usea fourth degree polynomial to calculate the pixel distor-tion across the DECam field-of-view ( distort degrees=4 ).Fig. 3 shows the pixel scale distortion map resulting from theuse of all 455 source catalogues in deriving the astrometricsolution. The pixel scale smoothly changes in a radially sym-metric way, with distortions of, at most, ∼ . (cid:48)(cid:48) ( ∼ . (cid:48)(cid:48) , and with typical σ (cid:46) . (cid:48)(cid:48) , indicating sub-arcsecond accuracy of our astro-metric calibration. During the nights we observed u (cid:48) g (cid:48) r (cid:48) i (cid:48) z (cid:48) standard star field LSE 44 centred at ( α, δ ) = (13 h : 52 m : 49 s, − ◦ : 09 (cid:48) : 09 (cid:48)(cid:48) ) Instrument A1: distortion map -01 °00 '-01 °00 '-00 °40 '-00 °40 '-00 °20 '-00 °20 '-00 °00 '-00 °00 '+00 °20 '+00 °20 '+00 °40 '+00 °40 '+01 °00 '+01 °00 ' -01 °00 '-01 °00 ' -00 °40 '-00 °40 ' -00 °20 '-00 °20 ' -00 °00 '-00 °00 ' +00 °20 '+00 °20 ' +00 °40 '+00 °40 ' +01 °00 '+01 °00 ' -1 ) a r c s ec pixel scale -0.2-0.10.00.10.2 % Figure 3.
The DECam pixel scale distortion map from the astro-metric calibration. The field of view of DECam is illustrated withpixel-scale distortion shown by the colour map in units of (cid:48)(cid:48) /pixel,and as an absolute percentage of the mean of 0.263 (cid:48)(cid:48) /pixel. Redrepresents positive pixel-scale distortions, while blue shows dis-tortions smaller than the mean. As a whole, the distortions arevery symmetric in nature, with a peak-to-valley difference nearthe 0.5 percent level. (Smith et al. 2007) with varying exposure time and airmasscombinations to facilitate photometric zero point estimates.For the estimates we use, zp = m std − m inst + kX (1)where zp is the magnitude zero point, k is the airmassterm, X is the airmass, m inst is the instrumental magni-tude from the standard star frames based on the SE el-liptical aperture measurements (i.e. mag auto ), and m std is the standard star catalogue AB magnitudes measured inthe u (cid:48) g (cid:48) r (cid:48) i (cid:48) z (cid:48) system. We estimate k via linear regressionin the ( m std − m inst ) – X plane, and then use Eq. 1 to ex-trapolate the zero point magnitude at X = 1 .
0. We com-pare to zero points and airmass terms based on photometricstandard star observations from the period 1–19 November2012 . Table 2 lists the values derived from our standardstar photometry and the CTIO values which are the aver-ages from all DECam CCDs. The listed CTIO errors arethose from the CTIO tables, added in quadrature to onestandard deviation of the measurements from the 62 CCDs.In general we find very good agreement between the two setsof calibration data, with the exception of the r (cid:48) -band, whichdiffers by ∆ r (cid:48) (cid:39) .
45 mag. The image quality during theearly part of the second night varied slightly between stan-dard star field exposures, which resulted in a larger scatterin the m r (cid:48) versus X relation and a correspondingly moreuncertain calibration. For this reason we adopt our derivedzero points and airmass terms for the four u (cid:48) , g (cid:48) , i (cid:48) , and z (cid:48) frames, and defer to the older CTIO calibration for the r (cid:48) -band, for which we incorporate errors from both calibrationsto reflect the larger calibration uncertainty (see § http://goo.gl/8h0xuW MNRAS000
45 mag. The image quality during theearly part of the second night varied slightly between stan-dard star field exposures, which resulted in a larger scatterin the m r (cid:48) versus X relation and a correspondingly moreuncertain calibration. For this reason we adopt our derivedzero points and airmass terms for the four u (cid:48) , g (cid:48) , i (cid:48) , and z (cid:48) frames, and defer to the older CTIO calibration for the r (cid:48) -band, for which we incorporate errors from both calibrationsto reflect the larger calibration uncertainty (see § http://goo.gl/8h0xuW MNRAS000 , 1–17 (2016)
M. A. Taylor et al.
Table 2.
Photometric calibration information. Col. (1) lists thefilters, while photometric zero points derived directly from ourstandard star field observations are shown in Col. (2), and Col. (3)lists those provided by CTIO. Similarly, Cols. (4) and (5) list thecorresponding airmass terms. The errors on zp CTIO are adoptedas the standard deviation between individual chip zero pointsadded in quadrature to the zp error listed in the CTIO calibrationdata.Filter zp SCABS zp CTIO k SCABS k CTIO (mag) (mag) u (cid:48) ± .
06 23.62 ± . − ± . − ± . g (cid:48) ± .
02 25.42 ± . − ± . − ± . r (cid:48) ± .
04 25.47 ± . − ± . − ± . i (cid:48) ± .
01 25.34 ± . − ± . − ± . z (cid:48) ± .
01 25.06 ± . − ± . − ± . N ( u g ) N ( u r ) N ( u i ) Separation (arcsec) N ( u z ) Figure 4.
Astrometric alignment accuracy. The distributions insource coordinate offsets for recovered sources in Tile 1 are shownwith respect to their u (cid:48) -band astrometry and demonstrate thatour image alignments are accurate to within ∼ . (cid:48)(cid:48) . Verticalblack dashed lines indicate the DECam pixel scale. The astrometric solution derived on all the individual framesis used to construct the final image stacks using the
Swarp software (v. 2.38). During the initial source detection algo-rithm, we use the i (cid:48) -band images as our reference frames,and align the others to these images in pixel-space. We findour image alignment to be excellent, as matching recoveredsources in each filter in Tile 1 to the u (cid:48) -band source catalogueresults in Fig. 4, which shows the distribution in source coor-dinate offsets. In each panel the offset distributions peak at ∼ . (cid:48)(cid:48) , with virtually no sources showing offsets of (cid:38) . (cid:48)(cid:48) .As noted in §
2, we failed to finish a complete ditherpattern for the u (cid:48) -band observations, and could not obtaindata of sufficient quality in the g (cid:48) and r (cid:48) filters for Tiles 13-23of the SCABS footprint. For this reason they were omittedfrom the data reduction cascade described above and thus no image stacks are available to be analyzed. With this unfor-tunate fact in mind, and not wanting to delay the release ofour current science-ready source catalogues, we restrict thisrelease to sources in Tiles 1–7, roughly corresponding to thespatially homogeneous halo within ∼
140 kpc of NGC 5128.While Tiles 8–12, and 24 are in principle ready, we will re-lease those data in a future release along with the results ofthe analysis of Tiles 13–24.
For all of the final stacked frames, photometric measure-ments on detected sources are performed using a combi-nation of SE and PSFEx (v. 3.16.1). Sources are first de-tected with an initial pass of SE , and vignette images arecreated upon which the PSF is modelled by PSFEx . Be-fore the PSF modelling, point-like sources are identified onthe images by identifying the stellar locus in mag auto – flux radius space, and bright but non-saturated sourcesare selected that bracket the mean flux radius of the lo-cus. We model the PSF on the corresponding vignettes with PSFEx using a 47 ×
47 pix kernel ( psf size=47,47 ) withvariations followed to third order ( psfvar degrees=3 ). Asummary of the modelling is shown in Table 3 that lists,for each tile–filter combination, the number of PSF modelsconstructed, as well as the modelled PSF FWHM in arcsec,which varies between 1 . (cid:48)(cid:48) and 1 . (cid:48)(cid:48) , with the best qualityobservations corresponding to the r (cid:48) -band, which shows amean FWHM of 1 . (cid:48)(cid:48) .The resulting PSF models are finally used in another SE run, which using them as input, measures the final,PSF-corrected magnitudes for the sources. During this fi-nal source extraction, the detection criterion was suchthat flux from at least six pixels adjacent to a source( detect minarea=6 ) fell at least 1.5 σ above the back-ground levels to be analyzed ( detect thresh=1.5 , anal-ysis thresh=1.5 ). Table 4 summarizes the final source cat-alogues produced by SE for each stacked frame, with totalsthroughout the area covered by Tiles 1–7 listed below. Weadopt statistical photometric errors as those reported by PS-FEx , and add the uncertainties from our adopted zero pointand airmass term calibrations (see § σ bootstrapped uncertainties based onour artificial star experiments ( § Galactic Extinction andReddening Calculator using the Galactic reddening mapsof Schlafly & Finkbeiner (2011) on a source-by-source ba-sis and include these for convenience in the source cata-logues described in § u (cid:48) -band, to take into account differential reddening acrossthe SCABS field-of-view. From the heat-map, which indi-cates the magnitude of reddening towards a given directionin Tiles 1–7, extinction can be as high as A u (cid:48) = 0 . u (cid:48) -band, with peak-to-valley differences across the region http://ned.ipac.caltech.edu/forms/calculator.html MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures Table 3.
Point-spread function modelling summary. Col. (1) indicates the tile over which the PSF modelling was conducted, followedby the number of PSF stars selected for the u (cid:48) -band in Col. (2), and for the g (cid:48) -, r (cid:48) -, i (cid:48) -, and z (cid:48) -bands in each second column thereafter.Similarly, Col. (3) lists the PSF FWHM measured in the u (cid:48) -band for each tile, with every second column following lists the sameinformation for the remaining filters. Finally, the bottom row lists the mean numbers of PSF stars, and the corresponding mean FWHMvalues. Tile N u (cid:48) , psf u (cid:48) fwhm N g (cid:48) , psf g (cid:48) fwhm N r (cid:48) , psf r (cid:48) fwhm N i (cid:48) , psf i (cid:48) fwhm N z (cid:48) , psf z (cid:48) fwhm ( (cid:48)(cid:48) ) ( (cid:48)(cid:48) ) ( (cid:48)(cid:48) ) ( (cid:48)(cid:48) ) ( (cid:48)(cid:48) )1 6 007 1.43 8 129 1.68 10 806 1.27 11 671 1.13 13 783 1.282 6 736 1.43 7 132 1.74 11 223 1.25 8 154 1.43 9 971 1.343 6 997 1.41 9 101 1.68 10 141 1.25 11 109 1.35 10 533 1.284 7 293 1.42 8 134 1.66 10 899 1.26 11 557 1.22 10 742 1.225 8 282 1.41 9 049 1.69 11 201 1.25 12 027 1.25 9 887 1.236 8 613 1.40 9 812 1.85 11 846 1.24 10 284 1.27 10 192 1.287 8 668 1.37 11 638 1.74 11 896 1.25 11 458 1.41 11 625 1.38Mean 7514 1.41 8999 1.72 11145 1.25 10894 1.29 10962 1.29 Table 4.
Source detection summary. The seven tiles are listed inCol. (1), while Cols. 2–6 show the total number of sources withPSF-corrected photometric measurements in each of the u (cid:48) -, g (cid:48) -, r (cid:48) -, i (cid:48) -, and z (cid:48) -band stacked images.Tile N u (cid:48) N g (cid:48) N r (cid:48) N i (cid:48) N z (cid:48) of as much as ∆ A u (cid:48) = 0 .
59 mag, with effects diminishing to-wards the redder filters to a maximum A z (cid:48) = 0 .
27 mag and∆ A z (cid:48) = 0 .
18 mag for the z (cid:48) -band. Artificial star experiments were conducted to quantify thepoint-source depths of the SCABS observations. In these ex-periments, artificial point-sources are added to images in arange of magnitudes, and the same source detection algo-rithm used on the science images are applied to the experi-mental images. The results are used to derive the magnitudelimit at which we fail to recover the mock sources.We use a set of
IDL - and
Python -based scripts to addstars by slicing images into 200 ×
200 pix regions, and gen-erating up to 10 random positions within each to add themock stars. To avoid artificial crowding, the positions aresuch that no real source lies within 16 pixels of the mockstar, which results in ∼
60 000 artificial star positions perimage. Magnitudes are then randomly assigned to the ar-tificial sources to create the final mock star catalogues. Toobtain sufficient statistics, the process of assigning randommagnitudes is repeated 10 times per image, so that the re-sults are in reality based on ∼
600 000 artificial sources perpointing.We simulate the SCABS observing conditions by query-ing our PSF model libraries (see § Figure 5.
The u (cid:48) -band foreground extinction ( A u (cid:48) ) map forTiles 1–7 of SCABS. The amount of extinction across the fieldof view differs by as much as 0 . u (cid:48) -band across thefield of view, and illustrates the importance of correcting for dif-ferential foreground reddening. luminosity to add appropriate “stellar” sources at the cor-responding positions to the real image. Another run of SE using the same parameters as for the real science frames iscarried out on the experimental images, and mock outputcatalogues are generated. These catalogues are compared tothe input mock catalogues to derive the completeness lim-its of the SCABS data. To account for the longer exposuretimes in the u (cid:48) , i (cid:48) , and z (cid:48) filters for Tile 1, this process iscarried out separately for all five filters in Tiles 1 and 2.Given the constant exposure times, FWHM dispersions notexceeding ∼ MNRAS000
The u (cid:48) -band foreground extinction ( A u (cid:48) ) map forTiles 1–7 of SCABS. The amount of extinction across the fieldof view differs by as much as 0 . u (cid:48) -band across thefield of view, and illustrates the importance of correcting for dif-ferential foreground reddening. luminosity to add appropriate “stellar” sources at the cor-responding positions to the real image. Another run of SE using the same parameters as for the real science frames iscarried out on the experimental images, and mock outputcatalogues are generated. These catalogues are compared tothe input mock catalogues to derive the completeness lim-its of the SCABS data. To account for the longer exposuretimes in the u (cid:48) , i (cid:48) , and z (cid:48) filters for Tile 1, this process iscarried out separately for all five filters in Tiles 1 and 2.Given the constant exposure times, FWHM dispersions notexceeding ∼ MNRAS000 , 1–17 (2016)
M. A. Taylor et al.
Figure 6.
Results of our artificial star experiments expressedas u (cid:48) -band point-source completeness estimates. ( Upper sub-panels: ) The difference between the input mock stellar cataloguemagnitudes and the output SE measurements as a function ofinput magnitude are shown in blue, while the red curve indicatesa non-parametric linear regression to the data, along with 1 σ bootstrapped errors in 0.2 mag bins, clipped to within 2.3 σ . Theblack dashed line indicates perfect agreement between input andoutput magnitude. ( Lower sub-panels: ) The fractions of sourcesrecovered from the input catalogues in 0.2 mag bins. The dashedblue line shows the data, while the solid red line shows a splineinterpolation used to predict the 50% and 90% completeness lim-its. The blue shading encloses sources with 100% representation(dark blue), 90% representation (blue), and 50% representation(light blue). Adopted 50% and 90% completeness limits are shown(see also Tbl. ?? . Finally, the top panel shows results correspond-ing to the central tile (Tile 1) of SCABS, while the bottom panelshows results representative of the outer ring (Tiles 2–7). and the good agreement between the Tile 1 and 2 g (cid:48) - and r (cid:48) -band sensitivities (see Table 5), the results of the artificialstar experiments for Tile 2 are adopted for Tile 3–7.The results of the artificial star experiments are listedin Table 5, and shown in Figs. 6–15. The lower panels ofFigs. 6–10 show, for each of the five filters in Tiles 1 and2, the fraction of recovered mock sources as a function of in-put magnitude. The blue shading bounded by dashed blacklines indicates areas corresponding to 100, 90, and 50 percentcompleteness, with darker shading indicating higher com- Figure 7.
Results of the artificial star experiments expressed as g (cid:48) -band point-source completeness estimates. See Fig. 6 for de-tailed descriptions of the panels. pleteness. The solid red relations indicate spline fits to theblue dashed curves, which represent the results of the ex-periments. In all cases, the analytic red curves represent thedata well, and are used to infer the numerical completenesslimits as labelled, and listed in Table 5. The upper panelsof Figs. 6–10 illustrate the robustness of the recovered pho-tometry of the mock stars. The difference between the inputPSF-based and recovered mag auto magnitudes is shown asa function of input brightness, with a non-parametric linearregression and associated 1 σ bootstrapped errors shown bythe red curves. These panels suggest that our point-sourcephotometry is reliable within (cid:46) . SE ’s mag auto not accounting for the tails of the PSFmodels at faint magnitudes, and do not exceed ∼ . Using the
Python/AstroML package.MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures Figure 8.
Results of the artificial star experiments expressed as r (cid:48) -band point-source completeness estimates. See Fig. 6 for de-tailed descriptions of the panels. which in any case, we conservatively include in our listedsystematic error budgets.Figs. 11–15 show the photometric stability across thefield of view of DECam. The left-hand panels show resultsfor Tile 1, with Tiles 2–7 shown on the right, while the up-per and lower rows correspond to the 50 and 90 percentcompleteness limits, respectively. In image pixel space, thecompleteness test images are sliced into 2D bins, and 50 and90 percent completeness limits are calculated for individualbins. The resulting grid of limiting magnitudes is then usedto infer values across the entire image using lanczos inter-polation, and is indicated by the colour maps with sensitivityincreasing from blue to red.Small patches that sharply transition to shallower (orNaN) limits persist across several of the panels of Figs. 11–15. These artifacts are a result of the stochastic process ofadding mock stars in regions that contain extended back-ground sources, combined with the random sampling ofmagnitudes assigned to the mock stellar catalogues. Test-ing larger and/or smaller bin sizes results in these patchesshifting position among the various panels, and thus we donot consider them to be physical in nature. With this inmind, the median limiting magnitudes across the field ofview are shown in the upper left corner of each panel, which Figure 9.
Results of the artificial star experiments expressed as i (cid:48) -band point-source completeness estimates. See Fig. 6 for de-tailed descriptions of the panels. agree well with the values based on Figs. 6–10 and listed inTable 5.The artificial star experiments indicate that for theTiles 2–7, we reach 90 percent completeness magnitudes of23.61, 22.26, 22.02, 21.63, and 21.33 mag in the u (cid:48) , g (cid:48) , r (cid:48) , i (cid:48) , and z (cid:48) filters, respectively, with correspondingly fainter50 percent completeness limits of 24.08, 22.67, 22.46, 22.00,and 21.71 mag. Meanwhile, the deeper Tile 1 u (cid:48) -, i (cid:48) -, and z (cid:48) -band imaging is reflected by the ∼ . ∼ .
8, and ∼ . (cid:46) . ∼ DECam foot-print, and generally tend towards variations of (cid:46) . r (cid:48) -band results, and a (cid:46) . g (cid:48) -band, which is likely due to seeing variations during theobservations and is in any case, less than the photometricvariation across the DECam field of view. MNRAS000
8, and ∼ . (cid:46) . ∼ DECam foot-print, and generally tend towards variations of (cid:46) . r (cid:48) -band results, and a (cid:46) . g (cid:48) -band, which is likely due to seeing variations during theobservations and is in any case, less than the photometricvariation across the DECam field of view. MNRAS000 , 1–17 (2016) M. A. Taylor et al.
Figure 10.
Results of the artificial star experiments expressedas z (cid:48) -band point-source completeness estimates. See Fig. 6 for de-tailed descriptions of the panels. The source catalogues are publicly available by querying http://vizier.u-strasbg.fr/viz-bin/VizieR . In accordswith Table 4, we provide photometric measurements for asfew as 595 136 sources in the u (cid:48) -band and up to 1 490 519sources with i (cid:48) -band photometry. Table 6 summarizes thecontents of the available measurements, alongside the corre-sponding catalogue parameter names (based on SE wherepossible). A total of nine photometric measurements areprovided with associated SE error estimates, correspondingto fluxes within seven fixed-circular apertures, an ellipticalaperture, and PSF-based magnitudes. Following the pho-tometric measurements we list the adopted statistical andsystematic error budgets (see § SE half-light ra-dius, source ellipticity, isophotal area measured within fiveisocontours, and a compactness parameter. SE measures iso-contours at seven surface brightness levels, the areas withineach are given by the iso parameters in units of pix .Here we list five corresponding to the two faintest isocon-tours ( iso ,
1) increasing to the smallest isophotal area con-taining the brightest pixels of the source ( iso
Table 5.
Point source completeness estimates. The Table is sep-arated into two subtables indicated by the subtitles. The firstfive rows list completeness limits for the central tile, where u (cid:48) , i (cid:48) ,and z (cid:48) photometry is deeper than the rest of the survey, whilethe bottom five rows list values that are assumed for Tiles 2–7(i.e. ”the Outer Ring”). Col. (1) indicates the filter under consid-eration, while Cols. (2) and (4) show the 50 and 90 percent com-pleteness limits, respectively. Cols. (3) and (5) show one standarddeviation from the mean of the 50 and 90 percent completenessmagnitudes across the DECam field of view, which is an indica-tion of the spatially varying photometric sensitivity of SCABS.Filter m σ m σ (mag) (mag) (mag) (mag)Tile 1 u (cid:48) g (cid:48) r (cid:48) i (cid:48) z (cid:48) u (cid:48) g (cid:48) r (cid:48) i (cid:48) z (cid:48) band is listed based on the reddening maps of Schlafly &Finkbeiner (2011). The value of these data covers myriad potential lines ofinvestigation. For example, the 90 percent completenessdepths of these data are sufficient to detect (cid:38)
95 percentof globular clusters (GCs) in the region, assuming a GC lu-minosity function peak at m V ≈ − . ∼ . ∼
140 kpc from NGC 5128, the suite of colour indicesare sufficient to distinguish most GCs from both foregroundstellar sources, and point-like background sources that mor-phologically masquerade as GCs. Moreover, the extent ofSCABS reaches into the extreme halo of NGC 5128 and thusshould begin to probe GCs associated with the intra-groupmedium of Centaurus A, including any dwarfs in the re-gion. Altogether, the potential for these data to identify thevast majority of GCs associated with NGC 5128 representsa powerful method of constraining its mass assembly historythrough cosmic time.The serendipitous location of NGC 5128 at relativelylow Galactic latitude ( b = 19 . ◦ ) makes this dataset valu-able for more than strictly extragalactic investigation. Thewide-field coverage of DECam, with pointings mildly off ofthe Galactic plane makes the removal of foreground starschallenging for background science, but the sampling of thefull optical SED eases this task and thus can provide a richlist of foreground star photometry. Indeed, the ∼
21 deg of coverage samples foreground stars through the thin andthick Galactic disks, and out into the halo toward NGC 5128, MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures Y Pixel Value X P i x e l V a l u e µ u , = 24 . σ u , = 0 . NE u -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ u , = 24 . σ u , = 0 . NE u -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ u , = 23 . σ u , = 0 . NE u -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ u , = 23 . σ u , = 0 . NE u -band Completeness Magnitude Figure 11.
Photometric depth variations in the u (cid:48) -band based on artificial star experiments. Results for the central tile (Tile 1) ofSCABS are shown in the left column, and results representative of the Outer Ring (Tiles 2–7) are shown on the right. Variations in thecompleteness magnitudes are parameterized by the colour bar, with the top row corresponding to the 50 percent completeness magnitude,and the bottom showing results for the 90 percent depth. Units along the axes are in DECam imaging pixels (see also Fig. ?? ), and theNorth and East cardinal directions are indicated in the lower-left corners of the panels. 1 σ variations are listed in the upper-left corners,below the mean photometric depths based on the variation maps upper-left corners, which are in good agreement with the adopted valuesshown in Tbl. 5 and Figs. 6–10. Spurious drops in the depth variations represent regions that were under-sampled during the binning,and not considered to be physical.MNRAS000
Photometric depth variations in the u (cid:48) -band based on artificial star experiments. Results for the central tile (Tile 1) ofSCABS are shown in the left column, and results representative of the Outer Ring (Tiles 2–7) are shown on the right. Variations in thecompleteness magnitudes are parameterized by the colour bar, with the top row corresponding to the 50 percent completeness magnitude,and the bottom showing results for the 90 percent depth. Units along the axes are in DECam imaging pixels (see also Fig. ?? ), and theNorth and East cardinal directions are indicated in the lower-left corners of the panels. 1 σ variations are listed in the upper-left corners,below the mean photometric depths based on the variation maps upper-left corners, which are in good agreement with the adopted valuesshown in Tbl. 5 and Figs. 6–10. Spurious drops in the depth variations represent regions that were under-sampled during the binning,and not considered to be physical.MNRAS000 , 1–17 (2016) M. A. Taylor et al.
Y Pixel Value X P i x e l V a l u e µ g , = 22 . σ g , = 0 . NE g -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ g , = 22 . σ g , = 0 . NE g -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ g , = 22 . σ g , = 0 . NE g -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ g , = 22 . σ g , = 0 . NE g -band Completeness Magnitude Figure 12.
Photometric depth variations in the g (cid:48) -band based on artificial star experiments. See Fig. 11 for a detailed description of theFigure. potentially providing a rich catalogue of foreground starswith ten permutations of optical colour indices. By com-parison to state-of-the-art stellar atmospheric models, richancillary science including studies of the ages and metallic-ities of the various populations of Galactic foreground starsare possible.The doubling of the Local Group dwarf galaxy popula- tion in the past decade (e.g. Willman et al. 2005; Belokurovet al. 2006, 2007; McConnachie et al. 2009; McConnachie2012) provides a potent window on the epoch of the firstgalaxies via near-field cosmology studies coupled with simu-lations (e.g. Salvadori & Ferrara 2009; Bovill & Ricotti 2009,2011a,b); however, little is known about dwarf galaxy pop-ulations beyond the Local Group. To this end, rich popula- MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures Y Pixel Value X P i x e l V a l u e µ r , = 22 . σ r , = 0 . NE r -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ r , = 22 . σ r , = 0 . NE r -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ r , = 21 . σ r , = 0 . NE r -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ r , = 21 . σ r , = 0 . NE r -band Completeness Magnitude Figure 13.
Photometric depth variations in the r (cid:48) -band based on artificial star experiments. See Fig. 11 for a detailed description of theFigure. tions of dwarf galaxies are being discovered in nearby galaxyclusters like Virgo and Fornax (Mu˜noz et al. 2015; S´anchez-Janssen et al. 2016) and NGC 5128 itself (Karachentsev et al.2007; Crnojevi´c et al. 2014, 2015; M¨uller et al. 2015, 2016)which provide direct observational tests of the bottom-uphierarchical formation of their hosts as favoured by Λ ColdDark Matter cosmology (e.g. Klypin et al. 1999; Moore et al. 1999). These observations, in conjunction with the iden-tification of the group’s extended GC system, will then beof excellent utility in placing new constraints on the assem-bly history of this iconic galaxy and its group environment,which is the natural extension of the ongoing detailed LocalVolume studies, such as SDSS, PanSTARRS, PANDAS, and MNRAS000
Photometric depth variations in the r (cid:48) -band based on artificial star experiments. See Fig. 11 for a detailed description of theFigure. tions of dwarf galaxies are being discovered in nearby galaxyclusters like Virgo and Fornax (Mu˜noz et al. 2015; S´anchez-Janssen et al. 2016) and NGC 5128 itself (Karachentsev et al.2007; Crnojevi´c et al. 2014, 2015; M¨uller et al. 2015, 2016)which provide direct observational tests of the bottom-uphierarchical formation of their hosts as favoured by Λ ColdDark Matter cosmology (e.g. Klypin et al. 1999; Moore et al. 1999). These observations, in conjunction with the iden-tification of the group’s extended GC system, will then beof excellent utility in placing new constraints on the assem-bly history of this iconic galaxy and its group environment,which is the natural extension of the ongoing detailed LocalVolume studies, such as SDSS, PanSTARRS, PANDAS, and MNRAS000 , 1–17 (2016) M. A. Taylor et al.
Y Pixel Value X P i x e l V a l u e µ i , = 22 . σ i , = 0 . NE i -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ i , = 22 . σ i , = 0 . NE i -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ i , = 22 . σ i , = 0 . NE i -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ i , = 21 . σ i , = 0 . NE i -band Completeness Magnitude Figure 14.
Photometric depth variations in the i (cid:48) -band based on artificial star experiments. See Fig. 11 for a detailed description of theFigure. the Dark Energy Survey (e.g. York et al. 2000; Abbott et al.2006; McConnachie et al. 2009; Kaiser et al. 2010).While the depth is not sufficient to probe the high-redshift ( z (cid:38)
1) universe, the optical luminosity function(LF) of intermediate redshift galaxies is sensitive to thestar formation/morphological properties of the underlyinggalaxy population. Meanwhile, similarly deep near-infrared (NIR) observations have already been conducted and willextend the SED coverage yet further. Together with the op-tical data discussed here these data will probe the mass func-tion (MF) of background cluster galaxies (e.g. Madau et al.1998). Since the evolution of the MF is directly predictedfrom hierarchical galaxy formation models that incorporatetheoretical SEDs (e.g. galform
Cole et al. 2000), the full
MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures Y Pixel Value X P i x e l V a l u e µ z , = 22 . σ z , = 0 . NE z -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ z , = 21 . σ z , = 0 . NE z -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ z , = 21 . σ z , = 0 . NE z -band Completeness Magnitude Y Pixel Value X P i x e l V a l u e µ z , = 21 . σ z , = 0 . NE z -band Completeness Magnitude Figure 15.
Photometric depth variations in the z (cid:48) -band based on artificial star experiments. See Fig. 11 for a detailed description of theFigure. NUV-NIR LFs represent an excellent test for model predic-tions such as these. Altogether the depth of the SCABS dataprovide an excellent opportunity to derive galaxy parame-ters for a rich sample of galaxies out to z ≈ This paper presents new optical ( u (cid:48) g (cid:48) r (cid:48) i (cid:48) z (cid:48) ) observations ofthe central ∼
21 deg ( ∼
62 000 kpc ) region of the Cen-taurus A galaxy group centred on NGC 5182, as part ofthe Survey of Centaurus A’s Baryonic Structures (SCABS) .The observations have a raw data reduction conducted by
MNRAS000
MNRAS000 , 1–17 (2016) M. A. Taylor et al.
Table 6.
Point source catalogue information with all magnitudeslisted in AB.Description Unit
Parameter
Description Unit
Parameter
Right Ascension deg. alpha j2000
Declination deg. delta j2000
Fixed Aperture Magnitudes (7) mag. mag aper
Errors (7) mag. magerr aper
Elliptical Aperture Magnitude mag. mag auto
Error mag. magerr auto
PSF Magnitude mag. mag psf
Error mag. magerr psf
Statistical Error (see § stat err Systematic Error (see § sys err Effective Radius pixel flux radius
Image FWHM pixel fwhm image
World FWHM deg. fwhm world
Major Axis pixel a image
Minor Axis pixel b image
Ellipticity 1 − ba ellipticity Position Angle deg. theta image
Isophotal Area 0 pixel iso0 Isophotal Area 1 pixel iso1 Isophotal Area 3 pixel iso3 Isophotal Area 5 pixel iso5 Isophotal Area 7 pixel iso7 Spread Parameter spread model
Spread Error spreaderr model
Phototmetry Flag
FLAGS
Foreground Reddening mag. reddening the
CTIO-DECam Community Pipeline (Valdes et al. 2014,v.3.1.1), from which we derive photometric and astrometriccalibrations using our custom built post-processing pipelinebased on the
Astromatic software suite (Bertin & Arnouts1996; Bertin et al. 2002; Bertin 2006, 2011). Individualframes are aligned to a common world coordinate solution,and co-added to produce images which are sufficient forthe analysis of point-like, or mildly extended sources. Ar-tificial star experiments are conducted to derive 50 and 90percent point-source completeness estimates, finding 90 per-cent completeness magnitudes of at least 23 .
62, 22 .
27, 22 . .
63, and 21 .
34 AB mag in the u (cid:48) -, g (cid:48) -, r (cid:48) -, i (cid:48) -, and z (cid:48) -bands,respectively, with very stable photometric sensitivity acrossthe field.We release our source catalogues for public use, whichcan be used to, as non-exhaustive examples, probe the com-pact stellar systems (UCDs and GCs) of the Centaurus Agalaxy group in the context of their stellar population pa-rameters and NGC 5128’s mass assembly history, study thebackground universe out to z ≈ .
0, and probe the proper-ties of the distinct populations of Galactic foreground starsin the direction of NGC 5128. We look forward to releas-ing future source catalogues including the central ∼
72 deg around Centaurus A, as well as finishing a detailed back-ground subtraction that will yield multi-band photometryof NGC 5128’s rich dwarf galaxy population, both old andnew. Furthermore, deep NIR imaging is already in-hand,which will be added to the current imaging to provide fullNUV-NIR SEDs for nearly one million sources in the fieldand provide myriad science results in the upcoming years. ACKNOWLEDGEMENTS
We wish to thank Simon ´Angel, Yasna Ordenes-Brice˜no,Mirko Simunovic, and Hongxin Zhang for fruitful discus-sion, and especially Eric Peng for additionally providingus with catalogues of new confirmed foreground stars andGCs prior to publication. M.A.T. acknowledges the finan-cial support through an excellence grant from the “Vicer-rector´ıa de Investigaci´on” and the Institute of AstrophysicsGraduate School Fund at Pontificia Universidad Cat´olicade Chile and the European Southern Observatory GraduateStudent Fellowship program. T.H.P. acknowledges supportby a FONDECYT Regular Project Grants (No. 1121005 andNo. 1161817) and the BASAL Center for Astrophysics andAssociated Technologies (PFB-06). M.S.B. was supported inpart by FONDECYT Project Grant (No. 3130549).This research has made use of the NASA AstrophysicsData System Bibliographic Services, the NASA Extragalac-tic Database, and the SIMBAD database, operated at CDS,Strasbourg, France (Wenger et al. 2000). Software usedin the analysis includes the
Python/NumPy v.1.11.0 and
Python/Scipy v0.17.0 (Jones et al. 2001; Van Der Waltet al. 2011, ), Python/astropy (v1.1.1; Astropy Collaboration et al. 2013, ), Python/matplotlib (v1.5.1; Hunter2007, http://matplotlib.org/ ), Python/scikit-learn (v0.16.1; Pedregosa et al. 2012, http://scikit-learn.org/stable/ ), and
Python/astroML (v0.3; VanderPlas et al.2012, ) packages.This work is based on observations at Cerro Tololo Inter-American Observatory, National Optical Astronomy Obser-vatory (CNTAC Prop. ID: 2014A-0610; PI: Matthew Tay-lor), which is operated by the Association of Universities forResearch in Astronomy (AURA) under a cooperative agree-ment with the National Science Foundation. This projectused data obtained with the Dark Energy Camera (DE-Cam), which was constructed by the Dark Energy Sur-vey (DES) collaboration. Funding for the DES Projectshas been provided by the U.S. Department of Energy, theU.S. National Science Foundation, the Ministry of Scienceand Education of Spain, the Science and Technology Fa-cilities Council of the United Kingdom, the Higher Edu-cation Funding Council for England, the National Centerfor Supercomputing Applications at the University of Illi-nois at Urbana-Champaign, the Kavli Institute of Cosmo-logical Physics at the University of Chicago, Center for Cos-mology and Astro-Particle Physics at the Ohio State Uni-versity, the Mitchell Institute for Fundamental Physics andAstronomy at Texas A&M University, Financiadora de Es-tudos e Projetos, Funda¸c˜ao Carlos Chagas Filho de Am-paro, Financiadora de Estudos e Projetos, Funda¸c˜ao Car-los Chagas Filho de Amparo ´a Pesquisa do Estado do Riode Janeiro, Conselho Nacional de Desenvolvimento Cient´ı-fico e Tecnol´ogico and the Minist´erio da Ciˆencia, Tecnolo-gia e Inova¸c˜ao, the Deutsche Forschungsgemeinschaft andthe Collaborating Institutions in the Dark Energy Survey.The Collaborating Institutions are Argonne National Lab-oratory, the University of California at Santa Cruz, theUniversity of Cambridge, Centro de Investigaciones En´er-geticas, Medioambientales y Tecnol´ogicas–Madrid, the Uni-versity of Chicago, University College London, the DES-Brazil Consortium, the University of Edinburgh, the Ei-dgen˜ossische Technische Hochschule (ETH) Z˜urich, FermiNational Accelerator Laboratory, the University of Illi-nois at Urbana-Champaign, the Institut de Ci`encies del’Espai (IEEC/CSIC), the Institut de F´sica d’Altes Ener-gies, Lawrence Berkeley National Laboratory, the Ludwig-Maximilians Universit¨at M˜unchen and the associated Ex-
MNRAS , 1–17 (2016) he Survey of Centaurus A’s Baryonic Structures cellence Cluster Universe, the University of Michigan, theNational Optical Astronomy Observatory, the University ofNottingham, the Ohio State University, the University ofPennsylvania, the University of Portsmouth, SLAC NationalAccelerator Laboratory, Stanford University, the Universityof Sussex, and Texas A&M University. REFERENCES
Abbott, T. M. C., Catarutti, R., Mondaca, E., et al. 2006, inSociety of Photo-Optical Instrumentation Engineers (SPIE)Conference Series, Vol. 6267, Society of Photo-Optical Instru-mentation Engineers (SPIE) Conference SeriesAstropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al.2013, A&A, 558, A33Belokurov, V., Zucker, D. B., Evans, N. W., et al. 2006, ApJ, 642,L137Belokurov, V., Zucker, D. B., Evans, N. W., et al. 2007, ApJ, 654,897Bertin, E., & Arnouts, S. 1996, A&AS, 117, 393Bertin, E., Mellier, Y., Radovich, M., et al. 2002, AstronomicalData Analysis Software and Systems XI, 281, 228Bertin, E. 2006, Astronomical Data Analysis Software and Sys-tems XV, 351, 112Bertin, E. 2011, Astronomical Data Analysis Software and Sys-tems XX, 442, 435Bovill, M. S., & Ricotti, M. 2009, ApJ, 693, 1859Bovill, M. S., & Ricotti, M. 2011, ApJ, 741, 17Bovill, M. S., & Ricotti, M. 2011, ApJ, 741, 18Chiboucas, K., Karachentsev, I. D., & Tully, R. B. 2009, AJ, 137,3009Cole, S., Lacey, C. G., Baugh, C. M., & Frenk, C. S. 2000, MN-RAS, 319, 168Cˆot´e, S., Freeman, K. C., Carignan, C., & Quinn, P. J. 1997, AJ,114, 1313Crnojeci´c, D., Sand, D. J., Caldwell, N., et al. 2014, ApJ, 795, 35Crnojevi´c, D., Sand, D. J., Spekkens, K., et al. 2015,arXiv:1512.05366Ferrarese, L., Cˆot´e, P., Cuillandre, J.-C., et al. 2012, ApJ, 200, 4Flaugher, B., Diehl, H. T., Honscheid, K., et al. 2015, AJ, 150,150Harris, W. E. 2001, Saas-Fee Advanced Course 28: Star Clusters,223Harris, G. L. H. 2010, PASA, 27, 457Hunter, J. D. 2007, CISE, 9, 90Jones E., Oliphant E., Peterson P., et al. 2001-, SciPy: OpenSource Scientific Tools for Python, [Online; accessed 2016-05-10]Kaiser, N., Burgett, W., Chambers, K., et al. 2010, in Society ofPhoto-Optical Instrumentation Engineers (SPIE) ConferenceSeries, Vol. 7733, Society of Photo-Optical InstrumentationEngineers (SPIE) Conference SeriesKarachentsev, I. D., Tully, B. R., Dolphin, A., et al. 2007, AJ,133, 504Klypin, A., Kravtsov, A. V., Valenzuela, O., & Prada, F. 1999,ApJ, 522, 82Madau, P., Pozzetti, L., & Dickinson, M. 1998, ApJ, 498, 106McConnachie, A. W., Irwin, M. J., Ibata, R. A., et al. 2009, Na-ture, 461, 66McConnachie, A. W. 2012, AJ, 144, 4McLeod, B., Geary, J., Conroy, M., et al. 2015, PASP, 127, 366Merritt, A., van Dokkum, P., & Abraham, R. 2014, ApJ, 787, L37Moore, B., Ghigna, S., Governato, F., et al. 1999, ApJ, 524, L19M¨uller, O., Jerjen, H., & Binggeli, B. 2015, A&A, 583, A79M¨uller, O., Jerjen, H., & Binggeli, B. 2016, arXiv:1605.04130Mu˜noz, R. P., Puzia, T. H., Lan¸con, A., et al. 2014, ApJS, 210, 4 Mu˜noz, R. P., Eigenthaler, P., Puzia, T. H., et al. 2015, ApJ, 813,L15Pedregosa, F., Varoquaux, G., Gramfort, A., et al. 2012, JMLR,12, 2825Salvadori, S., & Ferrara, A. 2009, MNRAS, 395, L6S´anchez-Janssen, R., Ferrarese, L., MacArthur, L. A., et al. 2016,ApJ, 820, 69Sand, D. J., Crnojevi´c, D., Strader, J., et al. 2014, ApJ, 793, L7Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103Skrutskie, M. F., Cutri, R. M., Stiening, R., et al. 2006, AJ, 131,1163Smith, J. A., Allam, S. S, Tucker, D. L, et al. 2007, AJ, submitted
Tully, B.R., Libeskind, N.I., Karachentsev, I.D., et al. 2015, ApJ,802, 25Valdes, F., Gruendl, R., & DES Project 2014, Astronomical DataAnalysis Software and Systems XXIII, 485, 379van den Bergh, S. 2000, AJ, 119, 609Van Der Walt, S., Colbert, S. C., & Varoquaux, G. 2011, Comp.in Sci. and Eng., 13, 22VanderPlas, J., Connolly, A. J., Ivezic, Z., & Gray, A. 2012,Proceedings of Conference on Intelligent Data Understand-ing (CIDU), pp. 47-54, 2012., 47Villegas, D., Jord´an, A., Peng, E. W., et al. 2010, ApJ, 717, 603Wenger, M., Ochsenbein, F., Egret, D., et al. 2000, A&AS, 143,9Willman, B., Blanton, M. R., West, A. A., et al. 2005, AJ, 129,2692York, D. G., Adelman, J., Anderson, J. E., Jr., et al. 2000, AJ,120, 1579This paper has been typeset from a TEX/L A TEX file prepared bythe author.MNRAS000