The SAURON project - XIII. SAURON-GALEX study of early-type galaxies: the ultraviolet colour-magnitude relations and Fundamental Planes
Hyunjin Jeong, Sukyoung K. Yi, Martin Bureau, Roger L. Davies, Jesus Falcon-Barroso, Glenn van de Ven, Reynier F. Peletier, Roland Bacon, Michele Cappellari, Tim de Zeeuw, Eric Emsellem, Davor Krajnovic, Harald Kuntschner, Richard M. McDermid, Marc Sarzi, Remco C. E. van den Bosch
aa r X i v : . [ a s t r o - ph . GA ] J un Mon. Not. R. Astron. Soc. , 000–000 (0000) Printed 13 June 2018 (MN L A TEX style file v2.2)
The SAURON pro ject - XIII. SAURON-GALEXstudy of early-type galaxies: the ultravioletcolour-magnitude relations and Fundamental Planes
Hyunjin Jeong, Sukyoung K. Yi, ⋆ Martin Bureau, Roger L. Davies, Jes´us Falc´on-Barroso, , Glenn van de Ven, † Reynier F. Peletier, Roland Bacon, Michele Cappellari, Tim de Zeeuw, , Eric Emsellem, Davor Krajnovi´c, Harald Kuntschner, Richard M. McDermid, Marc Sarzi, and Remco C. E. van den Bosch Department of Astronomy, Yonsei University, Seoul 120-749, Korea Sub-Department of Astrophysics, University of Oxford, Denys Wilkinson Building, Keble Road, Oxford OX1 3RH, United Kingdom European Space Agency / ESTEC, Keplerlaan 1, 2200 AG Noordwijk, The Netherlands Instituto de Astrof´ısica de Canarias, E-38200 La Laguna, Tenerife, Spain Institute for Advanced Study, Einstein Drive, Princeton, NJ 08540, U.S.A. Kapteyn Astronomical Institute, University of Groningen, P.O. Box 800, 9700 AV Groningen, The Netherlands Universit´e de Lyon 1, CRAL, Observatoire de Lyon, 9 av. Charles Andr´e, F-69230 Saint-Genis Laval; CNRS, UMR 5574;ENS de Lyon, France European Souther Observatory, Karl-Schwarzchild-Str. 2, 85748, Garching, Germany Leiden Observatory, Leiden University, Niels Bohrweg 2, 2333 CA Leiden, The Netherlands Space Telescope European Coordinating Facility, European Southern Observatory, Karl-Schwarzchild-Str. 2, 85748, Garching, Germany Gemini Observatory, 670 North A’Ohoku Place, Hilo, Hawaii 96720, U.S.A. Centre for Astrophysics Research, University of Hertfordshire, Hatfield, United Kingdom Department of Astronomy, University of Texas, Austin, TX 78712, U.S.A.
13 June 2018
ABSTRACT
We present
Galaxy Evolution Explorer ( GALEX ) far (FUV) and near (NUV)ultraviolet imaging of 34 nearby early-type galaxies from the
SAURON repre-sentative sample of 48 E/S0 galaxies, all of which have ground-based opticalimaging from the MDM Observatory. The surface brightness profiles of ninegalaxies ( ≈
26 per cent) show regions with blue UV − optical colours sug- c (cid:13) H. Jeong et al. gesting recent star formation. Five of these ( ≈
15 per cent) show blue inte-grated UV − optical colours that set them aside in the NUV integrated colour-magnitude relation. These are objects with either exceptionally intense andlocalised NUV fluxes or blue UV − optical colours throughout. They also haveother properties confirming they have had recent star formation, in particularH β absorption higher than expected for a quiescent population and a higherCO detection rate. This suggests that residual star formation is more com-mon in early-type galaxies than we are used to believe. NUV-blue galaxies aregenerally drawn from the lower stellar velocity dispersion ( σ e <
200 km s − )and thus lower dynamical mass part of the sample. We have also constructedthe first UV Fundamental Planes and show that NUV blue galaxies bias theslopes and increase the scatters. If they are eliminated the fits get closer toexpectations from the virial theorem. Although our analysis is based on a lim-ited sample, it seems that a dominant fraction of the tilt and scatter of the UVFundamental Planes is due to the presence of young stars in preferentially low-mass early-type galaxies. Interestingly, the UV − optical radial colour profilesreveal a variety of behaviours, with many galaxies showing signs of recent starformation, a central UV-upturn phenomenon, smooth but large-scale age andmetallicity gradients, and in many cases a combination of these. In addition,FUV − NUV and FUV − V colours even bluer than those normally associatedwith UV-upturn galaxies are observed at the centre of some quiescent galax-ies. Four out of the five UV-upturn galaxies are slow rotators. These objectsshould thus pose interesting challenges to stellar evolutionary models of theUV-upturn. Key words: galaxies: elliptical and lenticular, cD – galaxies: evolution –galaxies: fundamental parameters – galaxies: photometry – galaxies: structure– ultraviolet: galaxies.
The dominant formation mechanism of early-type galaxies remains one of the long-standingdebates of modern astrophysics. The classical monolithic model (e.g. Eggen, Lynden-Bell & Sandage1962; Larson 1974) suggests that early-type galaxies form in highly efficient starbursts athigh redshifts and evolve without much residual star formation from that point onward. On ⋆ E-mail: [email protected] † Hubble Fellow c (cid:13) , 000–000 he SAURON project - XIII
SAURON sample (e.g., de Zeeuw et al. 2002, hereafter Paper II; Krajnovi´c et al.2008, hereafter Paper XII).More recently, the
Galaxy Evolution Explorer ( GALEX ) satellite opened up the ultra-violet (UV) window, allowing to probe the recent star formation history of galaxies withmuch greater accuracy than was previously possible with only optical information. The rest-frame UV is one order of magnitude more sensitive than the optical to the presence ofhot stars, easily revealing populations younger than 1 Gyr. Yi et al. (2005) used
GALEX
Medium Imaging Survey data to construct the first near ultraviolet (NUV) CMR of early-type galaxies classified by the Sloan Digital Sky Survey (SDSS). They found a remarkablylarge scatter towards blue colours and interpreted it as evidence of recent star formation.Kaviraj et al. (2007) and Schawinski et al. (2007) subsequently found that the fraction ofearly-type galaxies that experienced recent star formation in the last billion years can begreater than 30 per cent. Young, Bendo & Lucero (2009) further concluded that the starformation rates of nearby early-type galaxies derived from mid-infrared imaging are in goodagreement with the UV results. There is thus clear evidence for residual star formation inthe local early-type galaxy population and for its influence on global scaling relations, atleast colour-magnitude relations.The Fundamental Plane (FP) is another key scaling relation of early-type galaxies, atwo-dimensional plane in the three-dimensional manifold of their global structural param-eters (effective radius, surface brightness and stellar velocity dispersion; e.g. Dressler et al. c (cid:13) , 000–000 H. Jeong et al. tilt of the FP away from the virial prediction. Numerous studieshave investigated this, and the observed tilt has been attributed to a variation of the mass-to-light ratio across the sequence of early-type galaxies and/or to the breaking of the homologyassumption (see, e.g., Dressler et al. 1987; Djorgovski & Davis 1987; Guzm´an et al. 1993;Jørgensen et al. 1996; Scodeggio 1997; Pahre et al. 1998; Kelson et al. 2000; Bernardi et al.2003; Treu et al. 2006; Cappellari et al. 2006, hereafter Paper IV; Bolton et al. 2008). Nev-ertheless, much debate still exists on the origin of the tilt and the scatter about the meanrelation (e.g. D’Onofrio et al. 2006 and references therein).Spatially-resolved galaxy data provide valuable information on the details of the stellarpopulation distributions, especially when combined with UV data. Hence, we aim here toobtain a UV database matching the
SAURON integral-field data, and present spatially-resolvedUV and optical imaging of most early-type galaxies in the
SAURON survey. Numerous inves-tigations can be performed with such a database, but we will focus here on revisiting theeffects of star formation on the scaling relations of early-types, particularly colour-magnituderelations and the Fundamental Plane. Longstanding enigmas such as the UV-upturn phe-nomenon (see O’Connell 1999 and references therein) are closely related.In this paper, we thus present and discuss our UV imaging of
SAURON early-type galaxiesobtained with
GALEX , along with ground-based optical imaging from the MDM Observa-tory. In §
2, we present a brief summary of the
SAURON survey and our optical observations,and then describe at length the
GALEX observations and surface photometry. In §
3, wediscuss the UV colour-magnitude relations, the first UV Fundamental Planes, and make anattempt at interpreting radial UV − optical colour profiles. We summarize our findings in § The
SAURON observations are aimed at determining the two-dimensional stellar kinematics(Emsellem et al. 2004, hereafter Paper III), stellar linestrengths (Kuntschner et al. 2006,hereafter Paper VI) and ionised gas kinematics (Sarzi et al. 2006, hereafter Paper V) of 48nearby early-type galaxies and 24 spiral bulges in the field and clusters (see Paper II). The c (cid:13)000
SAURON observations are aimed at determining the two-dimensional stellar kinematics(Emsellem et al. 2004, hereafter Paper III), stellar linestrengths (Kuntschner et al. 2006,hereafter Paper VI) and ionised gas kinematics (Sarzi et al. 2006, hereafter Paper V) of 48nearby early-type galaxies and 24 spiral bulges in the field and clusters (see Paper II). The c (cid:13)000 , 000–000 he SAURON project - XIII Ground-based optical imaging observations in the
Hubble Space Telescope ( HST ) filter F555W(similar to Johnson V ) were obtained using the MDM Observatory 1.3-m McGraw-Hill Tele-scope, as part of a large survey targeting the whole SAURON galaxy sample. The MDM ob-servations are described in detail in Falc´on-Barroso et al. (in prep.) and were reduced andcalibrated in the standard manner. The field-of-view of the MDM images is 17 . ′ × . ′ . ′′ × . ′′
508 pixels, allowing for accurate sky subtraction and proper sampling of the seeing.The seeing during the observations was typically ≈ . V magnitudes weuse in this paper are also from Falc´on-Barroso et al. (in prep.). We obtained both FUV (1350–1750 ˚A) and NUV (1750–2750 ˚A) images of 34 E/S0 galaxiesusing the medium-depth imaging mode of
GALEX , as part of an ongoing UV imaging surveyof the
SAURON sample (
GALEX guest investigator programmes GI1–109 and GI3–041) andthe
GALEX
Nearby Galaxy Survey (NGS; Gil de Paz et al. 2007). Of the 48 early-typegalaxies in the
SAURON sample, 6 are too close to UV-bright stars to be observed with
GALEX and 3 from our programmes have not been observed yet, similarly for 5 galaxiesfrom other guest investigator programmes, explaining the current sample of 34. The typicalexposure time per field was one orbit ( ≈ GALEX instruments, pipeline and calibration are described in Martin et al. (2005) andMorrissey et al. (2005, 2007). We note in particular a possible systematic error in the FUVand NUV zero-points of up to 0 .
15 mag. This is not included in the uncertainties quotedin the current paper as we are mostly interested in the relative colours of our objects. Thespatial resolutions of the images are approximately 4 . ′′ . ′′ . ′′ × . ′′ c (cid:13) , 000–000 H. Jeong et al.
Figure 1.
GALEX
UV and MDM optical images and surface photometry of 34 early-type galaxies in the
SAURON sample. (a)–(c)
FUV, NUV and F555W images. (d) V , NUV and FUV radial surface brightness profiles. (e)–(g) FUV − NUV, FUV − V and NUV − V radial colour profiles. (h) Ellipticity radial profile derived from the optical image. The two horizontal lines showFUV − NUV = 1 . − V = 5 .
0, respectively. The vertical lines show the I -band effective radius from the SAURON survey(Paper VI). c (cid:13)000
0, respectively. The vertical lines show the I -band effective radius from the SAURON survey(Paper VI). c (cid:13)000 , 000–000 he SAURON project - XIII Figure 1. continued.c (cid:13) , 000–000
H. Jeong et al.
Figure 1. continued. c (cid:13) , 000–000 he SAURON project - XIII Figure 1. continued.c (cid:13) , 000–000 H. Jeong et al.
Figure 1. continued. c (cid:13) , 000–000 he SAURON project - XIII Figure 1. continued.c (cid:13) , 000–000 H. Jeong et al.
Figure 1. continued.
The
GALEX
UV images are delivered pre-processed but we undertook our own estimateof the sky values. This was calculated for each image as the mean of the sky values in 90–100 image regions after masking-out Sextractor-detected sources. Typical mean sky valueswere 0 . − in FUV and 4 counts pixel − in NUV. We computed the mean skyvalue uncertainty in each image from the distribution of sky values in the small regionsassuming a Gaussian distritubion. These were typically 0 .
06 counts pixel − in FUV and0 . − in NUV. We took the low level flat fielding errors to be negligible, so thesemeasurements form the basis of the errors plotted in the individual FUV and NUV radialsurface brightness profiles of Figure 1. At large radii, the galaxies are faint and only a fewcounts are detected in each pixel. Accurately determining the level of the sky, even if faint,is thus the main factor limiting the depth of the images and our surface photometry. TheUV radial profiles in Figure 1 are truncated when the uncertainty in the surface brightnessexceeds 0 . − . The V band profiles are truncated at the same radial distancewhere their corresponding NUV profiles end. c (cid:13) , 000–000 he SAURON project - XIII Table 1.
Integrated properties.Galaxy E ( B − V ) D σ e R e h µ i e Total apparent magnitude UV typeFUV NUV FUV NUV FUV NUV(Mpc) (km s − ) (arcsec) (mag) (mag)NGC0474 0.034 26.1 150 74 67 28.96 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± R e from Paper IX. The uncertainty is taken as 5 per cent. (5)–(6)Effective radius in each band. The uncertainties are taken as 20 per cent. (7)–(8) Mean effective surface brightness in eachband and error. (9)–(10) Total apparent magnitude in each band and error. (11) UV − optical radial colour profilesclassification. RSF: recent star formation. UVX: UV-upturn. A/Z: large-scale age and/or metallicity gradient. We convolved the FUV and optical data to the spatial resolution of the NUV observationsbefore any analysis, to avoid spurious colour gradients in the inner parts. We carried outsurface photometry by measuring the surface brightness along elliptical annuli, using theiterative method described by Jedrzejewski (1987) and implemented in the
ELLIPSE taskwithin the
STSDAS ISOPHOTE package in
IRAF (Image Reduction and Analysis Facility). Thecentre of the isophotes was fixed to the centre of the light distribution and the position angle(PA), ellipticity ( ǫ ) and surface brightness ( µ ) were fitted as a function of the radius. Ellipseswere fitted to the V -band images only, which have a far superior signal-to-noise (S/N) ratioat all radii compared with the UV images, and were then imposed on the UV images sothat meaningful colours could be derived. Galactic extinction was corrected with R V = 3 . c (cid:13) , 000–000 H. Jeong et al.
Figure 2.
Comparison of our magnitudes and effective radii in the
F UV and
NUV passbands with those of Gil de Paz et al.(2007). (Cardelli, Clayton & Mathis 1989), A FUV = 8 . × E ( B − V ) and A NUV = 8 . × E ( B − V )(Wyder et al. 2005) using reddening maps from Schlegel, Finkbeiner & Davis (1998). Wepresent the GALEX
FUV, NUV and MDM F555W images as well as the surface-brightness,colour and ellipticity radial profiles in Figure 1. The vertical dashed lines in Figure 1 showthe effective radii determined by the
SAURON survey at I band (see Paper VI). For reference,a UV spectral slope of zero in the λ - F λ plane is roughly the same as FUV − NUV= 1 . − NUV colour profiles of Figure 1).We give comments on notable features in the profiles of individual galaxies in Appendix A.From the radial surface brightness profiles, we derived total apparent magnitudes by ex-trapolating the growth curves to infinity. For each profile, we first calculated the integratedapparent magnitude ( m ( R )) within elliptical isophotes up to the radius where the uncer-tainty in the surface brightness reaches 0 . − . We then computed the slope of thegrowth curve ( S ≡ dm ( R ) /dR ) within that region, plotted it as a function of the integratedmagnitude m ( R ), and fitted a straight line to the outer parts. The value of m where theextrapolated value of S is 0 was taken as the total apparent magnitude. Selecting a suitableradial range for growth curve fitting was occasionally difficult in FUV because of the low S/Nin the outer parts, but the errors quoted correspond to the uncertainties associated with thefit to the (slope of the) growth curve only, so they are probably slightly underestimated. Thistechnique is described in more detail in Cair´os et al. (2001) and Gil de Paz et al. (2007). Theeffective radius ( R e ; semi-major axis of the elliptical aperture containing half the light), themean surface brightness within R e ( h µ i e ) and their uncertainties were also calculated fromthe growth curve. Having said that, the uncertainty in R e arising from different measurement c (cid:13) , 000–000 he SAURON project - XIII ≈
20 per cent, which we adopt (see Falc´on-Barroso et al.,in prep.).In Figure 2, we show the comparison of our magnitudes and effective radii in both FUVand NUV with those of Gil de Paz et al. (2007). The agreement is generally good. The meanof the absolute values of the total apparent magnitude differences is ≈ .
16 mag in bothFUV and NUV, as expected generally somewhat larger than our formal uncertainties. Forthe effective radii, the mean absolute difference is ≈ ≈
15 per centin NUV, so generally within our errors. The integrated UV and other properties are listedin Table 1.
The simplest new data product obtained from our
GALEX observations is the integrated UVapparent magnitudes, from which we derive total absolute magnitudes using the distanceslisted in Table 1. Colour-magnitude relations are widely used to study the star formationhistory (SFH) of early-type galaxies. The NUV CMR (see, e.g., Yi et al. 2005; Kaviraj et al.2007) is a particularly good tool for tracking recent star formation, owing to its high sensi-tivity to the presence of young stellar populations. Figure 3 shows the UV–optical CMRs.It also shows the red sequence NUV − r fit of Yi et al. (2005), after correcting for the skymeasurement offset from the GALEX pipeline (see Gil de Paz et al. 2007). The scatter inthe CMRs of Figure 3 is significant and there are clearly outliers about the red sequencerelation. We henceforth attempt to quantify the effects of recent star formation on thoserelations.Most UV photons are emitted by stars younger than a few hundred megayears, so theyare useful to trace residual star formation, but FUV flux can also be generated by evolved hothelium-burning stars (the so-called UV-upturn phenomenon; see, e.g., Yi, Demarque & Oemler1997 and O’Connell 1999 for reviews). Star formation studies thus usually use the NUVpassband instead (Yi et al. 2005), although it is nevertheless important to know how muchof the NUV flux can originate from the UV-upturn phenomenon. NGC 4552 is a famousUV-upturn galaxy; its FUV − V colour is one of the bluest among nearby elliptical galaxieswith no sign of recent star formation. We measure a NUV − V colour of 5 . ± .
03 mag for c (cid:13) , 000–000 H. Jeong et al.
Figure 3.
UV colour-magnitude relations of our 34
SAURON early-type galaxies. The FUV − V ( top ), FUV − NUV ( middle ) andNUV − V ( bottom ) colours are shown as a function of the absolute magnitude in V . The dashed line indicates the NUV − V = 5 . β & − r is shown as a dotted line in theNUV − V colour-magnitude relation (with offset applied, see text). NGC 4552, so considering the measurement error we adopt a very conservative 5 σ upperlimit of NUV − V = 5 . − V colour profiles of Figure 1 and the NUV − V CMR ofFigure 3. We henceforth assume that the relative FUV strength of NGC 4552 is the maxi- c (cid:13)000
SAURON early-type galaxies. The FUV − V ( top ), FUV − NUV ( middle ) andNUV − V ( bottom ) colours are shown as a function of the absolute magnitude in V . The dashed line indicates the NUV − V = 5 . β & − r is shown as a dotted line in theNUV − V colour-magnitude relation (with offset applied, see text). NGC 4552, so considering the measurement error we adopt a very conservative 5 σ upperlimit of NUV − V = 5 . − V colour profiles of Figure 1 and the NUV − V CMR ofFigure 3. We henceforth assume that the relative FUV strength of NGC 4552 is the maxi- c (cid:13)000 , 000–000 he SAURON project - XIII ad hoc assumption,however it is frequently made and is empirically grounded.Using this empirical criterion, we have identified the galaxies that are likely to have ex-perienced a recent episode of star formation in Figure 3 onward (filled symbols). We labelgalaxies with NUV − V > . β & β indices are computed within one effective radius inthe optical and come from the SAURON data published in Paper VI. Supporting our empiricalthreshold, four out of five recent star formation candidate galaxies (NGC 3032, NGC 4150,NGC 4550 and NGC 7457) also show enhanced H β line strengths, a widely-used post-starburst signature. The only exception (NGC 474) is noted for its shell structures locatedin the outer regions, at R ≈
60 arcsec, and as shown in Figure 1 the total NUV flux isdominated by these outer regions, whereas the Balmer line was measured within R e in theoptical.A more in-depth look at Figure 3, with likely recent star formation galaxies identified,reveals that a non-negligible fraction of the scatter in the UV CMRs (and the departurefrom the Yi et al. 2005 red sequence relation) is due to galaxies with recent star formation.For example, the scatter in the (NUV − V )-M V CMR increases by 67 per cent betweenleast-square fits excluding and including RSF candidates. The overall fraction of galaxieswith RSF is 15 per cent (5 / − V or FUV − NUV on M V is visible in our sample. Finally, environment does not appear to play a significant rolein star formation, as the RSF galaxies selected contain both field and cluster galaxies. Werecall however that the densest environment probed by the SAURON survey is that of theVirgo cluster of galaxies (see Paper II), and that our sample is admittedly too small torobustly investigate environmental effects.We also plot the colour- σ e (integrated stellar velocity dispersion within R e ) and colour-mass (virial mass estimate) relations in Figure 4. The dynamical mass estimates are of theform M vir = (5 . ± . × R e σ /G (Paper IV). Recent star formation galaxies generallyhave lower velocity dispersions and correspondingly smaller virial masses than the majorityof red early-type galaxies, and there is no RSF early-type galaxy with a velocity disper-sion σ e >
200 km s − (see also Schawinski et al. 2006). This implies that only low-mass c (cid:13) , 000–000 H. Jeong et al.
Figure 4.
Colour-stellar velocity dispersion and colour-mass relations of our 34
SAURON early-type galaxies. Symbols are thesame as in Figure 3. early-type galaxies have formed stars recently, confirming the results of Schawinski et al.(2007). In addition, the more massive (high-velocity dispersion) early-type galaxies tend tobe bluer in both
FUV − V and FUV − NUV, confirming earlier studies (e.g. Burstein et al.1988; Donas et al. 2007).Interestingly, using
Spitzer
Infrared Array Camera (IRAC) imaging and Infrared Spec-trograph (IRS) spectroscopy, Shapiro et al. (in prep.) surveyed the mechanisms driving star c (cid:13) , 000–000 he SAURON project - XIII Table 2.
CO detection rates in recent star formation and quiescent early-type galaxies.RSF Partial RSF QSTNumber of sample galaxies 5 4 25Galaxies detected in CO 3 2 4Rate (percent) 60 ±
22 50 ±
25 16 ± formation in the SAURON early-type galaxies, and found that star formation happens exclu-sively in fast-rotating systems (see Paper IX). All of our recent star formation galaxies arealso classified as star-forming galaxies in their survey, except again for NGC 474, and all arefast-rotating except NGC 4550, the peculiar galaxy with co-spatial counter-rotating discs(extensively discussed in Paper IX).Our star formation interpretation is also consistent with the results of surveys of themolecular gas emission in early-type galaxies by Combes, Young & Bureau (2007), Crocker et al.(2008) and Young, Bureau & Cappellari (2008). CO emission is detected in three of the RSFcandidates (NGC 3032, NGC 4150 and NGC 4550). In the case of NGC 7457, a molecular gasmass of 4 . × M ⊙ was reported by Welch & Sage (2003), but Combes et al. (2007) laterfailed to detect it. NGC 474 was also not detected by Combes et al. (2007), but the 24 arcsecprimary beam did not encompass the shells and strong NUV emission. We summarize theCO detection rates of our sample galaxies in Table 2, where we have divided the sampleinto three categories: RSF, partial RSF and quiescent galaxies. Again, RSF are galaxiesselected by the empirical criterion NUV − V < .
0. Partial RSF galaxies do not satisfy theoverall blue colour criterion in integrated NUV − V , but they show significant blue regionsin their radial profiles. Quiescent galaxies have NUV − V > . Early-type galaxies follow a correlation between total luminosity and central velocity dis-persion of the form L ∝ σ q , where q ≈ c (cid:13) , 000–000 H. Jeong et al.
Figure 5.
Faber-Jackson relation for our sample of 34
SAURON early-type galaxies in the FUV ( top ) and NUV ( bottom )passbands. Symbols are the same as in Figure 3. The original Faber-Jackson relation ( σ ∝ L / ) is shown as a dotted line ineach panel. RSF early-types are included, the scatter increases by 65 and 69 per cent in the FUV andNUV, respectively, compared to that when only quiescent galaxies are considered . This isconsistent with the conclusions from our analysis of the colour-magnitude relations in theprevious sub-section.We now consider the Fundamental Plane (Djorgovski & Davis 1987; Dressler et al. 1987),a relation between the photometric effective radius ( R e ), the mean surface brightness within R e ( h µ i e ) and the effective stellar velocity dispersion ( σ e ) of the formlog( R e ) = a log( σ e ) + b log( h µ i e ) + c , (1)where a , b and c are constant coefficients derived by minimizing the residuals from theplane, and a = 2 . b = 0 . V band), while our photometricquantities are measured in the UV. The Fundamental Plane of course implicitly relies on allquantities used in the virial theorem being measured for the same stellar population. OurUV Fundamental Planes are thus somewhat hybrid, and one should keep in mind that thismay introduce some biases (because of both differing apertures and wavebands). However,as UV-derived stellar kinematics is still some time away, this approach is necessary andnevertheless illuminating.The coefficients a and b measured by Jørgensen et al. (1996) are 1 . ± .
07 and 0 . ± . For reference, simple linear fits to the whole sample yield the slopes of 3 . +0 . − . (FUV) and 3 . +0 . − . (NUV), whereas forthe quiescent sub-sample we find 4 . +0 . − . (FUV) and 3 . +0 . − . (NUV). c (cid:13) , 000–000 he SAURON project - XIII Table 3.
Fundamental Plane coefficients at FUV and NUV.Band a b
Scatter PopulationFUV 1.09 0.22 0.17 AllFUV 1.25 0.29 0.14 QSTNUV 1.11 0.23 0.14 AllNUV 1.82 0.30 0.08 QST in the Gunn r -band, respectively. Early investigations suggested that a variation of thecoefficient a in different filters would be expected if there were a systematic variationof the mass-to-light ratio of the stellar populations with galaxy luminosity or mass (e.g.Prugniel & Simien 1996). Pahre, De Carvalho & Djorgovski (1998) indeed reported an in-crease of the slope with increasing wavelength (from U to K ), although Bernardi et al.(2003) found that the FP coefficients were approximately the same in g , r , i and z bandsusing a much larger but more distant sample of early-type galaxies from SDSS. Meanwhile,Trujillo, Burkert & Bell (2004) suggested that the FP tilt is mostly driven by non-homologyand that stellar population effects account for only a small fraction of it. Early-types in fieldenvironments are also generally more diverse and are found to show a greater scatter inthe Fundamental Plane (see, e.g., De Carvalho & Djorgovski 1992; Zepf & Whitmore 1993).These authors thus suggested that environment plays an important role in the process ofgalaxy formation and evolution. Recently, new approaches have aimed to elucidate the FPtilt, for example via gravitational lensing (Treu et al. 2006; Bolton et al. 2008) and stellardynamical modeling (Paper IV). These studies differ from previous ones in that they do notdepend on simple virial assumptions, but measure accurate dynamical masses directly. Usingtwo different approaches, they consistently conclude that the tilt is almost entirely due toa genuine mass-to-light ratio variation and not to non-homology. However, it still remainsunclear just how much of this variation can be attributed to a change in the dark matterfraction or to differences in the stellar populations.For the first time, we present in Figures 6 and 7 the Fundamental Planes in the FUV andNUV bands, using again the same symbols as in Figure 3. Table 3 lists the FP coefficients andscatter at both FUV and NUV, obtained from a least-square fit by minimizing the variationsto the planes. Fits including and excluding RSF galaxies are listed. As mentioned above, RSFgalaxies tend to have lower velocity dispersions and smaller effective radii and thus smallermasses than the bulk of the sample galaxies. Crucially, RSF galaxies systematically deviatefrom the best-fit planes so as to create shallower slopes, and they significantly increase the c (cid:13) , 000–000 H. Jeong et al.
Figure 6.
Fundamental Plane and projections of the Fundamental Plane in the FUV band. Symbols are the same as in Figure 3and linear fits to the two sub-samples (whole sample, quiescent galaxies only) are shown as dotted and solid lines, respectively.We also present a linear fit to the whole sample as a solid line in the bottom-right plot (Kormendy relation). scatters. This is clearly shown by both Figures 6 and 7 and the FP parameters listed inTable 3. Although our sample is limited, this indicates that lower mass galaxies, forminga larger amount of young stars, more greatly (and systematically) deviate from the planedefined by the quiescent galaxies.The long-wavelength K -band slope is less sensitive to contamination from young stellarpopulations than optical bands ( a ≈ .
53; see Pahre et al. 1998). Considering our NUVFundamental Plane fit for the entire sample, our slope is significantly different from boththe virial expectation and the K -band value. However, comparing our NUV-derived slopeof 1 .
82 for quiescent galaxies only, we recover all of the K -band slope and almost all of thetilt away from the virial expectation. We thus conclude that a significant fraction of the FPtilt and scatter is due to low-mass early-type galaxies with stellar populations significantly c (cid:13) , 000–000 he SAURON project - XIII Figure 7.
Same as Figure 6 but for the NUV band. younger that those of high-mass galaxies, thus reducing their mass-to-light ratios. The FUVFundamental Plane is more complicated to interpret because of the additional presence ofold but hot stars (UV upturn phenomenon), but a similar if weaker trend is observed.The tilt of the Fundamental Plane was also considered in Paper IV using V -band pho-tometry and sophisticated dynamical modeling. The main result to emerge was a clear,continuous and monotonic increase of the total dynamical mass-to-light ratio with velocitydispersion, consistent with the FP tilt. As can be seen here in Figure 4, the NUV blue galax-ies affecting the tilt of the UV Fundamental Planes also have some of the smallest velocitydispersions and lowest virial masses. Furthermore, although there is non-negligible scatter,clear (NUV − V )- σ e and (FUV − NUV)- σ e (or (NUV − V )- M vir and (FUV − NUV)- M vir ) rela-tions are observed. The current results thus seem entirely consistent with that of Paper IV. c (cid:13) , 000–000 H. Jeong et al.
We are simply picking up the galaxies with the largest and/or most recent star formationevents here with our integrated NUV − V < . R e and h µ i e , i.e. the Kormendy relation (Kormendy 1977). A clear advantage ofthis relation is that it is based on photometric parameters only, which are easily accessibleeven at high redshifts. All of our sample galaxies follow the Kormendy relation, and wenote that our UV Kormendy relation slopes (3 . ± .
49 and 3 . ± .
36 at FUV and NUV,respectively, see Figures 6 and 7) are in rough agreement with the optical one ( h µ i e ≈ .
02 log R e in Hamabe & Kormendy 1987). It may however be significant that, comparedto the mean trends, RSF galaxies tend to be slightly offset toward smaller and/or fainterobjects. We use the radial colour profiles to discuss the stellar populations of the sample galaxies.We discuss our criteria to classify them in groups exhibiting similar properties below, andthen discuss in more detail galaxies with signs of recent star formation.First, our main tool are the NUV − V radial colour profiles. As for integrated magnitudes,we classify a galaxy as having had recent star formation if its NUV − V profile is below 5 . − optical colours normally show negative slopes (thecentral regions being redder) due to variations in the properties of the underlying stellarpopulations, mainly age and metallicity. It is well-known that there is a significant age-metallicity degeneracy (see, e.g., Worthey 1994), whereby it is difficult to distinguish theeffects of a small change in age from those of a small change in metallicity. Therefore, andeven though we recognise the qualitative aspect of such a criterion, if only a small part –usually outside the effective radius – of the UV − optical colour profiles goes smoothly andonly a little below the NUV − V = 5 . − V < . c (cid:13)000
02 log R e in Hamabe & Kormendy 1987). It may however be significant that, comparedto the mean trends, RSF galaxies tend to be slightly offset toward smaller and/or fainterobjects. We use the radial colour profiles to discuss the stellar populations of the sample galaxies.We discuss our criteria to classify them in groups exhibiting similar properties below, andthen discuss in more detail galaxies with signs of recent star formation.First, our main tool are the NUV − V radial colour profiles. As for integrated magnitudes,we classify a galaxy as having had recent star formation if its NUV − V profile is below 5 . − optical colours normally show negative slopes (thecentral regions being redder) due to variations in the properties of the underlying stellarpopulations, mainly age and metallicity. It is well-known that there is a significant age-metallicity degeneracy (see, e.g., Worthey 1994), whereby it is difficult to distinguish theeffects of a small change in age from those of a small change in metallicity. Therefore, andeven though we recognise the qualitative aspect of such a criterion, if only a small part –usually outside the effective radius – of the UV − optical colour profiles goes smoothly andonly a little below the NUV − V = 5 . − V < . c (cid:13)000 , 000–000 he SAURON project - XIII − V = 5 . − NUV profiles. If a galaxy without evidence of recent starformation has a central region with FUV − NUV < .
0, we identify it as a UV-upturn galaxy.These galaxies include NGC 4374, NGC 4486, NGC 4552, NGC4621 and NGC 5846. As ex-pected, these galaxies are among the most luminous and most massive in our sample. Allexcept NGC 4621 are slow-rotators (see Paper IX). Significantly, some UV-upturn galaxieshave extreme FUV − NUV and FUV − V colours at small radii, bluer than the integratedaperture colours of classic UV-upturn galaxies dating from the International UltravioletExplorer ( IUE ) era. As these galaxies provided important constraints on stellar evolution-ary models of the UV-upturn, it will be interesting to revisit the models with the currentdata. It is also probably worthwhile to point out that a number of galaxies exhibit evi-dence for both a UV-upturn (FUV − NUV < . − V < . − optical colours in someregions only (as opposed to all radii). Interestingly, this appears connected to kinematic (sub-)structures detected in SAURON and other studies. The case of NGC 474, known for its shells(e.g. Turnbull, Bridges & Carter 1999) and strong misalignment between the kinematic andphotometric major axes (see Paper III), shows blue UV − optical colours and enhanced H β line strength at large radii. NGC 1023, with a prominent twist in the centre of the velocitymap (Paper III), reveals a UV blob on the eastern side of the galaxy and blue UV − opticalcolours in the same region, probably related to an interacting gas-rich dwarf galaxy (seeMorganti et al. 2006 for H I observations). The presence of a UV-prominent outer ring re-lated to a bar in NGC 2974 was discussed by Jeong et al. (2007). Blue UV − optical coloursin this galaxy are observed in the central and outer regions. In the case of NGC 4459, blueUV − optical colours are observed only in the centre. Paper VI shows that this galaxy has c (cid:13) , 000–000 H. Jeong et al. strong H β absorption within a central dust ring associated with a decoupled stellar andionised gas disc (Paper III and Paper V, respectively). NGC 4526 is a similar object appear-ing more edge-on. Both NGC 4459 and NGC 4526 also harbor central molecular gas discs(Young et al. 2008).On the other hand, of the nine galaxies with recent star formation, NGC 3032, NGC 4150,NGC 4550 and NGC 7457 show overall blue UV − optical colours. NGC 3032 shows thebluest UV − optical colour in our sample and has a prominent peak in the H β map sug-gesting recent and ongoing star formation. McDermid et al. (2006) identified two types ofkinematically-decoupled components (KDCs) among the SAURON sample: compact KDCs,which are often young and occur in fast-rotating galaxies, and kiloparsec-scale KDCs, withhomogeneously old populations in slow-rotating galaxies (see also Paper IX). NGC 3032,NGC 4150 and NGC 7457 all have small-scale KDCs (Paper III) and are considered fastrotators (Paper IX). NGC 3032 and NGC 4150 further have substantial molecular gas discs(Young et al. 2008). In the case of NGC 4550, with two counter-rotating stellar discs (see,e.g., Rubin, Graham & Kenney 1992; Rix et al. 1992), the H β line strength is elevated inthe central region along the major axis (Paper VI). Its molecular gas content and detailedstar formation history are discussed in Crocker et al. (2008). NGC 7457 has a controversialCO detection discussed previously (see Welch & Sage 2003 and Combes et al. 2007). We have presented
GALEX
FUV and NUV imaging along with ground-based F555W imag-ing from the MDM Observatory for 34 early-type galaxies from the
SAURON survey sample.Nine of them show extended blue UV − optical colours, hinting at recent star formation. Fiveof these are also classified as RSF early-types by our integrated colour-magnitude relationcriterion. Supporting the findings from the UV colour-magnitude relation technique, fourout of the five candidate RSF galaxies also show enhanced H β absorption line strengths, awidely-used post-starburst signature. The only exception (NGC 474) is noted for its outershells. Roughly 15 per cent of the early-type galaxies in our sample are therefore classifiedas RSF early types. Considering that the UV flux from a starburst is only detectable forroughly 1 Gyr, this implies that residual star formation has been common in early-typegalaxies even up to the present day. This star formation interpretation is also consistent c (cid:13) , 000–000 he SAURON project - XIII σ e >
200 km s − , consistent with the in-dependent work of Schawinski et al. (2007). Recent star formation early types also tend tohave smaller effective radii and thus smaller (virial) masses. Despite the limited numberstatistics, a key result from the present study is that RSF early types change the slopes ofscaling relations (colour-magnitude relations and Fundamental Planes) and dominate thescatters in them. Most notably, much of the FP tilt and scatter can now be explained bythe fact that the properties of a substantial fraction of early-type galaxies are influencedby RSF, systematically biasing their mass-to-light ratios, especially at low masses. The UVFundamental Planes become significantly closer to the virial expectation (and tighter) whenonly quiescent early-type galaxies are considered. The same must be true at optical wave-lengths, although the effect will be smaller. Our result appears consistent with the M/L - σ e relation derived in Paper IV, although the current colour threshold only picks up the mostextreme objects.The radial UV − optical colour profiles not only reveal recent star formation galaxies,but also a number of galaxies with smooth large-scale age and/or metallicity gradients.Similarly, a number of galaxies exhibiting the UV-upturn phenomenon are identified. Somehave bluer FUV − NUV and FUV − V colours at small radii than the integrated aperturecolours of classic UV-upturn galaxies. A number of galaxies show both a central UV-upturnand large-scale gradients. This diversity of behaviours will, in due time, need to be explainedand reproduced by stellar population models.Early-type galaxies are no longer thought to be simple. Many of them have kinematicanomalies and sub-structures that are non-trivial to interpret. We have shown here thatthey are also composed of multiple generations of stars widely separated in age. This isconsistently found by other short-wavelength as well as far-infrared studies, mainly aidedby space experiments. The presence of young stars in seemingly old populations is no longerdebatable, yet it is still unclear what causes and regulates this residual star formation.We believe that our database, spatially resolving nearby early-type galaxies, will be key toanswering some of those questions. c (cid:13) , 000–000 H. Jeong et al.
ACKNOWLEDGMENTS
The authors thank the anonymous referee for useful comments which led to improvementsin the paper. This work was supported by the Korea Research Foundation Grant fundedby the Korean government (KRF-C00156) to SKY. MB acknowledges support from NASAthrough GALEX Guest Investigator program GALEXGI04-0000-0109. MB and SKY aregrateful to the Royal Society for an International Joint Project award (2007/R2) supportingthis work. RLD acknowledges support from the Royal Society in the form of a Wolfson MeritAward. MB and RLD are also grateful for postdoctoral support through STFC rolling grantPP/E001114/1. The STFC Visitors grant to Oxford also supported joint visits. GvdV ac-knowledges support provided by NASA through Hubble Fellowship grant HST-HF-01202.01-A awarded by the Space Telescope Science Institute, which is operated by the Association ofUniversities for Research in Astronomy, Inc., for NASA, under contract NAS 5-26555. MCacknowledges support from a STFC Advanced Fellowship (PP/D005574/1). Based on ob-servations made with the NASA Galaxy Evolution Explorer. GALEX is operated for NASAby the California Institute of Technology under NASA contract NAS5-98034. Photometricdata were also obtained using the 1.3m McGraw-Hill Telescope of the MDM Observatory.Part of this work is based on data obtained from the ESO/ST-ECF Science Archive Facil-ity. This project made use of the HyperLeda database (http://leda.univ-lyon1.fr) and theNASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Lab-oratory, California Institute of Technology, under contract with the National Aeronauticsand Space Administration.
REFERENCES
Bender R., Burstein D., Faber S. M., 1992, ApJ, 399, 462Bernardi M. et al., 2003, AJ, 125, 1866Bolton A. S., Treu T., Koopmans L. V. E., Gavazzi R., Moustakas L. A., Burles S., SchlegelD. J., Wayth R., 2008, ApJ, 684, 248Bower R. G., Lucey J. R., Ellis R., 1992, MNRAS, 254, 589Burstein D., Bertola F., Buson L. M., Faber S. M., Lauer T. R., 1988, ApJ, 328, 440Cair´os L. M., Caon N., V´ılchez J. M., Gonz´alez-P´erez J. N., Mu˜no-Tu˜n´on C., 2001, ApJS,136, 393Cardelli J. A., Clayton G. C., Mathis J. S., 1989, ApJ, 345, 245 c (cid:13) , 000–000 he SAURON project - XIII c (cid:13) , 000–000 H. Jeong et al.
Krajnovi´c D. et al. 2008, MNRAS, 390, 93Hamabe M., Kormendy J., 1987, in IAU Symposium 127, Structure and Dynamics ofElliptical Galaxies, ed. T. de Zeeuw (Dordrecht: Reidel), 379Kormendy J., Bender R., 1996, ApJ, 464, L119Kuntschner H. et al., 2006, MNRAS, 369, 497 [Paper VI]Larson R. B., 1974, MNRAS, 166, 385Martin D. C. et al., 2005, ApJ, 619, L1McDermid R. M. et al., 2006, MNRAS, 373, 906 [Paper VIII]Morganti R. et al., 2006, MNRAS, 371, 157Morrissey P. et al., 2005, ApJ, 619, L7Morrissey P. et al., 2007, ApJS, 173, 682Nolan L. A., Dunlop J. S., Panter B., Jimenez R., Heavens A., Smith G., 2007, MNRAS,375, 371O’Connell R. W., 1999, ARA&A, 37, 603Pahre M. A., Djorgovski S. G., De Carvalho R. R., 1998, AJ, 116, 1591Pahre M. A., De Carvalho R. R., Djorgovski S. G., 1998, AJ, 116, 1606Prugniel Ph., Simien F., 1996, A&A, 309, 749Raimann D., Storchi-Bergmann T., Quintana H., Hunstead R., Wisotzki L., 2005, MNRAS,364, 1239Rix H.-W., Franx M, Fisher D., Illingworth G., 1992, ApJ, 400, L5Rubin V. C., Graham J. A., Kenney J. D. P., 1992, ApJ, 394, L9Sarzi M. et al., 2006, MNRAS, 366, 1151 [Paper V]Schawinski K. et al., 2006, Nature, 442, 888Schawinski K. et al., 2007a, ApJS, 173, 512Schlegel D. J., Finkbeiner D. P., Davis M., 1998, ApJ, 500, 525Schweizer F., Seitzer P., 1992, AJ, 104, 1039Scodeggio M., 1997, Ph.D. thesis, Cornell Univ.Stanford S. A., Eisenhardt P. R. M., Dickinson M., 1998, ApJ, 492, 461Toomre A., Toomre J., 1972, ApJ, 178, 623Treu T., Koopmans L. V. E., Bolton A. S., Burles S., Moustakas L. A., 2006, ApJ, 640,662Trujillo I., Burkert A., Bell E., 2004, ApJ 600, 39Turnbull A. J., Bridges T. J., Carter D., 1999, MNRAS, 307, 967 c (cid:13)000
Krajnovi´c D. et al. 2008, MNRAS, 390, 93Hamabe M., Kormendy J., 1987, in IAU Symposium 127, Structure and Dynamics ofElliptical Galaxies, ed. T. de Zeeuw (Dordrecht: Reidel), 379Kormendy J., Bender R., 1996, ApJ, 464, L119Kuntschner H. et al., 2006, MNRAS, 369, 497 [Paper VI]Larson R. B., 1974, MNRAS, 166, 385Martin D. C. et al., 2005, ApJ, 619, L1McDermid R. M. et al., 2006, MNRAS, 373, 906 [Paper VIII]Morganti R. et al., 2006, MNRAS, 371, 157Morrissey P. et al., 2005, ApJ, 619, L7Morrissey P. et al., 2007, ApJS, 173, 682Nolan L. A., Dunlop J. S., Panter B., Jimenez R., Heavens A., Smith G., 2007, MNRAS,375, 371O’Connell R. W., 1999, ARA&A, 37, 603Pahre M. A., Djorgovski S. G., De Carvalho R. R., 1998, AJ, 116, 1591Pahre M. A., De Carvalho R. R., Djorgovski S. G., 1998, AJ, 116, 1606Prugniel Ph., Simien F., 1996, A&A, 309, 749Raimann D., Storchi-Bergmann T., Quintana H., Hunstead R., Wisotzki L., 2005, MNRAS,364, 1239Rix H.-W., Franx M, Fisher D., Illingworth G., 1992, ApJ, 400, L5Rubin V. C., Graham J. A., Kenney J. D. P., 1992, ApJ, 394, L9Sarzi M. et al., 2006, MNRAS, 366, 1151 [Paper V]Schawinski K. et al., 2006, Nature, 442, 888Schawinski K. et al., 2007a, ApJS, 173, 512Schlegel D. J., Finkbeiner D. P., Davis M., 1998, ApJ, 500, 525Schweizer F., Seitzer P., 1992, AJ, 104, 1039Scodeggio M., 1997, Ph.D. thesis, Cornell Univ.Stanford S. A., Eisenhardt P. R. M., Dickinson M., 1998, ApJ, 492, 461Toomre A., Toomre J., 1972, ApJ, 178, 623Treu T., Koopmans L. V. E., Bolton A. S., Burles S., Moustakas L. A., 2006, ApJ, 640,662Trujillo I., Burkert A., Bell E., 2004, ApJ 600, 39Turnbull A. J., Bridges T. J., Carter D., 1999, MNRAS, 307, 967 c (cid:13)000 , 000–000 he SAURON project - XIII APPENDIX A: DESCRIPTION FOR INDIVIDUAL GALAXIES
We briefly comment here on the
GALEX and MDM surface brightness profiles of the E/S0galaxies presented in this paper. Individual comments on the stellar kinematics, the emission-line maps and the line strength structures are given in Paper III, Paper V and Paper VI,respectively.
NGC 474:
This galaxy (Arp 227), noted for its shell structures (e.g. Turnbull, Bridges & Carter1999), shows blue NUV − V colours except in the central regions. The H β absorption linestrength rises toward larger radii (Paper VI). Considering the UV − V colours, it seems thatyoung stellar populations are present in the outer parts. NGC 524:
This galaxy shows red colours in both the FUV − V and the NUV − V colourprofile, making a good example of a quiescent elliptical galaxy. NGC 821:
This edge-on galaxy with a rapidly rotating disc-like component (Paper III)shows red FUV − NUV and UV − V colours except outside the effective radius. NGC 1023:
This SB0 galaxy shows a tilt in the FUV and NUV surface brightness profilesaround 25 arcsec, a central concentration in all metal lines (Paper VI) and significantlynegative values of both h and h (Paper III). We discovered a strong NUV blob, constitutingstrong evidence for recent/ongoing star formation, just outside the galaxy. This is likelythe result of a tidal interaction between NGC 1023 and a neighboring gas-rich galaxy (seeMorganti et al. 2006). c (cid:13) , 000–000 H. Jeong et al.
NGC 2695:
This galaxy shows blue FUV − NUV colours in the central regions with ahigh amplitude in h . The UV − V colours become bluer toward larger radii, albeit withlarge uncertainties. We thus consider this a red galaxy. NGC 2699:
The mean velocity map of this galaxy shows a rapidly rotating componentaround 5 arcsec (Paper III). Similar to NGC 2695, it shows moderately blue NUV − V coloursonly in the outer parts. NGC 2768:
This galaxy with a cylindrical velocity field shows red colours in the UV − V colour profiles. NGC 2974:
This galaxy with a UV outer ring shows a blue colour in the outer regions,suggesting young stellar populations associated with a large-scale bar (see Jeong et al. 2007),although the H β line strength map appears relatively flat (Paper VI). NGC 3032:
Blue UV − V colours are observed everywhere, suggesting a strong ongoingstarburst. This galaxy also shows strong H β absorption with negative Mg b and Fe5270 s linestrength gradients (Paper VI). NGC 4150:
This galaxy with a counter-rotating core (Paper III) shows blue UV − V colours and enhanced H β absorption with a strong drop in the Mg b line strength (Paper VI). NGC 4278:
This galaxy shows blue FUV − NUV colours within 20 arcsec, suggesting astrong FUV flux. It also has the strongest line emission (Paper V).
NGC 4374 (M84):
This giant elliptical galaxy shows a similar FUV − NUV colour trendas that in NGC 4278. It is known for its BL Lac nucleus and shows strong ionised-gasemission (Paper V).
NGC 4387:
This boxy galaxy shows red UV − V colours. NGC 4458:
The FUV image of this galaxy exhibits an indistinct shape except for thecentral regions. This faint disc-like object also has a small central kinematically-decoupledcore (Paper III).
NGC 4459:
This galaxy with a central dust ring (Paper V) shows blue UV − V colourswithin 10 arcsec, coincident with an intense H β absorption line strength and molecular gas. NGC 4473:
This galaxy with a high velocity dispersion along the major axis likely re-sulting from substantial counter-rotation (Paper III) shows extended UV emission in theoutskirts.
NGC 4477:
This galaxy with a prominent misalignment between the stellar kinematicand photometric major axes (Paper III shows overall red UV − V colours. NGC 4486 (M87):
This galaxy with a prominent jet shows blue UV − V colours in the c (cid:13) , 000–000 he SAURON project - XIII NGC 4526:
The UV − V plots of this galaxy exhibit blue colours in the centre, coincidentwith strong H β absorption (Paper VI) and a stellar velocity dispersion drop due to a star-forming fast-rotating disc (Paper III and Paper V). NGC 4546:
This galaxy shows moderately blue NUV − V colours in the outer parts. Thesecould be related to the likely presence of a bar and associated star formation. NGC 4550:
This galaxy with two counter-rotating stellar discs (see Rubin, Graham & Kenney1992; Rix et al. 1992) shows blue UV − V colours everywhere. NGC 4552 (M89):
This famous UV-upturn galaxy shows very blue FUV − NUV coloursin the centre.
NGC 4564:
This elongated galaxy with a disc-like component (Paper III) shows blueFUV − NUV colours within the inner 10 arcsec.
NGC 4570:
This fast-rotating edge-on galaxy shows slightly blue UV − V colours in theouter parts. The presence of a bar was claimed by van den Bosch & Emsellem (1998). NGC 4621:
This object has a disc-like component similar to NGC 4564 (Paper III).A counter-rotating component is also detected inside 2 arcsec (Wernli, Emsellem & Copin2002). Blue FUV − NUV colours are observed in the centre.
NGC 5198:
This galaxy with a central kinematically-decoupled core shows blue FUV − NUVcolours in the centre that may indicate the presence of a UV upturn.
NGC 5308:
This edge-on disc galaxy shows red UV − V colours. NGC 5813:
This galaxy with a well-known kinematically decoupled core shows marginallyblue NUV − V colours in the outer parts. NGC 5831:
Like NGC 5813, this galaxy with a kinematically-decoupled core shows blueNUV − V colours in the outer regions. There is a small bump in both FUV and NUV surfacebrightness profiles around 35 arcsec. NGC 5838:
This fast-rotating object shows red NUV − V colours throughout. NGC 5845:
This compact elliptical galaxy shows a small bump in the NUV profile at aradius of ≈
10 arcsec, with a corresponding distinct peak in the velocity map related to thecentral disk.
NGC 5846:
Similar to NGC 4552 and NGC 4564, this bright giant elliptical galaxy showsblue FUV − NUV colours in the centre that are suspected to be a UV upturn, with marginallyblue NUV − V colours. c (cid:13) , 000–000 H. Jeong et al.
NGC 5982:
This galaxy with a kinematically-decoupled core shows blue FUV − NUVcolours in the centre.
NGC 7457:
This kinematically-decoupled-core galaxy shows blue NUV − V colours through-out, suggesting recent star formation. The H β absorption is also high everywhere (Paper VI). c (cid:13)000