arXiv:astro-ph/0701018v2 3 Jan 2007 con o h omto fMln1tp aaisa rare as galaxies 1-type Malin might of which formation proposed the been for have account easily scenarios extende not enormously are although such halos of disks, matter properties the dark reproduce cold to in Mo able formation formation. disk galaxy of po of els 1 theories Malin of to the properties challenges be such the 2004), to interesting if al. appears et Even which Minchin rare, (e.g., case are effects. galaxies LSB selection likely giant brightness extreme would surface 1 b of Malin surveys cause galaxy like magnitude-limited galaxies in underrepresented (1997), be emphasiz al. As et significant. Bothun are by galaxies disk of population all and warping some motions. shows noncircular structure velocity the and uncertain ag s10kcfo h aaycne,ipyn dynam- a 10 implying center, order galaxy of the mass from ical kpc 110 as large naypoi au of value asymptotic an 97 ates a re,&MnirRgin 01.The 2001). Monnier-Ragaigne H & Driel, van Matthews, 1997; iul nw.Teepnnilsaelnt fti disk this of length be scale to exponential determined The was known. viously rgteso t ik ai samsieglx iha with galaxy massive a is 1 of Malin luminosity optical disk, total its of brightness e.Frhroe ti n ftems a-ihgalaxies gas-rich most the of one H surface is an a val it with with extreme known, Furthermore, these found approach been that ues. un- has length are scale disk 1 and galaxy Malin brightness other of 1995; no properties 1993, the usual; al. 1999), et sub- Sprayberry al. discovered et 1990; Beijersbergen galaxies in al. et LSB Even (Bothun giant sequently galaxy. other proto (LSB) the with brightness considered comparison surface is low it giant properties, these typical on Based 1997). ohv netaoae eta ufc rgteso only of brightness surface central extrapolated µ an of have extent to emis- to radial brightness out low-surface largest extending with the sion galaxy, by has spiral work known later it in any the (1989), and in (1987) Bothun described al. & et Impey As Bothun by decades. paper two discovery nearly for object puzzling 0 rpittpstuigL using typeset Preprint h mlctoso hsojc o nesadn h over- the understanding for object this of implications The I h aayMln1hsrmie iglryuuuland unusual singularly a remained has 1 Malin galaxy The ( V oaincremaue yPceige l 19)reaches (1997) al. et Pickering by measured curve rotation ) ≈ rgtesof brightness htMln1cnan omlselrds htwsntprevi not was that disk stellar normal a contains 1 Malin that o ufc rgtesds aais xmnto fa arc an of Examination galaxies. disk brightness surface low motneo h xeddotrds ein o ulunde headings: full Subject proper a disk for in regions disk outliers galaxies. outer extreme extended such the prese that of The importance suggests galaxies. 1 disk Malin brightnes known surface in other central all and to length relative scale outlier disk its of terms fMln1i hto yia B/ aay h eakbyex remarkably The galaxy. SB0/a typical a of that r is 1 Malin of 25 ≈ ic t icvr,Mln1hsbe osdrdteprototy the considered been has 1 Malin discovery, its Since eateto hsc n srnm,42 rdrc Reines Frederick 4129 Astronomy, and Physics of Department . a arcsec mag 5 0 p per ob htmtial itntcmoeta component distinct photometrically a be to appears kpc 100 A I T E asof mass tl mltajv 6/22/04 v. emulateapj style X 1. 12 µ v ∼ ∼ r 0 INTRODUCTION M 9 ms km 190 = − 20 = 0kc ept h an surface faint the Despite kpc. 70 0 p,adisds a found was disk its and kpc, 100 2 ⊙ M aais niiul(ai )—glxe:sia galaxie — spiral galaxies: — 1) (Malin individual galaxies: a ane hnayglx pre- galaxy any than fainter far , V ∼ lhuhteds nlnto is inclination disk the although . a arcsec mag 1 ≈ 5 OMLSELRDS NTEGLX AI 1 MALIN GALAXY THE IN DISK STELLAR NORMAL A × − 22 10 − 1 . 10 a Pceige al. et (Pickering mag 9 n xed ordias radii to extends and M ⊙ − 2 Pceige al. et (Pickering n cl egho . p.Ott aisof radius a to Out kpc. 4.8 of length scale a and A ARON ABSTRACT ed d- se e- d - - al nvriyo aiona rie A96747;bar 92697-4575; CA Irvine, California, of University Hall, .B J. l pc Telescope Space 1989). ble Bothun & (Impey formation star of thres onset and the lab stability for interesting disk of particularly issues a LSB understanding also for giant ratory is gas-rich 1 and Malin diffuse known, most galaxi galaxy by the disk As or ordinary in 2001). 1992) mixing (Noguchi Wyse radial & Silk, and (Hoffman, evolution secular voids in peaks density rjce nua cl f15 p arcsec kpc 1.53 of scale angular projected epnigt uioiydsac of distance luminosity a to responding nucleus. active a its dispersion of velocity the classification bulge the its of reexamine derive spectrum to to optical used also new the is A from galaxy point population. data aberrant galaxy central extremely observed gala of an real terms removing by by populated (in disks length) space scale and parameter significant brightness merely of observation surface not range new is the this This as truncates galaxy. distinction, brightness disk semantic surface known a low any extended of sat- regions most not gala outer the LSB does having an 1 as its Malin classified being despite disk for this criteria of standard fairly the presence spiral, a isfy the barred contains early-type on actually an based 1 of that typical Malin disk that stellar shown normal is ground-based It on fr based imaging. different determined been rather previously is had morphology what optical galaxy’s the that with spr of part as cosmology. this to rescaled been have paper this literatur in the listed from taken quantities observed consistency, iw ahobtwssltit oridvda xoue fo of exposures field individual WFPC2 four full into the split of was center orbit the Each near view. CCD WF3 the on 0 h aaywsosre o n ri ahi h F300W the in each orbit one for ( observed was galaxy The 10. srnm,Ic,udrNS otatNS5-26555. Universiti NAS Telescope of contract Association NASA Space under the the Inc., by Astronomy, at operated Archive is Data which the stitute, from obtained scope, ARTH ,Mln1wsoiial on ob eyremote very a be to found originally was 1 Malin s, U c fads fnra ieadsraebrightness surface and size normal of disk a of nce hival 1 hspprpeet exmnto fa archival an of reexamination a presents paper This ai a esito .85(ikrn ta.19) cor- 1997), al. et (Pickering 0.0825 of redshift a has 1 Malin mgso ai eeotie ihteWP2camera WFPC2 the with obtained were 1 Malin of Images bn)adF1W( F814W and -band) ul eonzd aigacentral a having recognized, ously sadn ftesrcueadfraino spiral of formation and structure the of rstanding eadms xrm xml ftecaso giant of class the of example extreme most and pe ae nosrain aewt h AAEAHbl pc Te Space Hubble NASA/ESA the with made observations on Based ispoal onteit u nesoe the underscores but exist, not do probably ties H ubeSaeTlsoeI Telescope Space Hubble dntasml xeso fteinrds.In disk. inner the of extension simple a not nd 0 edd an ue tutr eetdotto out detected structure outer faint tended, 1k s km 71 = HST 2. RHVLDT N OBSERVATIONS AND DATA ARCHIVAL rga O54 P:Ipy n19 March 1996 on Impey) (PI: GO-5946 program − 1 :structure s: ( Mpc HST I bn)fitr,wt h aayplaced galaxy the with filters, -band) − mg fMln1 hc reveals which 1, Malin of image ) 1 1 , ∼ Ω M 0kc h structure the kpc, 10 bn mg reveals image -band 0 = . 7 and 27, I bn surface -band D L [email protected] − 1 7 p n a and Mpc 370 = o cosmology a for Ω sfrRsac in Research for es Λ 0 = cec In- Science . 3 For 73. holds and e Hub- and om nd xy xy le- o- es ly r 2 BARTH removal of cosmic-ray hits. The F300W exposures were read point source component was added to allow for the possibility out with 2 × 2 pixel binning on the CCD, while the F814W of unresolved emission from the active nucleus. The point- images were read out in the standard unbinned mode. Total spread function for the F814W filter was generated using the exposure times were 2100 s in each filter. This dataset was TinyTim software package (Krist 1993). previously analyzed by O’Neil et al. (2000), who studied the Figure 2 shows the results of the GALFIT modeling. The properties of faint galaxies in the same WFPC2 field as Malin structure of the galaxy is reproduced well overall, although 1, but they did not discuss the structure of Malin 1 itself. there are systematic residuals around the bulge and inner bar The images were retrieved from the HST data archive and region. These residuals could be reduced somewhat if the individual exposures were combined using standard tasks in constraint on the Sérsic index of the bar were relaxed, at the IRAF/STSDAS2. Figure 1 shows a portion of the F814W expense of having an unrealistic bar profile with a very cen- WFPC2 mosaic centered on Malin 1. An image of the en- trally peaked structure. However, whether the bar profile is tire WFPC2 mosaic is shown by O’Neil et al. (2000). If the constrained or not in the fit has almost no effect on the cen- full HST image is viewed with extremely high contrast, some tral surface brightness or scale length of the disk component, of the very low surface brightness outer spiral structure can since the disk parameters are mainly determined from the re- just barely be seen at radii of up to ∼ 1′ from the galaxy cen- gion around r =4′′ − 6′′ where it dominates the galaxy profile. ter, and the faint spiral features match the structures visible Figure 3 shows the galaxy’s radial profile as measured from in Fig. 2 of Bothun et al. (1987). However, the S/N in the the WFPC2 image and from the GALFIT model. The ra- WFPC2 image at these large radii is too low to perform use- dial profiles were measured using the IRAF task ELLIPSE, ful measurements of the surface brightness of the outer disk. which fits elliptical isophotes to the image at a series of fixed The F300W image also has very low S/N and will not be dis- semimajor axis lengths, following the methods described cussed further in this paper. by Jedrzejewski (1987). Conversion from HST F814W fil- A new opticalspectrum of Malin 1 was obtainedat the Keck ter magnitudes to standard Cousins I-band Vega-based mag- II telescope on 2005 May 16 UT with the ESI spectrograph nitudes was done using the SYNPHOT package in IRAF, (Sheinis et al. 2002). The exposure time was 1800 s with a assuming an S0-type spectrum from the spectral atlas of 0.′′75-wide slit, yielding a spectrum with R ≈ 6000 over the Kinney et al. (1996). The brightness profiles include a cor- range 3800–10000 Å and S/N ≈ 20 per pixel in the contin- rection for Galactic extinction (AI =0.067mag; Schlegel et al. uum at 5100 Å. The slit was oriented at the parallactic angle 1998), a K-correction of 0.08 mag for the I band, determined for the midpoint of the observation. Spectral extraction was using the Kinney et al. (1996) S0 template spectrum, and a ′′ correction for cosmological surface brightness dimming of done using a 1 width, and the extracted echelle orders were ′′ wavelength-calibrated using observations of HgNe, Xe, and 0.34 mag. Brightness profiles derived from 0. 2-wide slices CuAr lamps and flux-calibrated using an observation of the along the major and minor axes of the bar are shown in Fig- standard star BD +28◦4211. The ten individual echelle orders ure 4. Aside from small deviations in the disk due to the spi- were combined into a single spectrum, weighted by the S/N ral arms, the major and minor axis brightness profiles are very in the overlap regions between orders. Several K-giant stars symmetric about the nucleus. were also observed on the same night during twilight for use Another issue to consider is the effect of inclination on the as velocity templates. disk surface brightness. One simple method that is commonly used to correct for inclination is to apply the relation µcorrected 3. MEASUREMENTS AND RESULTS = µobserved − 2.5C log(a/b), where a and b are the major and minor axis lengths of the disk, and C is a parameter whose 3.1. Imaging value ranges from 0 for an optically thick disk to 1 for the op- The WFPC2 image clearly reveals the morphologyof Malin tically thin case (de Jong 1996; Seigar & James 1998). From 1 at small angular scales where previousground-basedimages the GALFIT decomposition, the disk axis ratio b/a is 0.858, were unable to discern the details of its structure. The galaxy consistent with the mean ellipticity of ǫ =1−(b/a)=0.14 mea- is seen to have a compact bulge dominating the light profile sured by the ELLIPSE routine over the disk-dominated region out to r ≈ 1′′, a bar of length ∼ 6′′, and a nearly face-on disk from r = 4 to 6 arcsec. This corresponds to a surface bright- with a hint of spiral structure that can be traced out to ∼ 6 − ness correction of −0.17 mag in the optically thin case. Given 7′′. The disk morphology appears smooth and there are no the approximate nature of this method and the small magni- obvious dust lanes, knots, or star-forming complexes. Apart tude of the correction, we choose not to apply this correction from the remarkable but nearly invisible outer LSB structure, to the measured quantities, but we consider it as a minor effect Malin 1 has the morphology of an SB0/a galaxy. that would tend to increase the central surface brightness bya To decompose the galaxy’s structure into its subcompo- small amount. nents, model fits to the WFPC2 image were performed using The structural parameters derived from the GALFIT model- the 2-dimensionalfitting package GALFIT (Peng et al. 2002). ing are listed in Table 1. While the formal fitting uncertainties The components used in the model fit included an exponential from GALFIT are negligibly small, the actual uncertainties on disk, a Sérsic-law bulge, and a bar. Since GALFIT does not the derived model parameters are predominantly systematic include bar models such as a Freeman (1966) bar or a flat bar due to real deviations of the galaxy componentsfrom the sim- (Prieto et al. 1997), the bar component was modeled using a ple fitting functions used by GALFIT and are therefore dif- Sérsic profile with the index n constrained to be ≤ 0.5 to ap- ficult to estimate. Trial GALFIT runs with slightly different proximate a bar with a flat core, and with this constraint the fitting models (i.e., without constraints on the bar Sérsic in- bar has n =0.5 in the best-fitting model. In addition, a central dex n or without including a point source component) yielded magnitude differences of order ∼ 0.2 mag for the bulge and 2 IRAF is distributed by the National Optical Astronomy Observatories, bar components relative to the best-fit model listed in Table which are operated by the Association of Universities for Research in As- tronomy, Inc., under cooperative agreement with the National Science Foun- 1, which gives some indication of the likely systematic uncer- dation. tainties in the decomposition. Different choices of bulge and NORMAL STELLAR DISK IN MALIN 1 3 bar models or omission of the nuclear point source led to rel- the bulge, bar, and normal disk of the galaxy. Since the sur- atively small changes in the exponential disk parameters, at face brightnessof the disk within several arcsecondsof the nu- the level of < 0.1 mag in total magnitude and < 5% in scale cleus is much fainter than the surface brightness of the bulge length. or bar, it is not surprisingthat it would be difficultto recognize the disk in ground-based imaging of average seeing. Interest- 3.2. Spectroscopy ingly, Bothun et al. (1987) note that the nucleus “appears to As previously shown by Impey & Bothun (1989), the bulge be surrounded by extended nebulosity of somewhat high sur- spectrum is consistent with a predominantly old stellar popu- face brightness”. In retrospect, this “nebulosity” must have lation. The stellar velocity dispersion of the bulge was mea- been starlight from the disk itself. sured from the Keck spectrum by direct fitting of broadened Recent work by Aguerri et al. (2005) on the I-band photo- and diluted stellar spectra in the wavelength domain, follow- metric properties of SB0 galaxies provides an ideal compari- ing the methods described by Barth et al. (2002). The fit son sample to examine the structure of Malin 1 in the context was performed over the rest wavelength range 5210–5480 Å, of galaxies of similar Hubble type. Based on a sample of 14 which contains Fe λ5270 and several other reasonably strong nearby galaxies, they found that SB0 galaxies generally have features. The Mgb lines were excluded from the fit since the nearly exponential bulge profiles with n =1.48 ± 0.16 and re galaxy spectrum has a lower Fe/α abundance ratio than typical in the range 0.3 to 1.0 kpc with a mean of 0.6 kpc, so Malin nearby K giant stars, as is generally found for high-dispersion 1’s bulge properties are typical for its Hubble type. With a I bulge absolute magnitude of MI = −20.9 mag and σ =196 km galaxies. Weak [N ] λ5200 emission appears to be present − as well. Fits were performed using nine template stars with s 1, Malin 1 also falls within the normal range of SB0 galax- spectral types between G8 III and K3 III were. The best fit ies in the Faber-Jackson relation (see Fig. 5 of Aguerri et al. was found with a K0 star, giving σ = 196 ± 15km s−1, where 2005). the final uncertainty is the sum in quadrature of the fitting The disk of Malin 1 also has properties similar to the SB0 uncertainty from the best-fitting template (10 km s−1) and the sample. Figure 7 plots the central I-band surface brightness standard deviation of the results from fitting all nine templates against disk scale length. While Malin 1 falls toward the (11 km s−1). Figure 6 shows the fit of the broadened KO star lower end of the sample in terms of its surface brightness, to the spectrum of Malin 1 over this region. it is still consistent with the normal range for this Hubble The velocity dispersion fitting code was also used to gener- type. The only aspect of Malin 1’s structure that appears ate starlight-subtracted spectra in the regions surrounding the different from the SB0 sample is the ratio of bulge to disk Hβ and Hα emission lines. Using the best-fitting K-type stel- radius. Aguerri et al. (2005) find a surprisingly tight cou- lar template and allowing for a featureless continuum dilution pling with re/h = 0.20 ± 0.01 for their sample, while Ma- yielded an adequate continuum subtraction over these regions, lin 1 has re/h =0.13. A portion of this difference could be giving a pure emission-line spectrum (Figure 5). the result of the different bar model adopted for the GAL- Reddening-corrected emission line flux ratios measured FIT analysis, however, since Aguerri et al. (2005) used Free- from the nuclear spectrum are [O II] λ3727 / [O III] λ5007 man bar or flat bar models in their 1-dimensional radial pro- = 3.1, [O III] λ5007 / Hβ = 1.9, [O I] λ6300 / Hα = 0.3, and file fits, and different assumptions for the bar profile will di- [N II] λ6584 / Hα = 0.85. These measurements are consis- rectly affect the flux inferred to arise from the outer part of tent with a LINER classification (Heckman 1980; Ho et al. the bulge. In any case, a bulge-to-disk scale length ratio of 1997a). Impey & Bothun (1989) had previously found [O III] 0.13 is well within the normal range for low or high surface /Hβ = 4.8 and classified Malin 1 as a high-excitation Seyfert brightness spiral galaxies of a range of Hubble types (de Jong galaxy; presumably this resulted from not having performed 1996; Seigar & James 1998; Beijersbergen et al. 1999). Over- a starlight subtraction and therefore missing most of the Hβ all, these results indicate that out to r ≈ 10 kpc, the optical flux. Approximately 30% of S0–Sa galaxies contain LINER structure of Malin 1 does not deviate very significantly from nuclei (Ho et al. 1997b) so Malin 1 is not unusual in exhibit- the general population of early-type barred galaxies. ing this type of activity. There is at best weak evidence for The low-surface brightness structure at large radii appears a broad component to the Hα emission line. Model fits in- to be a photometrically distinct component of the galaxy’s cluding 3 Gaussian components for the three narrow emission structure rather than a smooth extension of the normal inner lines are somewhat improved by the addition of a broad com- disk. For a disk scale length of 4.8 kpc, the extrapolated sur- ± face brightness of the inner disk at r =50′′ (equivalent to ∼ 75 ponent (having a best-fitting width of FWHM = 2010 80 − km s−1), but the relatively low S/N of the spectrum precludes kpc) would be undetectably faint: 36.1 mag arcsec 2 in the B any definitive conclusions regarding the possible presence of band, which is far fainter than the actual outer disk brightness −2 a broad-line component. of µV ≈ 27 mag arcsec at this radius (Bothun et al. 1987). There may be a break in the exponential light profile of the 4. DISCUSSION disk at radii greater than ∼ 7′′, but deeper images would be The measurements described above indicate a very differ- needed to trace the surface brightness profile in the transition ent structure for Malin 1 than that which has previously been region between the inner disk and the extended outer disk. found. Bothun et al. (1987) fit the galaxy profile with a model Given the presence of this newly identified disk in Malin 1, consisting of an r1/4-law bulge and an exponential disk, and should it still be considered an LSB galaxy? LSB galaxies are ′′ ′′ determined that the bulge had re = 2. 9 ± 0. 5, which cor- generally classified as such based on the extrapolated central responds to 4.4 kpc for the cosmology assumed in this pa- surface brightness of their disk components. A crucial point per. Pickering et al. (1997) found similar results from newer is that the LSB classification applies only to the disk com- ground-based imaging data. The HST image shows that the ponent. Giant LSB disk galaxies have high surface bright- bulge radius is smaller than this value by nearly an order of ness bulges that dominate the light of the central regions (e.g., magnitude, and it seems likely that the bulge region described Sprayberry et al. 1995). Thus, the total light profiles of giant by Bothun et al. (1987) was actually the combined light from LSB galaxies apparently always have high surface brightness 4 BARTH central regions, and a radial profile decomposition is required Bothun et al. (1987) are far outside the ranges found for nor- in order to determine whether a galaxy should be classified mal spirals or even for other giant LSB galaxies. In the giant as having a giant LSB disk or not. Clearly, to perform such LSB sample of Sprayberry et al. (1995), for example, the disk classifications in a meaningful way, observations with suffi- scale lengths range from 5 to 13 kpc and the faintest disk has −2 cient spatial resolution to decompose the bulge and inner disk µB(0) = 24.2 mag arcsec . Clearly, however, when the nor- components accurately are required. mal disk of Malin 1 is compared with the disk properties of While there is perhaps no universal agreement on the sur- other galaxies in these diagrams, it is no longer an outlier. face brightness threshold for a galaxy disk to be considered an It might be argued that the extended outer structure in Ma- LSB disk, a common criterion is a central surface brightness lin 1 still has the form of a disk with a very low extrapolated −2 fainter than µB = 23.0 mag arcsec (Impey & Bothun 1997). central surface brightness and large scale length and that this As noted by Bothun et al. (1997), a galaxy with a central sur- extended disk remains an outlier relative to all other known face brightness fainter than this level would represent a > 4σ galaxies. However, it seems more reasonable and consistent deviation from the distribution of surface brightness found by to compare the normal, central disk of Malin 1 with those of Freeman (1970) and would therefore be a very unusual out- other galaxies, and to consider its extended outer disk in the lier if the actual distribution of disk surface brightnesses were contextofthe extremeouterdisks ofother nearby spiral galax- described by Freeman’s law. To compare the disk properties ies, which sometimes display structures similar to (albeit less of Malin 1 with this criterion, the B magnitude of the Malin 1 extended than) that of Malin 1. disk was estimated by performingsynthetic photometryon the Recent work has revealed a great deal of detail and a S0 template spectrum from Kinney et al. (1996). The (B − I) diversity of properties among the extreme outer disk re- color index for this spectrum is 2.2 mag. Assuming this (B−I) gions of some nearby spiral galaxies. Deep narrow-band ob- color for the disk of Malin 1, it has an estimated µ0(B)=22.3 servations of some late-type spirals have detected H II re- mag arcsec−2, which would not qualify it as an LSB disk3. In gions far beyond the galaxies’ optical radii (Ferguson et al. the classification system of McGaugh (1996), Malin 1 would 1998b; Lelièvre & Roy 2000), and GALEX observations of have an “intermediate surface brightness” disk. the nearby spiral M83 have detected UV-bright stars in the Another related criterion that has been applied to distin- outer disk at distances of greater than 20 kpc from the galaxy guish low and high surface brightness galaxies is a “dif- center (Thilker et al. 2005). In these examples, the regions of fuseness index” for the disk, which combines central surface recent star formation have filamentary, spiral-like structures brightness and scale length into a single parameter. The dif- with an appearance similar to the outer disk regions of Malin fuseness index criterion given by Sprayberry et al. (1995) is 1. Ferguson et al. (1998a) noted that the distant outer regions µB(0) + 5log(h) > 27.0 for a galaxy to be classified as a gi- of spiral disks have physical properties (such as low column ant LSB disk. Based on the GALFIT results and the as- densities of gas and high gas mass to stellar mass ratios) that sumed (B − I) color index of 2.2 mag, the disk in Malin 1 has are very similar to the inferred properties of Malin 1. It is µB(0)+5log(h)=25.0, which lies well inside the high surface now apparent that this similarity is not merely a coincidence– brightnessrange for this parameter. Thus, based on either cen- the parameters previously found for Malin 1 were precisely tral surface brightness or diffuseness, the disk of Malin 1 out those of its extreme outer disk region. Extended outer disk to r ≈ 10 kpc should not be considered an LSB disk according structure has recently been found out to r ≈ 40 kpcin M31 as to conventional criteria. well (Ferguson et al. 2002; Ibata et al. 2005), but in the case Malin 1 still might be considered an LSB galaxy based on of M31 the extended outer structure appears to be a smooth the average surface brightness over its entire disk area, how- continuation of the exponential profile of the main disk, in ever. According to Pickering et al. (1997), the galaxy’s total contrast to the photometrically decoupled outer structure of luminosity is MV = −22.9 ± 0.4mag. The HST measurements Malin 1. The stellar disk of the late-type spiral NGC 300 yield MI = −22.4 mag for the combined bulge, bar, and nor- also follows a single exponential profile from its inner re- mal disk components. Assuming (V − I)=1.3 mag for an S0 gions to the outer disk at distances up to 10 scale lengths galaxy as determined from SYNPHOT, this corresponds to (Bland-Hawthorn et al. 2005). MV = −21.1 mag. Thus, the comparison between the HST and New observations of early-type barred galaxies ground-based measurements indicates that the majority of the have revealed surprising outer disk structures as well. galaxy’s total optical luminosity lies in the extended disk re- Erwin, Beckman, & Pohlen (2005) have found that at least gion beyond the normal disk. It would be useful to confirm 25% of SB0-SBb galaxies have photometric profiles charac- this result with new ground-based imaging data and a sin- terized by an outer exponential disk component with a larger gle profile decomposition over the entire radial range of the scale length than the inner, main disk component. The transi- galaxy; the HST image could be used to exclude background tion from the inner to outer disk occurs at 3–6 scale lengths galaxies and improve the measurement accuracy for the outer of the inner disk and at surface brightnesses of 22.6–25.6 disk brightness profile. mag arcsec−2, and spiral structure is often seen in the outer In comparison with all other known disk galaxies, Malin 1 disk region. Erwin et al. call such structures “antitruncated” has always appeared as a unique and distant outlier in plots disks. Malin 1 appears to fit within this category, although of central surface brightness against disk scale length (see, deeper images would be needed to determine the structure of e.g., Bothun et al. 1987; Sprayberry et al. 1995; Bothun et al. the transition region between the inner and outer disks. 1997; Dalcantonetal. 1997; vandenBosch 1998). The Simulations by Peñarrubia et al. (2006) have explored one previously measured disk scale length and central surface possible mechanism that could form such highly extended −2 brightness (h ≈ 70 kpc; µ0(V) = 25.5 mag arcsec ) from outer disks, by the tidal shredding of dwarf galaxies in the outskirts of an M31-sized galaxy. With an initial apocenter of 3 For comparison, Fukugita et al. (1995) find an average (B−I) color of 2.0 r = 75 kpc, the dwarf satellites were tidally distorted into ex- mag for S0 galaxies; using this value instead of the SYNPHOT color would −2 tended disks having spiral-like filamentary structure, extend- yield µ0(B) = 22.1 mag arcsec . ing to over 100 kpc from the center of the larger galaxy. The NORMAL STELLAR DISK IN MALIN 1 5 tidal debris tends to evolve into a nearly exponential profile, a normal early-type barred spiral galaxy, and the central sur- with scale lengths of up to 50 kpc, dependingon the compact- face brightness of its disk is ∼ 4 magnitudes brighter than ness of the original stellar distribution in the dwarf galaxy. the value originally found by Bothun et al. (1987). Thus, it Could other giant LSB galaxies possess inner, high surface seems reasonable to consider Malin 1 not as an LSB galaxy, brightness disks that have gone undetected? Several of the but rather as a galaxy with the normal structural components other known giant LSB galaxies lie at distances comparableto of an early-type barred spiral, which is embedded in a remark- that of Malin 1 (e.g., Sprayberry et al. 1995), and inadequate ably extended, optically faint, and gas-rich outer structure be- spatial resolution could cause a normal disk to be missed in a yond its normal disk. bulge-disk decomposition. This would be a natural pitfall of Bothun et al. (1997) remark that giant LSB galaxies such as fitting a simple 2-component model to a galaxy with a bright Malin 1 are “truly enigmatic in that ’normal’ formation pro- central region (including a bulge and disk) and a very diffuse cesses were at work to create the bulge componentbut no con- and extended outer disk that was brighter than a simple ex- spicuous stellar disk ever formed around this bulge.” Malin 1 tension of the inner disk profile at large radii. The outer disk is perhaps a somewhat less enigmatic galaxy once its normal would dominate the fit at large radii and therefore determine disk has been recognized. the scale length and central surface brightness of the expo- nential disk component in the fit, and the inner, high surface brightness disk, if present, would be subsumed in the outer I am grateful to Joe Shields and Karen O’Neil for discus- part of the bulge profile. Deep, high-resolution imaging of a sions that motivated this work and to Luis Ho, Marc Seigar, representative sample of giant LSB galaxies would be useful Peter Erwin, Chien Peng, and the anonymous referee for to clarify the nature of these objects. many helpful suggestions. Support for the Keck observations was generously provided by the UC Irvine Physical Sciences 5. CONCLUSIONS Innovation Fund. The author wishes to recognize and ac- The stellar disk of Malin 1 has long been thought to be an knowledge the very significant cultural role and reverence that extreme and distant outlier compared to the disk properties the summit of Mauna Kea has always had within the indige- of all other known spiral galaxies. The archival HST data nous Hawaiian community. We are most fortunate to have the demonstrates that the central structure of Malin 1 is that of opportunity to conduct observations from this mountain.

REFERENCES Aguerri, J. A. L., Elias-Rosa, N., Corsini, E. M., & Muñoz-Tuñón, C. 2005, Jedrzejewski, R. I. 1987, MNRAS, 226, 747 A&A, 434, 109 Kinney, A. L., Calzetti, D., Bohlin, R. C., McQuade, K., Storchi-Bergmann, Barth, A. J., Ho, L. C., & Sargent, W. L. W. 2002, AJ, 124, 2607 T., & Schmitt, H. R. 1996, ApJ, 467, 38 Beijersbergen, M., de Blok, W. J. G., & van der Hulst, J. M. 1999, A&A, 351, Krist, J. 1993, in ASP Conf. Ser. 52: Astronomical Data Analysis Software 903 and Systems II, ed. R. J. Hanisch, R. J. V. Brissenden, & J. Barnes, 52, 536 Bland-Hawthorn, J., Vlajic,´ M., Freeman, K. C., & Draine, B. T. 2005, ApJ, Lelièvre, M., & Roy, J.-R. 2000, AJ, 120, 1306 629, 239 Matthews, L. D., van Driel, W., & Monnier-Ragaigne, D. 2001, A&A, 365, 1 Bothun, G. D., Impey, C. D., Malin, D. F., & Mould, J. R. 1987, AJ, 94, 23 McGaugh, S. S. 1996, MNRAS, 280, 337 Bothun, G., Impey, C., & McGaugh, S. 1997, PASP, 109, 745 McGaugh, S. S., & Bothun, G. D. 1994, AJ, 107, 530 Bothun, G. D., Schombert, J. M., Impey, C. D., & Schneider, S. E. 1990, ApJ, Minchin, R. F., et al. 2004, MNRAS, 355, 1303 360, 427 Noguchi, M. 2001, MNRAS, 328, 353 Dalcanton, J. J., Spergel, D. N., & Summers, F. J. 1997, ApJ, 482, 659 O’Neil, K., Bothun, G. D., & Impey, C. D. 2000, ApJS, 128, 99 de Jong, R. S. 1996, A&A, 313, 45 Peñarrubia, J., McConnachie, A., & Babul, A. 2006, ApJ, 650, L33 Erwin, P., Beckman, J. E., & Pohlen, M. 2005, ApJ, 626, L81 Peng, C. Y., Ho, L. C., Impey, C. D., & Rix, H. 2002, AJ, 124, 266 Ferguson, A. M. N., Gallagher, J. S., & Wyse, R. F. G. 1998a, AJ, 116, 673 Pickering, T. E., Impey, C. D., van Gorkom, J. H., & Bothun, G. D. 1997, AJ, Ferguson, A. M. N., Wyse, R. F. G., Gallagher, J. S., & Hunter, D. A. 1998b, 114, 1858 ApJ, 506, L19 Prieto, M., Gottesman, S. T., Aguerri, J.-A. L., & Varela, A.-M. 1997, AJ, Ferguson, A. M. N., Irwin, M. J., Ibata, R. A., Lewis, G. F., & Tanvir, N. R. 114, 1413 2002, AJ, 124, 1452 Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525 Freeman, K. C. 1966, MNRAS, 134, 15 Seigar, M. S., & James, P. A. 1998, MNRAS, 299, 672 Freeman, K. C. 1970, ApJ, 160, 811 Sheinis, A. I., Bolte, M., Epps, H. W., Kibrick, R. I., Miller, J. S., Radovan, Fukugita, M., Shimasaku, K., & Ichikawa, T. 1995, PASP, 107, 945 M. V., Bigelow, B. C., & Sutin, B. M. 2002, PASP, 114, 851 Heckman, T. M. 1980, A&A, 87, 152 Sprayberry, D., Impey, C. D., Irwin, M. J., McMahon, R. G., & Bothun, G. Ho, L. C., Filippenko, A. V., & Sargent, W. L. W. 1997a, ApJS, 112, 315 D. 1993, ApJ, 417, 114 Ho, L. C., Filippenko, A. V., & Sargent, W. L. W. 1997b, ApJ, 487, 568 Sprayberry, D., Impey, C. D., Bothun, G. d., & Irwin, M. J. 1995, AJ, 109, Hoffman, Y., Silk, J., & Wyse, R. F. G. 1992, ApJ, 388, L13 558 Ibata, R., Chapman, S., Ferguson, A. M. N., Lewis, G., Irwin, M., & Tanvir, Thilker, D. A., et al. 2005, ApJ, 619, L79 N. 2005, astro-ph/0504164 van den Bosch, F. C. 1998, ApJ, 507, 601 Impey, C. D., & Bothun, G. 1989, ApJ, 341, 89 Impey, C., & Bothun, G. 1997, ARA&A, 35, 267 6 BARTH

TABLE 1 STRUCTURAL PROPERTIESOF MALIN 1.

Parameter Value

Bulge re 0.6 kpc Bulge Sérsic index n 1.24 Bulge I magnitude 17.0 mag Bar I magnitude 17.6 mag Point source I magnitude 20.6 mag −2 Disk µ0(I) 20.1 mag arcsec Disk scale length h 4.8 kpc Bulge velocity dispersion 196 ± 15 km s−1 NOTE. — Properties listed here are determined from the GALFIT decomposition of the WFPC2 image and from the Keck spectrum. The magnitudes listed above include a correction for Galactic ex- tinction of 0.067 mag. Additionally, a K-correction of 0.08 mag and a correction for cosmological surface brightness dimming have been applied to the value of the disk central surface brightness µ0(I). NORMAL STELLAR DISK IN MALIN 1 7

FIG. 1.— Inner portion of the HST WFPC2 F814W mosaic of Malin 1, displayed using a logarithmic stretch, showing the bulge, bar, and disk components. The faint noisy vertical and horizontal lines are the boundaries between the WFPC2 CCDs. 8 BARTH

′′ ′′ FIG. 2.— GALFIT decomposition of Malin 1. Left panel: A 20 × 20 portion of the WFPC2 image. Middle panel: GALFIT model including bulge, bar, and disk components. Right panel: Residuals after subtraction of galaxy model. NORMAL STELLAR DISK IN MALIN 1 9

FIG. 3.— I-band surface brightness profile for the inner region of Malin 1 (crosses), measured with the IRAF ELLIPSE task. Model components are the central point source (dot-dashed curve), bulge (dotted), bar (short dashed), and disk (long-dashed), with the total galaxy model plotted as a solid line. The lower panel displays the ellipticity profile for the galaxy and for the GALFIT model. 10 BARTH

FIG. 4.— Surface brightness cuts along the major and minor axes of the bar (upper and lower panels, repsectively). Crosses represent measured surface brightness averaged over a 2 pixel-wide extraction through the image. Model components are the central point source (dot-dashed curve), bulge (dotted), bar (short dashed), and disk (long-dashed), with the total galaxy model plotted as a solid line. NORMAL STELLAR DISK IN MALIN 1 11

FIG. 5.— Keck ESI spectrum of Malin 1. Upper panel: A portion of the complete ESI spectrum. Lower panels: The starlight-subtracted spectrum in the regions around the Hβ and Hα emission lines. 12 BARTH

FIG. 6.— Best fit of a broadened K-giant template star, HD 92068 (K0 III), to the spectrum of Malin 1. NORMAL STELLAR DISK IN MALIN 1 13

FIG. 7.— I-band central surface brightness vs. scale length for disk components in SB0 galaxies. Open squares are from Aguerri et al. (2005). Malin 1 is represented by an × symbol.