A Deep Study of the Dwarf Satellites Andromeda XXVIII & Andromeda XXIX
Colin T. Slater, Eric F. Bell, Nicolas F. Martin, Erik J. Tollerud, Nhung Ho
PPreprint typeset using L A TEX style emulateapj v. 5/2/11
A DEEP STUDY OF THE DWARF SATELLITES ANDROMEDA XXVIII & ANDROMEDA XXIX
Colin T. Slater and Eric F. Bell
Department of Astronomy, University of Michigan, 1085 S. University Ave., Ann Arbor, MI 48109, USA
Nicolas F. Martin
Observatoire astronomique de Strasbourg, Universit´e de Strasbourg, CNRS, UMR 7550, 11 rue de l’Universit´e, F-67000 Strasbourg,France
Erik J. Tollerud and Nhung Ho
Astronomy Department, Yale University, P.O. Box 208101, New Haven, CT 06510, USA
ABSTRACTWe present the results of a deep study of the isolated dwarf galaxies Andromeda XXVIII andAndromeda XXIX with Gemini/GMOS and Keck/DEIMOS. Both galaxies are shown to host old,metal-poor stellar populations with no detectable recent star formation, conclusively identifying bothof them as dwarf spheroidal galaxies (dSphs). And XXVIII exhibits a complex horizontal branch mor-phology, which is suggestive of metallicity enrichment and thus an extended period of star formationin the past. Decomposing the horizontal branch into blue (metal poor, assumed to be older) and red(relatively more metal rich, assumed to be younger) populations shows that the metal rich are alsomore spatially concentrated in the center of the galaxy. We use spectroscopic measurements of theCalcium triplet, combined with the improved precision of the Gemini photometry, to measure themetallicity of the galaxies, confirming the metallicity spread and showing that they both lie on theluminosity-metallicity relation for dwarf satellites. Taken together, the galaxies exhibit largely typicalproperties for dSphs despite their significant distances from M31. These dwarfs thus place particularlysignificant constraints on models of dSph formation involving environmental processes such as tidalor ram pressure stripping. Such models must be able to completely transform the two galaxies intodSphs in no more than two pericentric passages around M31, while maintaining a significant stellarpopulations gradient. Reproducing these features is a prime requirement for models of dSph forma-tion to demonstrate not just the plausibility of environmental transformation but the capability ofaccurately recreating real dSphs.
Subject headings: galaxies: dwarf — galaxies: individual (And XXVIII, And XXIX) — Local Group INTRODUCTION
The unique physical properties and environments ofdwarf galaxies make them excellent test cases for improv-ing our understanding of the processes that affect thestructure, stellar populations, and evolution of galaxies.Because of their shallow potential wells, dwarf galaxiesare particularly sensitive to a wide range of processes thatmay only weakly affect larger galaxies. These processesrange from cosmological scales, such as heating by theUV background radiation (Gnedin 2000), to interactionsat galaxy scales such as tidal stripping and tidal stirring(Mayer et al. 2001; Klimentowski et al. 2009; Kravtsovet al. 2004), resonant stripping (D’Onghia et al. 2009),and ram pressure stripping (Mayer et al. 2006), to the ef-fects of feedback from from the dwarfs themselves (Dekel& Silk 1986; Mac Low & Ferrara 1999; Gnedin & Zhao2002; Sawala et al. 2010).Many studies have focused on understanding the differ-ences between the gas-rich, star forming dwarf irregulargalaxies (dIrrs) and the gas-poor, non-star-forming dwarfspheroidals. While a number of processes could suitablyrecreate the broad properties of this differentiation, find-ing observational evidence in support of any specific the-ory has been difficult. One of the main clues in thiseffort is the spatial distribution of dwarfs; while dIrrs can be found throughout the Local Group, dSphs princi-pally are only found within 200-300 kpc of a larger hostgalaxy such as the Milky Way or Andromeda (Einastoet al. 1974; van den Bergh 1994; Grebel et al. 2003).This is trend is also reflected in the gas content of LocalGroup dwarfs (Blitz & Robishaw 2000; Grcevich & Put-man 2009). This spatial dependence seems to indicatethat environmental effects such as tides and ram pres-sure stripping are likely to be responsible for creatingdSphs. However, there are outliers from this trend, suchas Cetus, Tucana, and Andromeda XV, which are dSphsthat lie more than 700 kpc from either the Milky Wayor Andromeda. The existence of such distant dSphs maysuggest that alternative channels for dSph formation ex-ist (Kazantzidis et al. 2011b), or it could be an incidentaleffect seen in galaxies that have passed through a largerhost on very radial orbits (Teyssier et al. 2012; Slater &Bell 2013).The set of isolated dwarf galaxies was recently enlargedby the discovery of Andromeda XXVIII and XXIX,which by their position on the sky were known to beapproximately 360 and 200 kpc from Andromeda, respec-tively (Slater et al. 2011; Bell et al. 2011). While AndXXIX was identified as a dSph by the images confirm-ing it as a galaxy, there was no comparable data on AndXXVIII (beyond the initial SDSS discovery data) with a r X i v : . [ a s t r o - ph . GA ] M a y Slater et al.which to identify it as a dSph or dIrr. We thus soughtto obtain deeper imaging of both galaxies down to thehorizontal branch level which would enable a conclusiveidentification of the galaxies as dSphs or dIrrs by con-straining any possible recent star formation. In addition,the deep photometry permits more precise determinationof the spatial structure and enables the interpretationof the spectroscopic Calcium triplet data from Tollerudet al. (2013) to obtain a metallicity measurement. Aswe will discuss, the information derived from these mea-surements along with dynamical considerations imposedby their position in the Local Group can together placesignificant constraints on plausible mechanisms for theorigin of these two dSphs.This work is organized as follows: we discuss the imag-ing data and the reduction process in Section 2, and illus-trate the general features of the color-magnitude diagramin Section 3. Spectroscopic metallicities are presented inSection 4, and the structure and stellar populations ofthe dwarfs are discussed in Section 5. We discuss the im-plications of these results for theories of dSph formationin Section 6. IMAGING OBSERVATIONS & DATA REDUCTION
Between 22 July 2012 and 13 August 2012 we ob-tained deep images of And XXVIII and XXIX with theGMOS instrument on Gemini-North (Gemini programGN-2012B-Q-40). The observations for each dwarf con-sisted of a total of 3150 seconds in SDSS-i band and 2925seconds in r, centered on the dwarf. Because the dwarfseach nearly fill the field of view of the instrument, wealso obtained a pair of flanking exposures for each dwarfto provide an “off-source” region for estimating the con-tamination from background sources. These exposuresconsisted of at least 1350 s in both r and i, though somefields received a small number of extra exposures. Theimages were all taken in 70th percentile image qualityconditions or better, which yielded excellent results withthe point source full width at half maximum ranging be-tween 0.47 (cid:48)(cid:48) and 0.8 (cid:48)(cid:48) .All of the images were bias subtracted, flat fielded, andcoadded using the standard bias frames and twilight flatsprovided by Gemini. The reduced images can be seen inFigure 1. Residual flat fielding and/or background sub-traction uncertainty exists at the 1% level (0.01 magni-tudes, roughly peak to valley). PSF photometry was per-formed using DAOPHOT (Stetson 1987), which enabledaccurate measurements even in the somewhat crowdedcenters of the dwarfs. In many cases the seeing in onefilter was much better than the other, such as for thecore of And XXVIII where the seeing was 0 . (cid:48)(cid:48) in iand 0 . (cid:48)(cid:48) in r. In these cases we chose to first detectand measure the position of stars in the image with thebest seeing, and then require the photometry of the otherband to reuse the positions of stars detected in the bet-ter band. This significantly extends our detection limit,which would otherwise be set by the shallower band, butwith limited color information at these faint magnitudes.The images were calibrated to measurements from theSloan Digital Sky Survey (SDSS), Data Release 9 (Ahnet al. 2012). For each stacked image we cross-matchedall objects from the SDSS catalog that overlapped ourfields, with colors between − . < ( r − i ) < .
6, andclassified as stars both by SDSS and DAOPHOT. Star- galaxy separation was performed using the “sharp” pa-rameter from DAOPHOT. From this we measured theweighted mean offset between the SDSS magnitudes andthe instrumental magnitudes to determine the zeropointfor each field. Between the saturation limit of the Gem-ini data, mitigated by taking several exposures, and faintlimits of the SDSS data (corresponding to approximately19 < i < . . < r < .
5) there were of order100 stars used for the calibration of each frame. Basedon the calculated stellar measurement uncertainties theformal uncertainty on the calibration is at the millimag-nitude level, but unaccounted systematic effects likelydominate the statistical uncertainty (e.g., precision red-dening measurements). All magnitudes were dereddenedwith the extinction values from Schlafly & Finkbeiner(2011).The photometric completeness of each stacked imagewas estimated by artificial star tests. For each field wetook the PSF used by DAOPHOT for that field and in-serted a large grid of artificial stars, with all of the starsat the same magnitude but with Poisson noise on theactual pixel values added to each image. This was per-formed for both r and i band images simultaneously, andthe resulting pair of images was then run through thesame automated DAOPHOT pipeline that was used onthe original image. Artificial stars were inserted over aa grid of i band magnitudes and r-i colors, producingmeasurements of the recovery rate that cover the entireCMD. The 50% completeness limit for both dwarfs is atleast r = 25 .
5, with slightly deeper data in the i-bandfor And XXVIII.The observed CMDs suffer from both foreground andbackground contamination. Foreground dwarf stars inthe Milky Way tend to contribute at the bright end ofthe CMD. At the faint end, distant galaxies that are toosmall to be resolved become the dominant source of con-tamination. This effect can quickly become significant atfainter magnitudes due to the rapid rise in the observedgalaxy luminosity function. This effect was minimized bythe superb seeing at the Gemini observatory, which al-lowed smaller galaxies to be resolved and excluded fromour sample. OBSERVED CMDS
The CMDs of And XXVIII and XXIX are shown inthe left panels of Figures 2 and 3, respectively. A 12Gyr old isochrone from Dotter et al. (2008) is overlaid atthe distances and spectroscopic metallicities determinedlater in this work. Both dwarfs show a well-populatedgiant branch with a very prominent red clump/red hor-izontal branch (RC/RHB) near r ∼ . .
0. Thisfeature is particularly clear as a large bump in the lumi-nosity functions of each dwarf, shown by the thick blackline in the right panels of Figure 2 and 3. In addition tothe RC/RHB, And XXVIII also shows a blue horizontalbranch (BHB) slightly fainter than r ∼ . − . < ( r − i ) < . r − i ) < . − σ upperlimit on the total HI mass of 2 . × M (cid:12) (T. Oosterloo,private communication). For comparison, the similarlylow-mass dwarf Leo T has had recent star formation andcontains ∼ . × M (cid:12) of H I (Ryan-Weber et al. 2008),while most dSphs have upper limits at this level or less(Grcevich & Putman 2009). This stringent limit on thegas in And XXVIII adds further evidence that it is adSph. Distance and Luminosity
The clear HB in both dwarfs enables an accurate mea-surement of the distance to the dwarfs, and hence theirdistance to M31. We fit a Gaussian plus a linear back-ground model to the r-band luminosity function of eachdwarf in the region of the HB, using only stars redderthan ( r − i ) = 0. The measured HB position is indi-cated in the right panels of Figure 2 and 3 by the hori-zontal arrow, and is m g, = 24 .
81 for And XXVIII and m g, = 24 .
84 for And XXIX. We use the RHB abso-lute magnitude calibration of Chen et al. (2009), whichis based on globular clusters RHBs measured directly inthe SDSS filter set. In the r-band this calibration, us-ing a linear metallicity dependence and without the ageterm, is M r = 0 . . . (1) Table 1
Properties of And XXVIII & XXIXParameter And XXVIII And XXIX α (J2000) 22 h m . s h m . s δ (J2000) 31 ◦ (cid:48) . (cid:48)(cid:48) ◦ (cid:48) . (cid:48)(cid:48) E(B-V) 0.080 a a Ellipticity 0 . ± .
02 0 . ± . ◦ ± ◦ ◦ ± ◦ r h . (cid:48) ± . (cid:48)
03 1 . (cid:48) ± . (cid:48) r h ±
20 pc 315 ±
15 pc D ±
48 kpc 829 ±
42 kpc( m − M ) . ± .
13 24 . ± . r M31b +18 − kpc 198 +18 − kpc M V − . ± . − . ± . (cid:104) [Fe / H] (cid:105) − . ± . − . ± . σ ([Fe / H]) 0 . ± .
15 0 . ± . < . × M (cid:12) a Schlafly & Finkbeiner (2011) b
3D distance, rather than projected.
The resulting distances are 811 ±
48 kpc for AndXXVIII and 829 ± . Based on thesedistances, both dwarfs lie well away from the plane ofsatellites from Conn et al. (2013) and Ibata et al. (2013).As seen from M31 the satellites are 80 ◦ (And XXVIII)and 60 ◦ (And XXIX) from the plane. The closest galaxyto And XXVIII is And XXXI at 164 kpc, while AndXXIX’s closest neighbor is And XIX at 88 kpc, makingboth relatively isolated from other dwarfs.We measured the total luminosity of both dwarfs bycomparing the portion of the LF brighter than the HBto the LF of the Draco dwarf. Using data from S´egall etal. (2007) we constructed a background-subtracted LFfor Draco inside r h , then scaled the LF of the dwarfssuch that they best matched the Draco LF. The result-ing luminosities are M V = − . ± . M V = − . ± . SPECTROSCOPIC METALLICITY
To complement the imaging data, we also make use ofmetallicities derived from Keck/DEIMOS spectroscopyof the brightest RGB stars. The source data and spectro-scopic reductions are described in Tollerud et al. (2013),and a sample spectrum is shown in Figure 4. We derivemetallicities from the λ ∼ The distance between And XXIX and M31 reported in Bell etal. (2011) was incorrect due to a geometry error; it is fixed in thiswork.
Slater et al.
And XXVIII And XXIX
Figure 1.
Stacked i-band image of And XXVIII on the left, and of And XXIX on the right. North is up, and East is to the left. Bothimages are approximately 5.6 (cid:48) on a side. The saturated feature near the center of And XXIX is a combination of a foreground star and twobackground galaxies. ( r − i ) r And XXVIII 0.3 0.0 0.3 0.6 ( r − i ) Background 40 80 120 160 N LF Figure 2.
CMD of And XXVIII on the left (inside 2 r h ), with the CMD of an equal-sized background region in the center. The red dashedline indicates the 50% completeness limit, while the vertical red line indicates the approximate division between red and blue horizontalbranches. The luminosity function of the dwarf is shown on the right, separated into a thick line showing stars with ( r − i ) > r − i ) <
0. A 12 Gyr old, [Fe/H] = -1.84 isochrone is overplotted, and the measured apparent magnitude of theHB is indicated with an arrow. (Section 2), these data can be calibrated to act as effec-tive proxies for [Fe / H] of these stars. For this purpose,we adopt the Carrera et al. (2013) metallicity calibra-tion to convert our photometry and equivalent widths to[Fe / H].A table of the spectroscopic metallicity measurementsof individual stars in each dwarf is presented in Table 2.We determine the uncertainty in the galaxy mean [Fe/H]by performing 1000 Monte Carlo resamplings of the dis-tribution. For each resampling, we add a random offsetto the metallicity of each star drawn from a Gaussianwith width of the per-star [Fe/H] uncertainty, and com-pute the mean of the resulting distribution. For mea-suring each galaxy’s metallicity spread σ ([Fe / H]), we re-port the second moment of the individual measurementdistribution and derive uncertainties from a resamplingprocedure like that for the galaxy mean [Fe/H]. The resulting metallicity distributions for And XXVIIIand XXIX are shown as cumulative distribution func-tions in Figure 5. From this it is immediately clear that,while the number of stars are relatively small, the medianof the distribution peaks at [Fe / H] ∼ − − . ± .
3, but at higher precision. STRUCTURE & STELLAR POPULATIONS wo Satellites of Andromeda 5 ( r − i ) r And XXIX 0.3 0.0 0.3 0.6 ( r − i ) Background 30 60 90 120 N LF Figure 3.
The same panels as Figure 2, but for And XXIX. A 12 Gyr old, [Fe/H] = -1.92 isochrone is overplotted. As with And XXVIIIthere are no indications of recent star formation. Though there may be some hints of a BHB, if it does exist it is substantially less prominentthan in And XXVIII.
Table 2
Metallicities of And XXVIII & XXIX StarsGalaxy RA (deg) Dec (deg) r ( r − i ) [Fe/H] σ [Fe / H]a
And XXVIII 338.16549 31.20840 21.397 0.53 -2.29 0.5And XXVIII 338.18538 31.22404 21.404 0.44 -1.58 0.5And XXVIII 338.15561 31.18421 21.381 0.45 -1.54 0.8And XXVIII 338.14847 31.15615 21.001 0.24 -0.50 0.8And XXVIII 338.17702 31.21802 21.548 0.42 -2.06 0.5And XXVIII 338.17499 31.22058 21.969 0.35 -2.91 0.7And XXVIII 338.18206 31.21668 21.509 0.41 -2.74 0.4And XXVIII 338.16849 31.22444 22.344 0.42 -1.90 0.6And XXVIII 338.18357 31.21526 21.332 0.38 -1.81 0.4And XXVIII 338.17542 31.23720 21.57 0.70 -1.28 0.3And XXVIII 338.15091 31.20916 21.578 0.46 -1.68 0.2And XXVIII 338.18428 31.23235 21.861 0.37 -1.15 0.3And XXVIII 338.22622 31.21862 21.78 0.37 -2.57 0.2And XXIX 359.73912 30.74974 22.113 0.36 -1.83 0.5And XXIX 359.72546 30.74484 21.467 0.45 -1.94 0.3And XXIX 359.72690 30.76834 21.592 0.44 -1.29 0.4And XXIX 359.74259 30.75986 22.084 0.42 -0.62 0.5And XXIX 359.74561 30.75100 21.854 0.39 -2.40 0.4And XXIX 359.71503 30.74976 21.369 0.40 -2.54 0.3And XXIX 359.71755 30.74150 21.968 0.37 -3.14 0.5And XXIX 359.71880 30.73644 22.003 0.40 -1.36 0.4And XXIX 359.71957 30.76735 22.211 0.35 -1.86 0.6And XXIX 359.75409 30.76225 21.172 0.45 -1.97 0.3And XXIX 359.75959 30.76464 22.111 0.36 -1.77 0.6And XXIX 359.73776 30.80015 21.266 0.20 -2.31 0.3And XXIX 359.73609 30.79734 22.137 0.33 -1.68 0.5And XXIX 359.68681 30.72895 21.959 0.36 -2.68 0.6And XXIX 359.74074 30.76867 21.407 0.44 -1.89 0.4And XXIX 359.74687 30.76948 21.751 0.29 -1.47 0.5And XXIX 359.75467 30.75391 21.752 0.37 -1.63 0.5 a Individual star [Fe/H] uncertainity; not to be confused with the overall metallicity spread in Table 1.
Slater et al. λ [ ] F l u x [ r e l a t i v e t o m e a n o f s p e c t r u m ] Figure 4.
An example spectrum of an individual star in AndXXVIII, focusing on the triplet of Calcium lines (marked with ver-tical dashed lines, and shifted to the velocity of And XXVIII). Thedotted line indicates the RMS uncertainty at each point in thespectrum. [Fe / H] f s t a r s ( < [ F e / H ] ) And XXIX (n=17)And XXVIII (n=13)
Figure 5.
Cumulative distribution of [Fe / H] for And XXVIII(blue solid line) and XXIX (red dotted line).
We determined the structural properties of the dwarfsusing an updated version of the maximum likelihoodmethod presented in Martin et al. (2008). This methodfits an exponential radial density profile to the galaxieswithout requiring the data to be binned, which enablesmore precise measurements of the structure in galax-ies with only a small number of observed stars. Theupdated version samples the parameter space with aMarkov Chain Monte Carlo process, and can more easilyaccount for missing data (Martin et al. 2015, submit-ted.) This is necessary to account for the limited field ofview of GMOS, which could cause a systematic size error(Mu˜noz et al. 2012), as well as the very center of AndXXIX where an inconveniently-located bright foregroundstar contaminates the very center of the image and pre-vents reliable photometry in the surrounding region.The resulting radial profiles and posterior probabil-ity distributions are shown in Figures 7 and 8. Thehalf-light radii and ellipticities all have fairly typical val-ues for other dwarfs of similar luminosities (Brasseur etal. 2011). The results are also consistent with the pa-rameters estimated from the much shallower SDSS data M V [ F e / H ] MW satellitesM31 satellitesThis work L V /L fl Figure 6.
Luminosity-metallicity relation for Local Group satel-lites, adapted from the compilation presented in Ho et al. (2015),which includes data from Kirby et al. (2011) and Collins et al.(2013). Squares are MW satellites, diamonds are M31 satellites,and the error bars are from the Monte Carlo resampling of the[Fe/H] distribution for each galaxy. And XXVIII and XXIX areshown as the larger red diamonds. This demonstrates that AndXXVIII and XXIX lie on the same metallicity-luminosity relationas other Local Group satellites.
Figure 7.
Posterior probability distributions for the structuralparameters fit for And XXVIII are shown on the left. From top-left to bottom-right, these show these correspond to the ellipticity( (cid:15) ), the position angle from north to east ( θ ), the number of starsunder the profile for the assumed depth limit ( N ∗ ), the angularmajor-axis half-light radius ( r h ), and its corresponding physicallength assuming the distance modulus measured above. The radialprofile is shown on the right, with the best fit exponential profileshown by the solid line and the dashed line showing the backgroundlevel. (Slater et al. 2011; Bell et al. 2011).The separation between the red and blue horizontalbranches in And XXVIII enables us to examine the spa-tial distribution of the metal-poor, older, and the moremetal-rich, younger, stellar populations. Radial profilesof the two horizontal branches (separated at ( r − i ) =0 .
0) are shown in Figure 9. The difference in the ra-dial profiles is easily seen in the right panel, and theposterior probability distributions for the half-light ra-dius confirm the statistical significance of the difference.This behavior has been seen in other dwarf galaxies, suchas Sculptor (Tolstoy et al. 2004), Fornax (Battaglia etal. 2006), Canes Venatici I (Ibata et al. 2006), And IIwo Satellites of Andromeda 7
Figure 8.
The same posterior probability distributions and radialprofile as in Figure 7, but for And XXIX. The two innermost radialprofile points (open circles) were not used in the fit due to the brightcontamination in the center of the galaxy. (McConnachie et al. 2007), and Leo T (de Jong et al.2008). In all of these cases the more metal-rich popu-lation is the more centrally concentrated one, consistentwith And XXVIII. Measuring the spatial structure of thetwo components independently shows that they appearto be simply scaled versions of each other; the half-lightradii are 370 ±
60 pc and 240 ±
15 pc (blue and red,respectively), while the ellipticities of 0 . ± .
06 and0 . ± .
03, along with position angles of 45 ◦ ± ◦ and34 ◦ ± ◦ , agree well with each other. Taken together thisimplies that the process that transformed the dwarf intoa pressure-supported system did so without randomizingthe orbital energies of individual stars enough to com-pletely redistribute the older and younger populations,but both populations did end up with the same generalmorphology.Simulations of isolated dwarfs by Kawata et al. (2006)are able to reproduce a radial metallicity gradient, butwith some uncertainty over the number of stars at thelowest metallicity values and the total luminosity of thesimulated dwarfs (and also see Revaz & Jablonka (2012)for simulated dwarfs without gradients). In these simula-tions the metallicity gradient is produced by the contin-uous accretion of gas to the center of the galaxy, whichtends to cause more metal enrichment and a youngerpopulation (weighted by mass) at small radii when com-pared to the outer regions of the galaxy. This explana-tion suggests that the “two populations” we infer fromthe RHB and BHB of And XXVIII are perhaps moreproperly interpreted as two distinct tracers of what is re-ally a continuous range of ages and metallicities presentin the dwarf. In this scenario, the lack of observed mul-tiple populations in And XXIX could be the result ofthe dwarf lacking sufficient gas accretion and star forma-tion activity to generate a strong metallicity gradient. Ifthis is the case, then there may be a mass dependence tothe presence of such gradients, which makes it particu-larly significant that And XXVIII is a relatively low-massgalaxy to host such a behavior. Whether this is merelystochasticity, or the influence of external forces, or if itrequires a more complex model of the enrichment processis an open question. DISCUSSION AND CONCLUSIONS
The analysis of And XXVIII and XXIX shows thatboth galaxies are relatively typical dwarf spheroidals,with old, metal-poor stellar populations and no measur-able ongoing or recent star formation. The significanceof these galaxies in distinguishing models of dSph forma-tion comes from their considerable distances from M31.If environment-independent processes such as supernovafeedback or reionization are responsible for transformingdIrrs into dSphs, then finding dSphs at these distancesis quite natural. However, such models are by them-selves largely unable to reproduce the radial dependenceof the dSph distribution around the Milky Way and M31.An environment-based transformation process, based onsome combination of tidal or ram pressure forces, canpotentially account for the radial distribution, but cor-rectly reproducing the properties of dSphs large radii isthe critical test of such models. It is in this light that An-dromeda XXVIII and Andromeda XXIX have the mostpower to discriminate between models.Models of tidal transformation have been studied ex-tensively and can account for many of the observed struc-tural properties of dSphs (Mayer et al. 2001; (cid:32)Lokas etal. 2010, 2012). However, a critical component of un-derstanding whether these models can reproduce the en-tire population of Local Group dSphs is the dependenceof the transformation process on orbital pericenter dis-tances and the number of pericentric passages. At largeradii the weaker tidal force may lose its ability completelytransform satellites into dSphs, potentially leaving ob-servable signatures in satellites on the outskirts of hostgalaxies.Observationally we cannot directly know the orbitalhistory of individual satellites without proper motions(of which there are very few), and must test the radialdistribution of dSphs in a statistical way. Slater & Bell(2013) used the Via Lactea simulations to show that asignificant fraction of the dwarf galaxies located between300 and 1000 kpc from their host galaxy have made atleast one pericentric passage near a larger galaxy. How-ever, the fraction of dwarfs that have undergone two ormore pericentric passages decreases sharply near 300 kpc.This suggests that it is unlikely for And XXVIII to haveundergone multiple pericentric passages.This presents a clear question for theories of dSph for-mation based on tidal interactions: can a dwarf galaxybe completely transformed into a dSph with only a sin-gle pericenter passage? Simulations of tidal stirring orig-inally seemed to indicate that the answer was no, andwhen dwarfs were placed on different orbits it was onlythe ones with several ( ∼ −
5) pericenter passages thatwere transformed into dSphs (Kazantzidis et al. 2011a).However, more recent simulations that used cored darkmatter profiles for the dwarfs suggest that multiple peri-center passages might not be required. Kazantzidis et al.(2013) show that dwarfs with very flat central dark mat-ter profiles (inner power-law slopes of 0.2) can be trans-formed into pressure supported systems after only oneor two pericenter passages. This result is encouraging,but it also comes with the consequence that cored darkmatter profiles also tend to make the dwarfs susceptibleto complete destruction by tidal forces. In the simula-tions of Kazantzidis et al. (2013), five out of the sevendwarfs that were successfully transformed into dSphs af-ter only one or two pericenter passages were subsequently Slater et al.
Figure 9.
Posterior probability distributions for the structural fit of And XXVIII, performed separately for stars in the RHB (red lines)and the BHB (blue lines). The difference in the radial profile clearly visible in the panel on the right, and the significance is confirmed bythe difference in half light radius ( r h ). The ellipticities and position angles are similar in the two populations. destroyed. Taken together, these results indicate thatrapid formation of a dSph is indeed plausible, but theremay only be a narrow range of structural and orbital pa-rameters compatible with such a process. Recent propermotion measurements of the dSph Leo I support thispicture even further, as it appears to have had only onepericentric passage (Sohn et al. 2013) yet is unambigu-ously a dSph.The properties of And XXVIII add an additional con-straint that any tidal transformation must not have beenso strong as to completely mix the older and youngerstellar populations. A simple test case of this problemhas been explored by (cid:32)Lokas et al. (2012), in which par-ticles were divided into two populations by their initialposition inside or outside of the half light radius. Thedwarfs were then placed on reasonable orbits around ahost galaxy, and evolved for 10 Gyr. The resulting radialprofiles of the two populations are distinct in nearly allcases, with some variation depending on the initial condi-tions of the orbit. These tests may be overly optimistic,since initial differentiation into two populations is per-formed by such a sharp radius cut, but the simulationsillustrate the plausibility of a dwarf retaining spatiallydistinct populations after tidal stirring.An additional piece of the puzzle is provided by themetallicities. And XXVIII and XXIX are both consis-tent with the luminosity-metallicity relation shown byother Local Group satellites (see Section 4). This im-plies that they could not have been subject to substan-tial tidal stripping , as this would drive them off this re-lation by lowering the luminosity without substantiallyaltering their metallicities. This point is further rein-forced by the similarity of the luminosity-metallicity re-lation of both dSph and dIrr galaxies in the Local Group(Kirby et al. 2013), making it unlikely that the mea-sured luminosity-metallicity relation itself is significantly altered by tidal stripping. Whether or not more gen-tle tidal effects can induce morphological transformationwithout altering the luminosity-metallicity relation re-mains to be seen.Taken together, the properties of And XXVIII andXXIX present a range of challenges for detailed mod-els of dwarf galaxy evolution to explain. Particularly forAnd XXVIII, the wide separation and low mass of thesystem add significant challenges to reproducing the gas-free spheroidal morphology with a stellar population gra-dient, while there may be similar challenges for explain-ing the apparent absence (or at least low-detectability)of such gradients in And XXIX. Though plausible ex-planations have been shown to exist for many of thesefeatures individually and under ideal conditions, whetherthe combination of these conditions can be accurately re-produced in a simulation is unknown. Further modelingof these types of systems is required before we can un-derstand the physical drivers of these observed features.We thank the anonymous referee for their helpful com-ments which improved the paper. This work was par-tially supported by NSF grant AST 1008342. Supportfor EJT was provided by NASA through Hubble Fellow-ship grant Facility: