1: \documentclass[12pt,preprint]{aastex}
2: %% You can insert a short comment on the title page using the command below.
3: \slugcomment{ApJ, in press}
4:
5: \shorttitle{An Andromedean Triplet}
6: \shortauthors{McConnachie et al.}
7:
8: \begin{document}
9:
10: \title{A trio of new Local Group galaxies with extreme properties}
11:
12: \author{Alan W. McConnachie$^1$\email{alan@uvic.ca}}
13: \author{Avon Huxor$^{2}$\email{huxor@roe.ac.uk}}
14: \author{Nicolas F. Martin$^{3}$\email{martin@mpia-hd.mpg.de}}
15: \author{Mike J. Irwin$^{4}$\email{mike@ast.cam.ac.uk}}
16: \author{Scott C. Chapman$^{4}$\email{schapman@ast.cam.ac.uk}}
17: \author{Gregory Fahlman$^{5}$\email{greg.fahlman@nrc-cnrc.gc.ca}}
18: \author{Annette M. N. Ferguson$^{2}$\email{ferguson@roe.ac.uk}}
19: \author{Rodrigo A. Ibata$^{6}$\email{ibata@astro.u-strasbg.fr}}
20: \author{Geraint F. Lewis$^{7}$\email{gfl@Physics.usyd.edu.au}}
21: \author{Harvey Richer$^{8}$\email{richer@phas.ubc.ca}}
22: \author{Nial R. Tanvir$^{9}$\email{nrt3@star.le.ac.uk}}
23:
24: \affil{$^1$Department of Physics and Astronomy, University of Victoria, Victoria, B.C., V8P 1A1, Canada}
25: \affil{$^2$Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, U.K.}
26: \affil{$^3$Max-Planck-Institut f{\"u}r Astronomie, K{\"o}nigstuhl 17, D-69117 Heidelberg, Germany}
27: \affil{$^4$Institute of Astronomy, Madingley Road, Cambridge, CB3 0HA, U.K.}
28: \affil{$^5$NRC Herzberg Institute of Astrophysics, 5071 West Saanich Road, Victoria, B.C., V9E 2E7, Canada}
29: \affil{$^6$Observatoire de Strasbourg, 11, rue de l'Universite, F-67000, Strasbourg, France}
30: \affil{$^7$Institute of Astronomy, School of Physics, A29, University of Sydney, NSW 2006, Australia}
31: \affil{$^8$Department of Physics and Astronomy, University of British Columbia, Vancouver, B.C., V6T 1Z1, Canada}
32: \affil{$^9$Department of Physics and Astronomy, University of Leicester, Leicester, LE1 7RH, U.K.\\}
33:
34: \begin{abstract}
35: We report on the discovery of three new dwarf galaxies in the Local
36: Group. These galaxies are found in new CFHT/MegaPrime $g,i$ imaging of
37: the south-western quadrant of M31, extending our extant survey area to
38: include the majority of the southern hemisphere of M31's halo out to
39: 150\,kpc. All these galaxies have stellar populations which appear
40: typical of dwarf spheroidal (dSph) systems. The first of these
41: galaxies, Andromeda~XVIII, is the most distant Local Group dwarf
42: discovered in recent years, at $\sim 1.4$\,Mpc from the Milky Way
43: ($\sim 600$\,kpc from M31). The second galaxy, Andromeda~XIX, a
44: satellite of M31, is the most extended dwarf galaxy known in the Local
45: Group, with a half-light radius of $r_h \sim 1.7$\,kpc. This is
46: approximately an order of magnitude larger than the typical half-light
47: radius of many Milky Way dSphs, and reinforces the difference in scale
48: sizes seen between the Milky Way and M31 dSphs (such that the M31
49: dwarfs are generally more extended than their Milky Way
50: counterparts). The third galaxy, Andromeda~XX, is one of the faintest
51: galaxies so far discovered in the vicinity of M31, with an absolute
52: magnitude of order $M_V \sim -6.3$. Andromeda~XVIII, XIX and XX
53: highlight different aspects of, and raise important questions
54: regarding, the formation and evolution of galaxies at the extreme
55: faint-end of the luminosity function. These findings indicate that we
56: have not yet sampled the full parameter space occupied by dwarf
57: galaxies, although this is an essential pre-requisite for successfully
58: and consistently linking these systems to the predicted cosmological
59: dark matter sub-structure.
60: \end{abstract}
61:
62: \keywords{surveys --- galaxies: dwarf --- Local Group --- galaxies: individual (Andromeda XVIII, Andromeda XIX, Andromeda XX)}
63:
64: \section{Introduction}
65:
66: Edwin Hubble first coined the term ``Local Group'' in his 1936 book
67: ``The Realm of the Nebulae'', to describe those galaxies that were
68: isolated in the general field but were in the vicinity of the
69: Galaxy. In recent years, the galaxies of the Local Group have been at
70: the focus of intense and broad-ranging research, from providing
71: laboratories for the investigation of dark matter properties (e.g.,
72: \citealt{gilmore2007} and references therein) to determinations of the
73: star formation history of the Universe (e.g., \citealt{skillman2005}
74: and references therein). Understanding individual galaxies in the Local
75: Group offers important contributions to galaxy structure and evolution
76: studies; understanding the properties of the population is central to
77: galaxy formation in a cosmological context.
78:
79: Hubble originally identified nine members of the Local Group: the
80: Galaxy and the Large and Small Magellanic Clouds; M31, M32 and NGC205;
81: M33, NGC6822 and IC1613; along with three possible members NGC6946,
82: IC10 and IC342. The distances of the latter three were highly
83: uncertain due to heavy extinction; IC10 has since been confirmed as a
84: member (\citealt{sakai1999}) although the other two lie outside the
85: Local Group (NGC6946; \citealt{sharina1997}; IC342:
86: \citealt{krismer1995}).
87:
88: The discovery of new Local Group members continued at a relatively
89: constant rate up to the start of 2004 (e.g.,
90: \citealt{ibata1994,whiting1997,whiting1999,
91: armandroff1998,armandroff1999,karachentsev1999}), at which point the
92: discovery rate has increased sharply. This has mostly been due to
93: large area photometric CCD-based surveys of the Milky Way and M31 stellar
94: haloes: by searching for overdensities of resolved stars in certain
95: regions of colour-magnitude space, it is possible to identify very
96: faint dwarf satellites which have previously eluded detection.
97:
98: Around the Milky Way, this technique has so far lead to the discovery
99: of 9 new satellites since 2005 (including possible diffuse star
100: clusters)
101: (\citealt{willman2005,willman2006,belokurov2006,belokurov2007,zucker2006b,walsh2007}). All
102: of these discoveries have been made using the Sloan Digitized Sky
103: Survey (SDSS). In addition, two new isolated dwarf galaxies have been
104: identified: Leo T, more than 400kpc from the Milky Way
105: (\citealt{irwin2007}), was discovered in the SDSS, and a revised
106: distance estimate for the previously known UGC4879 has moved this
107: galaxy from $> 10$\,Mpc to being placed on the periphery of the Local
108: Group (a scant $\sim 1.1$\,Mpc from the Milky Way;
109: \citealt{kopylov2008}).
110:
111: Around M31, 9 new dwarf galaxy satellites have been discovered since
112: 2004 (not including results presented herein). Two of these galaxies
113: (Andromeda IX, X) were found in special SDSS scans of M31
114: (\citealt{zucker2004a,zucker2007}) and one (Andromeda~XIV) was
115: discovered serendipitously by \cite{majewski2007} in Kitt Peak 4m imaging of fields in
116: the south-east halo of M31. The remaining new dwarf galaxies have been
117: discovered as part of our ongoing photometric survey of this galaxy
118: and its environs using the INT/WFC (Andromeda~XVII,
119: \citealt{irwin2008}) and CFHT/MegaPrime (Andromeda~XI, XII and XIII,
120: \citealt{martin2006}; Andromeda XV and XVI,
121: \citealt{ibata2007}). Despite its name, Andromeda~XVII is only the
122: fifteenth dwarf spheroidal satellite of M31 to be discovered;
123: Andromeda~IV is a background galaxy (\citealt{ferguson2000}) and
124: Andromeda~VIII was originally identified using planetary nebulae
125: (\citealt{morrison2003}) which were later shown to belong to M31 and
126: not to a separate entity (\citealt{merrett2006}). Additionally, only
127: thirteen of these dwarfs are actually located in the constellation of
128: Andromeda (Andromeda~VI $\equiv$ the Pegasus dSph; Andromeda~VII
129: $\equiv$ the Cassiopeia dSph).
130:
131: The unique, panoramic, perspective of the resolved stellar populations
132: of galaxies provided by Local Group members make them ideal targets
133: for observational programs aimed at understanding the detailed
134: structure of galaxies, their formation processes and their
135: evolutionary pathways. Dwarf galaxies are of particular interest,
136: given that they are thought to be the lowest mass, most dark matter
137: dominated systems which contain baryons (e.g.,
138: \citealt{mateo1998a}). They are therefore particularly sensitive
139: probes of external processes, such as tides and ram pressure stripping
140: (e.g., \citealt{mayer2006,mcconnachie2007c,penarrubia2008b}), and
141: internal processes such as feedback from star formation (e.g.,
142: \citealt{dekel1986,dekel2003}). Further, their potential as probes of
143: dark matter (e.g., \citealt{gilmore2007,strigari2007a}) and their
144: probable connection to cosmological sub-structures (e.g.,
145: \citealt{moore1999,bullock2000,kravtsov2004,penarrubia2008a}) give
146: them an importance to galaxy formation not at all in proportion to
147: their luminosity.
148:
149: Here we report on the discovery of three new dwarf galaxies in the
150: Local Group, all of which have been found as part of our ongoing
151: CFHT/MegaPrime photometric survey of M31. This new imaging extends our
152: survey area from the south-eastern quadrant discussed in
153: \cite{ibata2007} to the west, and currently includes an additional
154: $49$\,sq.\,degrees of M31's halo out to a maximum projected radius of
155: 150\,kpc. Section~2 summarises the observations and data-reduction
156: procedures and Section~3 presents a preliminary analysis of the new
157: dwarfs and quantifies their global properties. In Section~4, we
158: discuss our results in relation to some of the key questions which
159: have been prompted with the discoveries of so many new low luminosity
160: galaxies in the Local Group. Section~5 summarises our results.
161:
162: \section{Observations}
163:
164: \cite{martin2006} and \cite{ibata2007} presented first results from our
165: CFHT/MegaPrime survey of the south-west quadrant of M31, obtained in
166: semesters S02B -- 06B. Since S06B, we have initiated an extension
167: to this survey with the aim of obtaining complete coverage of the
168: southern hemisphere of M31's halo out to a maximum projected radius of
169: 150\,kpc from the center of M31. Figure~1 shows the locations of these
170: new fields relative to M31 in a tangent-plane projection. Red hatched
171: fields represent those fields previously presented in
172: \cite{ibata2007}. Magenta open fields represent the new survey area,
173: where solid lines denote fields which were observed in S06B -- 07B,
174: and dotted lines denote fields yet to be observed. Black stars mark
175: the positions of known M31 satellite galaxies, and open stars mark the
176: positions of the three new dwarfs presented herein.
177:
178: Our observing strategy is very similar to that described in
179: \cite{ibata2007}, to which we refer the reader for further details. In brief,
180: CFHT/MegaPrime consists of a mosaic of thirty-six $2048 \times
181: 4612$\,pixel CCDs with a total field of view of $0.96 \times
182: 0.94$ sq.\,degrees at a pixel scale of $0.187$\,arcsec\,pixel$^{-1}$. We
183: observe in the CFHT $g$ and $i$ bands for a total of 1350 seconds
184: each, split into $3 \times 450$\,seconds dithered sub-exposures, in $<
185: 0.8$\,arcsec seeing. This is sufficient to reach $g \sim 25.5$ and $i
186: \sim 24.5$ with a signal-to-noise of 10. In some cases, more than
187: three exposures were taken (at the discretion of CFHT staff to ensure
188: the requested observing conditions were met), and in these cases the
189: viable images were included in the stacking procedure, weighted
190: according to noise/seeing. We have chosen a tiling pattern which
191: typically has very little overlap between fields, and so we use short,
192: 45\,second exposures in $g$ and $i$ offset by half a degree in the right
193: ascension and declination directions in order to establish a
194: consistent photometric level over the survey. This typically has a rms
195: scatter of 0.02\,mags over our survey area.
196:
197: The CFHT/MegaPrime data were pre-processed by CFHT staff using the
198: {\it Elixir} pipeline, which accomplishes the bias, flat, and fringe
199: corrections and also determines the photometric zero point of the
200: observations. These images were then processed using a version of the
201: CASU photometry pipeline (\citealt{irwin2001}) adapted for
202: CFHT/MegaPrime observations. The pipeline includes re-registration,
203: stacking, catalogue generation and object morphological
204: classification, and creates band-merged $g,i$ products for use in the
205: subsequent analysis. The CFHT $g$ and $i$ magnitudes are de-reddened
206: using the \cite{schlegel1998} IRAS maps, such that $g_0 = g - 3.793
207: E(B - V)$ and $i_0 = i - 2.086 E(B - V)$, where $g_0$ and $i_0$ are
208: the de-reddened magnitudes.
209:
210: \section{Analysis}
211:
212: In this section we present an initial analysis of the three new dwarf
213: galaxies using the CFHT/MegaPrime discovery data. The measured
214: parameters of the dwarfs are summarised in Table~1.
215:
216: \subsection{Discovery and stellar populations}
217:
218: Two of the new dwarf galaxies (Andromeda XVIII and XIX) stand out as
219: prominent overdensities of stars in our survey and can be clearly
220: identified by eye in maps of the distribution of stellar
221: sources. Andromeda~XX, on the other hand, is considerably fainter and
222: its CMD is far more sparsely populated. Despite this, it was initially
223: identified by one of us (A. Huxor) through visual examination of the
224: individual CCDs during a search for globular clusters. An automated
225: detection algorithm, based upon a boxcar matched-filter search for
226: local overdensities with a variable width, was subsequently applied
227: after these preliminary searches. As well as highlighting these three
228: dwarfs, some other dwarf galaxy candidates were identified and are
229: being followed up. A subsequent paper will deal in detail with the
230: automated detection of dwarf galaxies around M31 to enable a full
231: completeness study, although such an analysis requires more contiguous
232: coverage of M31 than we currently possess. Prior to such a study, we
233: do not make any claims regarding the completeness of the satellite
234: sample so far discovered.
235:
236: The top panels of Figure~2 shows the $i_o$ versus $(g - i)_o$
237: colour-magnitude diagrams (CMDs) for the three new dwarf galaxies
238: discovered in the south-west quadrant of M31 and whose positions
239: relative to this galaxy are indicated in Figure~1. The bottom panels
240: of Figure~2 show reference fields with equivalent areas offset from
241: the center of each of the galaxies by several half-light radii. Each
242: of the CMDs has been corrected for foreground extinction. In each of
243: the three cases, a red giant branch (RGB) is clearly visible, although
244: in the case of Andromeda~XX it is poorly populated. To the depth of
245: these observations, it appears that there are very few, if any, bright
246: main-sequence and blue loop stars which would be indicative of younger
247: stellar populations, and it is likely therefore that these galaxies do
248: not host a dominant young population. Stars to the red of the RGB
249: (with $2 \lesssim (g - i)_o \lesssim 3$) are likely foreground Milky
250: Way disk stars, although intermediate-age asymptotic giant branch
251: stars can also occupy this colour locus and have a luminosity similar
252: to or brighter than the tip of the red giant branch (although this is
253: probably only relevant for Andromeda~XIX). In the Andromeda~XIX CMD
254: and reference field, the vertical feature at $(g - i)_o \sim 0.3$ is
255: the foreground Milky Way halo locus (see \citealt{martin2007} for an
256: analysis of this feature in our extant M31 survey). Given these
257: current data, all of the CMDs appear to show stellar populations
258: typical of dSph galaxies. The faint blue objects centered around $i_o
259: \sim 25.2$ with a mean colour of $(g - i)_o \sim 0.5$ in the
260: Andromeda~XIX CMD may be a horizontal branch component. However, as
261: the reference field shows, contamination from misclassified background
262: galaxies is considerable in this region of colour - magnitude
263: space. There is also some evidence of a very weak RGB population in
264: the Andromeda~XIX reference field, which is likely due to the
265: background M31 halo and stellar overdensities in the vicinity of
266: this dwarf galaxy (see Section 4.3.2).
267:
268: Figure~3 shows various properties for each of the three new dwarf
269: galaxies. The left-most panels show $I_o$ versus $(V - I)_o$ CMDs for
270: each galaxy. We have transformed CFHT $gi$ to Landolt $VI$ using a
271: two-stage transformation; we first change CFHT $gi$ into INT $V^\prime
272: i$ using the relations derived in \cite{ibata2007}, and we then
273: transform INT $V^\prime i$ into Landolt $VI$ using the transformations
274: given in \cite{mcconnachie2004a}\footnote{see
275: http://www.ast.cam.ac.uk/$\sim$wfcsur/colours.php for details}. In
276: each CMD, only those stars which lie within two half-light radii from
277: the center of each galaxy (shown by the red dashed ellipse in the
278: second panel) have been plotted. The dashed lines define a colour cut
279: designed to preferentially select stars which are members of the dwarf
280: galaxies. The solid line shows a 13\,Gyr isochrone with the
281: representative metallicity of the dwarf from \cite{vandenberg2006},
282: shifted to the distance modulus of the dwarf (the distance and
283: metallicity of each dwarf is calculated in Section~3.2).
284:
285: The second panel in each row of Figure~3 shows the spatial
286: distribution of candidate RGB stars in the vicinity of each galaxy,
287: defined by the colour cuts discussed previously. Black dashed lines
288: show the edges of the CFHT/MegaPrime CCDs. Both Andromeda~XVIII and XX
289: appear as obvious concentrations of stars, despite Andromeda~XX being
290: poorly populated. Andromeda~XVIII lies at the corner of one of the
291: CCDs, and much of this galaxy hides behind the large gap between the
292: second and first rows of CCDs in the CFHT/MegaPrime field (see Section
293: 3.3). Andromeda~XIX is a much more extended and diffuse system than
294: the other two, and contours have been overlaid to more clearly show
295: its structure. The first contour is set $3 - \sigma$ above the
296: background, and subsequent contour levels increase by $1.5\,\sigma$
297: over the previous level. This galaxy is located on the boundary of our
298: survey, overlapping slightly with the extant survey region from
299: \cite{ibata2007}. We include some adjacent fields from this earlier
300: part of the survey to obtain complete coverage of Andromeda~XIX.
301:
302: \subsection{Distances and metallicities}
303:
304: The upper-right panels in each row of Figure~3 show, for each galaxy,
305: the de-reddened $I$-band luminosity functions of stars in the CMD
306: which satisfy the colour and spatial cuts defined previously. These
307: have been corrected for foreground/background contamination by
308: subtracting a nearby ``reference'' field, scaled by area. The scaled
309: reference field is shown by the dotted line, to illustrate the
310: contribution from the foreground/background as a function of
311: magnitude. The $I-$band magnitude of the tip of the red giant branch
312: (TRGB; corresponding to the point in the evolution of a RGB star
313: immediately prior to it undergoing the core helium flash) is a
314: well-calibrated standard candle which is used extensively for nearby
315: galaxies (e.g.,
316: \citealt{lee1993a,salaris1997,mcconnachie2004a,mcconnachie2005a} and
317: references therein). In a well populated luminosity function, it is
318: normally taken to be equal to the luminosity of the brightest RGB
319: star. However, when dealing with faint dwarfs - particularly systems
320: like Andromeda~XX with a very sparse RGB - this assumption is likely
321: to be flawed due to sampling errors. However, for this initial
322: analysis of these galaxies we assume that the TRGB position measured
323: in this way is a good estimate of its actual position. We note that
324: the resulting distance modulus of Andromeda~XX in particular is
325: uncertain and will be refined once deeper data reaching below the
326: horizontal branch is available.
327:
328: Our best estimates for the (extinction-corrected) $I-$band magnitude
329: of the TRGB are highlighted on each of the luminosity functions in
330: Figure~3 and are listed in Table~1. For Andromeda~XX, we have adopted
331: very conservative error bars; the lower limit is an estimate of the
332: possible offset of the brightest RGB star from the true TRGB from our
333: experience with the comparably faint Andromeda~XII (\citealt{chapman2007}); the upper limit
334: assumes that the few brightest stars we have identified are actually
335: foreground contamination, and that the true TRGB is represented by the
336: group of stars at $I_o \sim 21.2$. Adopting $M_I = -4.04 \pm 0.12$
337: (\citealt{bellazzini2001}) yields a preliminary distance to each of
338: the new dwarf galaxies; the derived distance moduli and distances are
339: given in Table~1. Most notable is the distance to Andromeda~XVIII,
340: which has a well-defined TRGB, and which places it approximately
341: 1.4\,Mpc from the Milky Way ($\sim 600$\,kpc distant from M31), at the
342: periphery of the Local Group.
343:
344: As an independent check of our distance estimates (particularly that
345: for Andromeda~XX), we construct $g_o$ luminosity functions for each
346: galaxy using stars within two half-light radii of the centers. These
347: are shown in Figure~4 as solid lines. Also shown as dotted lines are
348: luminosity functions for nearby reference fields, scaled by
349: area. These luminosity functions go deeper than the previous CMDs
350: since stars are only required to be detected in the $g$-band. Our data
351: start to become seriously incomplete below $g_o \sim 25.5$, and
352: photometric errors at this magnitude are of order $\Delta\,g \simeq
353: 0.15$. For reference, the horizontal branch in M31 has a magnitude of
354: $g_o \sim 25.2$ (\citealt{ibata2007}). For Andromeda~XIX and XX, peaks
355: of stars are visible at $g \sim 25.3$ and $g \sim 25.6$, respectively,
356: which are not present in the reference fields, and which are marked in
357: Figure~4 by dashed lines. We attribute these peaks to the detection of
358: horizontal branch stars in each of these galaxies. While the peak for
359: Andromeda~XIX is less apparent than that for Andromeda~XX, its
360: position coincides with the expected luminosity of the horizontal
361: branch from inspection of its CMD in Figure~2, reinforcing our
362: interpretation of this feature. In contrast, no such feature is
363: visible for Andromeda~XVIII, which is expected given that we measure
364: it to be much more distant than the other two and so our observations
365: will not be deep enough to observe the horizontal branch
366: population. Similarly, our measurements of the positions of the
367: horizontal branches in Andromeda~XIX and XX are consistent with the
368: positions we measure for the TRGB in these galaxies. These detections
369: (and non-detection) of the horizontal branches are therefore
370: consistent with the distances derived from the TRGB, and suggest that
371: the uncertainty in the distance to Andromeda~XX may be less than we
372: currently adopt in Table~1.
373:
374: The lower-right panels of Figure~3 show the observed photometric
375: metallicity distribution (MDF) function, constructed using the same
376: technique as detailed in \cite{mcconnachie2005a}, using a bi-linear
377: interpolation of stars in the top two magnitudes of the RGB with
378: 13\,Gyr isochrones, [$\alpha$/Fe] $= 0$, from \cite{vandenberg2006}
379: with $BVRI$ colour-$T_{eff}$ relations as described by
380: \cite{vandenberg2003}. Each MDF has been corrected for
381: foreground/background contamination by subtraction of a MDF for a
382: reference field, scaled by area. The MDF for the scaled reference
383: field is shown as a dotted line in each panel. The mean metallicity
384: and metallicity spread, as quantified by the inter-quartile range
385: (IQR), are highlighted in Figure~3, and an isochrone corresponding to
386: the mean metallicity of the dwarf is overlaid on the CMD in the first
387: panels, shifted to the distance modulus of the dwarf galaxy.
388:
389: The metallicity spread in each of the three galaxies is similar,
390: although the IQR for Andromeda~XIX appears slightly smaller than for
391: the other two. Certainly, the colour spread of the RGB seen from the
392: CMDs is much smaller for Andromeda~XIX than for Andromeda~XVIII and
393: XX. That this does not correspond to a much smaller spread in
394: metallicity probably reflects the metal poor nature of Andromeda~XIX,
395: since RGB colour is a poor indicator of metallicity variation at very
396: low metallicities. It is also tempting to suggest that the narrow
397: spread in RGB colour indicates that Andromeda~XIX is a simple stellar
398: population; however, lessons learned from the Carina dSph, which has a
399: large age and metallicity spread but conspires to have a narrow RGB
400: (\citealt{smeckerhane1994}), suggests a note of caution against this
401: interpretation.
402:
403: The metallicity information is summarised in Table~1. The formal
404: uncertainties in the metallicity and metallicity spread estimates are
405: of order 0.1\,dex. In addition to uncertainties in the stellar models,
406: our metallicity estimates assume that (i) the dwarfs are all dominated
407: by a $13$\,Gyr stellar population, and (ii) the distance modulus for
408: each galaxy is well estimated. The former assumption is likely
409: reasonable, and should not lead to an error $\gtrsim 0.2$\,dex unless
410: the dwarfs are dominated by intermediate-age and young stellar
411: populations (for which there is no current evidence). The latter
412: assumption looks to be reasonable for Andromeda~XVIII and XIX, where
413: the RGB is reasonably well populated, but for Andromeda~XX the
414: uncertainty introduced through the distance estimate could be more
415: significant. We note that the metallicities of Andromeda~XVIII and XIX
416: look to be significantly lower than the median metallicity of the
417: kinematically-selected halo of M31, which has [Fe/H]$ \simeq -1.4$
418: (\citealt{chapman2006,kalirai2006b}).
419:
420: \subsection{Structures and magnitudes}
421:
422: We quantify the structures of Andromeda~XVIII, XIX and XX through the
423: spatial distributions of their resolved stars. However, the analysis
424: is made more complex since Andromeda~XIX is very diffuse, Andromeda~XX
425: has very few bright stars on which to base our analysis, and each of
426: the dwarf galaxies lies close to or at the edges of CCDs. In the
427: extreme case of Andromeda~XVIII, we are clearly missing a significant
428: part of the galaxy which lies behind the large gap between the second
429: and first rows of the CFHT/MegaPrime mosaic. To illustrate this, the
430: top panel of Figure~5 shows the $i-$band image of Andromeda~XVIII with
431: linear scaling; while Andromeda~XVIII is clearly visible to the naked
432: eye, much of the galaxy falls off the edge of the detector. To
433: determine how large this effect is, the lower panel of Figure~5 shows
434: a $10 \times 10$\,arcmins image centered on the coordinates of
435: Andromeda~XVIII from the POSSII/UKSTU (Blue) survey which we retrieved
436: through the Digitized Sky Survey, and which covers the entirety of
437: this galaxy.
438:
439: Given the several complications discussed above, we choose to derive
440: the structural parameters for the dwarfs based upon the maximum
441: likelihood technique developed by \cite{martin2008} instead of the
442: usual technique which bins the data spatially and uses smoothing
443: kernels (e.g., \citealt{irwin1995,mcconnachie2006b}). The procedure
444: has been modified from \cite{martin2008}, to which we refer the reader
445: for details, to account for incomplete coverage of the dwarfs due to
446: CCD edges. In brief, this technique calculates simultaneously the most
447: plausible values for the centroid, ellipticity, position angle and
448: half-light radius of the dwarf under the assumption that the surface
449: brightness radial profile is well described by an exponential curve, without
450: any need for smoothing or binning of the data. However, for
451: Andromeda~XVIII this approach is still insufficient since our data
452: only samples one segment of the galaxy, as shown by comparing the
453: POSSII/UKSTU image with the CFHT/MegaPrime image in Figure~5. Thus, for this
454: galaxy, we estimate its center from the POSSII/UKSTU data and
455: approximate it as circular. The half-light radius is then calculated
456: via the same technique as for Andromeda~XIX and XX using the CFHT/MegaPrime
457: data.
458:
459: The centroid, half-light radius ($r_h$), position angle (measured east
460: from north) and ellipticity ($\epsilon = 1 - b/a$) for each dwarf
461: galaxy, derived using the maximum likelihood technique (with the above
462: caveat for Andromeda~XVIII), are listed in Table~1. In addition,
463: Figure~6 shows the (background-corrected) stellar density profile
464: (equivalent to the surface brightness profile), derived using the same
465: technique as in \cite{mcconnachie2006b}, for each of the three dwarf
466: galaxies. We use elliptical annuli with the position angle,
467: ellipticity and centroid listed in Table~1. Overlaid on these profiles are
468: exponential profiles with the appropriate half light radii (the
469: exponential scale radius, $r_e \simeq 0.6 r_h$). These profiles are
470: the most probable exponential models for the stellar density
471: distribution of the dwarf galaxy derived using the maximum likelihood
472: method, and are not fits to the averaged data-points.
473:
474: We estimate the magnitude of Andromeda~XIX and XX in a similar way as
475: \cite{martin2006} and \cite{ibata2007}. First, we sum the total
476: $V-$band flux from candidate member stars which are within the
477: half-light radius of each dwarf galaxy and which are within $2 - 3$
478: magnitudes of the TRGB. However, this flux does not take into account
479: the contribution to the total light from fainter stars, most of which
480: we do not detect. To determine the appropriate correction to apply, we
481: compare the half-light flux of Andromeda~III measured in this way
482: (using similar CFHT/MegaPrime observations) to its apparent magnitude
483: of $m_v = 14.4 \pm 0.3$, directly measured by
484: \cite{mcconnachie2006b}. We then apply the appropriate correction to
485: the fluxes for each dwarf galaxy. Clearly, the uncertainties
486: associated with this method are considerable, and we make the implicit
487: assumption that the luminosity functions of Andromeda~III, XIX and XX
488: are similar. Under this assumption, we estimate an accuracy of $\sim
489: 0.6$\,mags in the final magnitude of Andromeda~XIX, although we
490: estimate a larger uncertainty of $\sim 0.8$\,mags for Andromeda~XX due
491: to the small number of bright stars available. The central surface
492: brightness of Andromeda~XIX and XX are estimated by normalizing the
493: exponential profiles shown in Figure~6 so that the surface integral
494: over the dwarf out to the half-light radius is equal to half the total
495: flux received from the dwarf. These numbers are also given in Table~1.
496:
497: It is not possible to derive the magnitude of Andromeda~XVIII
498: in the same way as above given that we only sample a segment of this
499: galaxy with our data. Comparison of the POSSII/UKSTU images of
500: Andromeda~XVIII with those of Andromeda~V, VI and VII show that it is
501: considerably lower surface brightness than either Andromeda~VI or VII,
502: but is similar to - and perhaps brighter than - that of Andromeda~V,
503: which has $S_o = 25.6 \pm 0.3$ (\citealt{mcconnachie2006b}). We
504: therefore adopt this as a faint-end limit to the central surface
505: brightness of Andromeda~XVIII. A faint-end limit to its magnitude can
506: then be calculated by normalizing its radial surface brightness
507: profile to this central value, integrating over its area out to the
508: half light radius, and multiplying the answer by two. The magnitude
509: derived in this way is given in Table~1. We note that updated
510: magnitudes and surface brightnesses will be derived for each of the
511: three new galaxies using the unresolved light component from
512: dedicated, follow-up, photometric studies.
513:
514: \section{Discussion}
515:
516: Andromeda~XVIII, XIX and XX have a range of relatively unusual
517: properties. In particular, Andromeda~XVIII is one of the most distant
518: Local Group galaxies discovered for several years, and is one of the
519: most isolated systems in the Local Group. Andromeda~XIX is extremely
520: extended, with a very large half-light radius and extremely faint
521: central surface brightness. Andromeda~XX, on the other hand, is one of
522: the lowest luminosity dwarf galaxies so far discovered around M31,
523: with a magnitude of $M_V \simeq -6.3^{+1.0}_{-0.7}$, comparable to the
524: luminosity of Andromeda~XII ($M_V = -6.4 \pm 1.0$;
525: \citealt{martin2006}). In this section, we discuss the properties of
526: these galaxies in the larger context of the main science questions
527: raised by the recent discoveries of so many new dwarf galaxies.
528:
529: \subsection{Completeness}
530:
531: Prior to 2004, there were 15 dSph galaxies known in the Local Group
532: (nine Milky Way satellites, six M31 satellites and two isolated
533: systems, Cetus and Tucana). Since this time, 22 new dwarf galaxies
534: (including possible diffuse star clusters around the Milky Way) have
535: been discovered in the Local Group, the overwhelming majority of which
536: are dSph satellites of the Milky Way and M31. For the Milky Way, the
537: SDSS has been responsible for all the discoveries to date, and most of
538: the galaxies discovered have been extremely faint; no new Milky Way
539: satellites with $M_V \lesssim -8$ have been found. Thus, apart from
540: satellites hidden by the Milky Way disk, our satellite system is
541: probably complete to this approximate magnitude limit, as originally
542: argued by \cite{irwin1994}.
543:
544: Around M31, it is more difficult to identify extremely faint dwarf
545: galaxies since we cannot probe as far down the stellar luminosity
546: function. Andromeda~XII and Andromeda~XX are the two faintest M31
547: satellites found so far, both with $M_V \sim -6.3$. For comparison,
548: the faintest Milky Way satellite found to date is probably Willman I,
549: with $M_V \sim -2.7$ (\citealt{willman2006,martin2008}).
550:
551: Andromeda~XVIII is considerably brighter than Andromeda~XX, and has a
552: central surface brightness similar to or brighter than Andromeda~V
553: ($S_o = 25.6 \pm 0.3$\,mags\,arcsec$^{-2}$). Andromeda~XVIII is
554: clearly visible in the POSSII/UKSTU (Blue) survey image which we
555: retrieved through the Digitized Sky Survey and which is reproduced in
556: the lower panel of Figure~5. However, its identification is made more
557: complicated by numerous nearby bright stars and nebulosity in its
558: vicinity, which may act to explain why it was not discovered using
559: these data. We have also confirmed that it is visible in the original
560: POSSI (Blue) survey. Its belated discovery indicates that previous
561: surveys for relatively {\it bright} dwarf galaxies around M31 were
562: incomplete and that some dwarfs were missed. Variable and unknown
563: completeness is problematic for studies of satellite distributions and
564: highlights the vital need for more systematic studies such as those
565: now being conducted.
566:
567: It is fortuitous that Andromeda~XVIII lies within our survey area
568: given its considerable distance from M31. Indeed, even as current and
569: future surveys help improve the completeness of the M31 and Milky Way
570: satellite systems, many isolated Local Group galaxies can be expected
571: to continue to elude detection: unlike the Milky Way satellites, they
572: are not nearby, and unlike the M31 satellites, they are not
573: necessarily clustered in an area amenable to systematic
574: searches. PanStarrs $3\pi$ will survey a large fraction of the sky a
575: magnitude deeper than SDSS, and should discover isolated Local Group
576: galaxies, particularly those within 500\,kpc or so from the Milky
577: Way. However, very faint galaxies much further away than this ($\sim
578: 1$\,Mpc) may prove more difficult to spot. Exactly how many very faint
579: dwarf galaxies are to be found at the periphery of the Local Group is
580: likely to remain uncertain for some time yet.
581:
582: \subsection{Spatial distribution}
583:
584: Several recent studies of the spatial distributions of satellites
585: around the Milky Way and M31
586: (\citealt{willman2004,kroupa2005,mcconnachie2006a,koch2006,metz2007,irwin2008})
587: have generally concluded that the distributions appear anisotropic:
588: \cite{mcconnachie2006a} highlight the fact that (at the time) 14 out
589: of the 16 candidate satellites of M31 are probably on the near side of
590: M31, while others (\citealt{kroupa2005,koch2006,metz2007,irwin2008})
591: conclude that many of the Milky Way and M31 satellites are aligned in
592: very flattened, disk-like, distributions (an observation originally
593: made by \citealt{lyndenbell1976,lyndenbell1982}).
594:
595: Andromeda~XVIII, XIX and XX do not lie near any of the principle
596: satellite planes previously proposed to exist around M31. As discussed
597: in the previous sub-section, the census of Local Group galaxies is
598: clearly not complete, and it is too early to draw definitive
599: conclusions regarding the distributions of satellites. This is
600: particularly true around M31, where relatively bright satellites are
601: still being discovered. For the Milky Way, the SDSS covers roughly
602: one-fifth of the Milky Way halo in the direction of the north galactic
603: cap; depending upon how many satellites are found in future surveys at
604: lower latitudes, the statistical significance of the proposed streams
605: of satellites may change substantially.
606:
607: In terms of spatial distributions, Andromeda~XVIII is unusual insofar
608: as it is very distant - roughly 1.4\,Mpc from the Milky Way, and
609: roughly 600\,kpc from M31. Thus it is probably not a satellite of M31,
610: although kinematics may help reveal whether it is approaching M31 and
611: the Local Group for the first time (like Andromeda~XII,
612: \citealt{chapman2007}) or if it has been thrown out from M31 following
613: an interaction (like Andromeda~XIV, \citealt{majewski2007,sales2007}).
614:
615: \subsection{Environment and structures}
616:
617: \subsubsection{Andromeda~XVIII, position and morphology}
618:
619: Andromeda~XVIII appears to possess stellar populations typical of dSph
620: galaxies. If it is subsequently confirmed to be gas poor, then it will
621: be the third dSph galaxy found in isolation in the Local Group (in
622: addition to Cetus and Tucana). The fact that isolated galaxies are
623: preferentially more gas-rich compared to satellites
624: (\citealt{einasto1974}) has lead to the proposition that satellite
625: galaxies are stripped of their gas via ram-pressure stripping and
626: tidal harassment in the halo of the host galaxy (e.g.,
627: \citealt{mayer2006}). However, for isolated systems such as
628: Andromeda~XVIII, Cetus and Tucana, prolonged interactions with massive
629: galaxies are unlikely to have occurred. Likewise, the gas-deficient
630: satellite Andromeda~XII is not believed to have undergone any past
631: interactions with a large galaxy since it appears to be on its first
632: infall into the potential of M31 (\citealt{chapman2007}). Further, the
633: most compelling case of a dwarf galaxy thought to be undergoing
634: ram-pressure stripping is Pegasus (DDO216;
635: \citealt{mcconnachie2007c}), an {\it isolated} galaxy more than
636: 400\,kpc from M31. Clearly, understanding if these observations are
637: consistent with the present models for dwarf galaxy evolution requires
638: a more complete inventory of nearby galaxies and their properties than
639: we currently possess.
640:
641: \subsubsection{Andromeda~XIX, tides and substructure}
642:
643: The half-light radius of Andromeda~XIX is $6.2$\,arcmins. At the
644: distance we derive for it, this corresponds to $r_h \simeq 1.7$\,kpc,
645: which is the largest value yet recorded for any dSph in the Local
646: Group. The average half-light radius for Milky Way dSphs is an order
647: of magnitude less, at $r_h \sim 150$\,pc, and none have half-light
648: radii larger than $r_h \simeq 550$\,pc (with the exception of the
649: tidally disrupting Sagittarius dSph; \citealt{majewski2003}) . M31
650: dSphs, on the other hand, have typical half-light radii of $r_h \sim
651: 300$\,pc, with the previous extremes being Andromeda~II, with $r_h
652: \simeq 1.1$\,kpc, and Andromeda~VII, with $r_h \simeq 750$\,pc
653: (\citealt{mcconnachie2006b}). The extremely diffuse and extended
654: nature of Andromeda~XIX is reminiscent of the ``outer component'' of
655: Andromeda~II, as traced by horizontal branch stars by
656: \cite{mcconnachie2007a}.
657:
658: It is tempting to attribute the diffuse structure of Andromeda~XIX to
659: tidal interactions. In this respect, it is relevant to note that
660: Andromeda~XIX lies very close to the major axis substructure
661: identified by \cite{ibata2007}. No independent distance estimate to
662: this substructure currently exists; \cite{ibata2007} assumed it to be
663: at the distance of M31 but if it is at the same distance as
664: Andromeda~XIX then the photometric metallicity estimates of these
665: features will be very similar. Figure~7 shows the surroundings of
666: Andromeda~XIX as a stellar density map; the first two contour levels
667: are $2$ and $3 - \sigma$ above the background, and the levels then
668: increase by $1.5\,\sigma$ over the previous level. As well as showing
669: Andromeda~XIX as a prominent overdensity, there is some evidence of
670: stellar material in its outskirts (also visible in the contours of
671: Figure~3). Whether or not Andromeda~XIX is the source of the major
672: axis substructure identified in \cite{ibata2007}, or is being tidally
673: perturbed, will require detailed kinematics in this region. We note
674: that \cite{penarrubia2008b} show that the effect of tides on dwarf
675: galaxies in cosmological haloes is to decrease the central surface
676: brightness and {\it decrease} the half light radius of the bound
677: component. This would argue against tidal effects explaining the
678: structure of Andromeda~XIX.
679:
680: The large scale-size of Andromeda~XIX reinforces the difference in
681: scale-size between the Milky Way and M31 satellites first highlighted in
682: \cite{mcconnachie2006b}, such that the M31 dSphs are more extended
683: than their Milky Way counterparts.
684: \cite{penarrubia2008a,penarrubia2008b} have investigated the cause of
685: this disparity in an attempt to relate it to either differences in the
686: underlying dark matter properties of the dwarfs or differences in
687: their evolution around their hosts. They conclude that tidal effects
688: are insufficient to explain the magnitude of the effect. However, if
689: the different scale sizes reflect intrinsic differences between the
690: Milky Way and M31 sub-haloes then this should reveal itself in the
691: kinematics of the two populations (with the M31 dwarfs being
692: dynamically hotter than their Milky Way counterparts). Whatever the
693: cause, the comparison of Andromeda~XIX and the other M31 satellites to
694: the Milky Way population highlights the importance of sampling dwarfs
695: in a range of environments so as to obtain a fuller appreciation of
696: the range of properties that these systems possess. In turn, this
697: helps us understand the physical drivers behind the differences and
698: similarities we observe. We note that studies of the star clusters of
699: M31 (\citealt{huxor2005,huxor2008}) have already extended the known
700: parameter space for these objects, with the M31 population containing
701: extended star clusters not found in the Milky Way population.
702:
703: \subsection{Satellites that are missing and ``the missing satellites''}
704:
705: Andromeda~XX is an exceptionally faint galaxy with a very poorly
706: populated RGB. This makes an accurate derivation of its properties
707: particularly difficult. However, the star formation history of
708: Andromeda~XX and the other ultra-faint satellites is particularly
709: relevant to the ``missing satellites'' question (do all the thousands
710: of dark matter sub-haloes predicted to exist in the haloes of galaxies
711: like the Milky Way and M31 contain stars and, if they do, where are
712: they?). Until recently, only a dozen or so dwarf satellites were
713: observed, and it was noted that the cumulative mass distribution of
714: these satellites was dramatically different to that of predicted dark
715: matter sub-haloes, even at relatively large masses
716: (\citealt{moore1999,klypin1999}). To solve this discrepancy without
717: altering the underlying cosmology, it was suggested that either there
718: were a large number of luminous satellites awaiting discovery or that
719: not all sub-haloes have a luminous component.
720:
721: Despite many new galaxies in the Local Group being discovered, and
722: many more undoubtedly awaiting discovery, we consider it very unlikely
723: that these discoveries will resolve the discrepancy between theory and
724: observation. The original comparison between the observed and
725: predicted satellite mass functions shows that the discrepancy sets in
726: for dwarfs as luminous as the Small Magellanic Cloud ($M_V \simeq
727: -16$) and Fornax ($M_V \simeq -13 $). Finding thousands of very faint
728: (and presumably less massive?) satellites would not solve the
729: disagreement at the more massive end and there is no evidence to
730: suggest that a dozen galaxies the luminosity of Fornax have been
731: missed (e.g., \citealt{irwin1994}). Further, as higher resolution dark
732: matter simulations make clear (e.g., \citealt{diemand2007}), the
733: sub-halo mass function appears to continue to increase at the low mass
734: end. It seems reasonable, therefore, that at some point these haloes
735: will not be massive enough to be able to accrete and/or retain baryons
736: and form stars, and this implies that there is a minimum mass halo
737: which can host a luminous component (\citealt{kravtsov2004}).
738:
739: A re-analysis of the observed dynamics of the dwarf galaxies by
740: \cite{penarrubia2008a,penarrubia2008b} within the $\Lambda-$CDM
741: çframework has shown that few-if-any of these galaxies (including
742: recent discoveries) occupy a halo with a circular velocity less than
743: $\sim 10 - 20$\,km\,s$^{-1}$. Further, these estimates bring the
744: cumulative distribution of luminous satellites and dark matter
745: sub-haloes into good agreement at the high-mass end. Using a different
746: technique, \cite{strigari2007b} find a similar result. Given that
747: these authors find good agreement between observations and theory down
748: to a certain mass limit, their results support the idea of a mass
749: threshold in dark matter haloes below which star formation becomes
750: highly inefficient. Therefore, by continuing to identify new,
751: ultra-faint dwarfs, we probe the astrophysics of galaxy formation at
752: low mass limits where the sensitivity to complex feedback mechanisms -
753: such as star formation (\citealt{kravtsov2004}) and reionization
754: (\citealt{bullock2001}) - is greatest.
755:
756: \section{Summary}
757:
758: We have presented three new Local Group dwarf galaxies discovered as
759: part of our ongoing CFHT/MegaPrime survey of M31 and its
760: environs. These galaxies - christened Andromeda~XVIII, XIX and XX
761: after the constellation in which they are found - have stellar
762: populations which appear typical of dSph galaxies. Individually, each
763: of these galaxies has relatively unusual properties compared to the
764: previously known dwarfs in the vicinity of M31:
765:
766: \begin{itemize}
767: \item Andromeda~XVIII is extremely distant, at $1355 \pm 88$\,kpc from
768: the Milky Way, placing it nearly $600$\,kpc from M31. Thus it is one
769: of the most isolated galaxies in the Local Group. It is clearly
770: observed through its integrated light (it appears to have a central
771: surface brightness similar to or brighter than that of Andromeda~V)
772: and suggests that there could be several other relatively bright dwarf
773: galaxies within the Local Group which have so far eluded detection;
774:
775: \item Andromeda~XIX is extremely extended, with a half-light radius of
776: $r_h = 1683 \pm 113$\,kpc. This is an order of magnitude more extended
777: than typical Milky Way dSphs. While its integrated luminosity is $M_V = -9.3 \pm 0.6$, its central surface brightness is exceptionally
778: low, at $S_o = 29.3 \pm 0.7$. Andromeda~XIX reinforces the difference in
779: scale-size between the Milky Way and M31 satellites first discussed in
780: \cite{mcconnachie2006b}. This galaxy may be being tidally disrupted,
781: and could be related to major axis substructure first identified in
782: \cite{ibata2007} and which lies near to Andromeda~XIX in
783: projection. However, we note that calculations by
784: \cite{penarrubia2008b} show that the net effect of tides on a dwarf
785: galaxy is to decrease the central surface brightness and {\it
786: decrease} the half-light radius of the bound component;
787:
788: \item Andromeda~XX is extremely faint, with an absolute magnitude of
789: order $M_V = -6.3^{+1.0}_{-0.7}$. It is one of the faintest galaxy so far
790: discovered in the vicinity of M31 (comparable in luminosity to
791: Andromeda XII) and as such many of its key parameters are extremely
792: uncertain at this stage. A full inventory of these systems is required
793: to properly define the faint-end of the galaxy luminosity function,
794: and to determine where, if anywhere, we encounter a lower limit to the
795: galaxy mass/luminosity function.
796: \end{itemize}
797:
798: \acknowledgements Based on observations obtained with
799: MegaPrime/MegaCam, a joint project of CFHT and CEA/DAPNIA, at the
800: Canada-France-Hawaii Telescope (CFHT) which is operated by the
801: National Research Council (NRC) of Canada, the Institute National des
802: Sciences de l'Univers of the Centre National de la Recherche
803: Scientifique of France, and the University of Hawaii. We are indebted
804: to the CFHT staff for their help and careful observations, and we
805: thank the anonymous referee for useful comments which improved the
806: clarity of this paper. AWM thanks Evan Skillman, Jorge Pe{\~n}arrubia
807: and Andrew Cole for useful discussions. AWM is supported by a Research
808: Fellowship from the Royal Commission for the Exhibition of 1851, and
809: thanks Sara Ellision and Julio Navarro for additional financial
810: assistance. AH and AMNF acknowledge support from a Marie Curie
811: Excellence Grant from the European Commission under contract
812: MCEXT-CT-2005-025869.
813:
814: \begin{thebibliography}{75}
815: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
816:
817: \bibitem[{{Armandroff} {et~al.}(1998){Armandroff}, {Davies}, \&
818: {Jacoby}}]{armandroff1998}
819: {Armandroff}, T.~E., {Davies}, J.~E., \& {Jacoby}, G.~H. 1998, \aj, 116, 2287
820:
821: \bibitem[{{Armandroff} {et~al.}(1999){Armandroff}, {Jacoby}, \&
822: {Davies}}]{armandroff1999}
823: {Armandroff}, T.~E., {Jacoby}, G.~H., \& {Davies}, J.~E. 1999, \aj, 118, 1220
824:
825: \bibitem[{{Bellazzini} {et~al.}(2001){Bellazzini}, {Ferraro}, \&
826: {Pancino}}]{bellazzini2001}
827: {Bellazzini}, M., {Ferraro}, F.~R., \& {Pancino}, E. 2001, \apj, 556, 635
828:
829: \bibitem[{{Belokurov} {et~al.}(2007){Belokurov}, {Zucker}, {Evans}, {Kleyna},
830: {Koposov}, {Hodgkin}, {Irwin}, {Gilmore}, {Wilkinson}, {Fellhauer},
831: {Bramich}, {Hewett}, {Vidrih}, {De Jong}, {Smith}, {Rix}, {Bell}, {Wyse},
832: {Newberg}, {Mayeur}, {Yanny}, {Rockosi}, {Gnedin}, {Schneider}, {Beers},
833: {Barentine}, {Brewington}, {Brinkmann}, {Harvanek}, {Kleinman}, {Krzesinski},
834: {Long}, {Nitta}, \& {Snedden}}]{belokurov2007}
835: {Belokurov}, V., {Zucker}, D.~B., {Evans}, N.~W., {Kleyna}, J.~T., {Koposov},
836: S., {Hodgkin}, S.~T., {Irwin}, M.~J., {Gilmore}, G., {Wilkinson}, M.~I.,
837: {Fellhauer}, M., {Bramich}, D.~M., {Hewett}, P.~C., {Vidrih}, S., {De Jong},
838: J.~T.~A., {Smith}, J.~A., {Rix}, H.-W., {Bell}, E.~F., {Wyse}, R.~F.~G.,
839: {Newberg}, H.~J., {Mayeur}, P.~A., {Yanny}, B., {Rockosi}, C.~M., {Gnedin},
840: O.~Y., {Schneider}, D.~P., {Beers}, T.~C., {Barentine}, J.~C., {Brewington},
841: H., {Brinkmann}, J., {Harvanek}, M., {Kleinman}, S.~J., {Krzesinski}, J.,
842: {Long}, D., {Nitta}, A., \& {Snedden}, S.~A. 2007, \apj, 654, 897
843:
844: \bibitem[{{Belokurov} {et~al.}(2006){Belokurov}, {Zucker}, {Evans},
845: {Wilkinson}, {Irwin}, {Hodgkin}, {Bramich}, {Irwin}, {Gilmore}, {Willman},
846: {Vidrih}, {Newberg}, {Wyse}, {Fellhauer}, {Hewett}, {Cole}, {Bell}, {Beers},
847: {Rockosi}, {Yanny}, {Grebel}, {Schneider}, {Lupton}, {Barentine},
848: {Brewington}, {Brinkmann}, {Harvanek}, {Kleinman}, {Krzesinski}, {Long},
849: {Nitta}, {Smith}, \& {Snedden}}]{belokurov2006}
850: {Belokurov}, V., {Zucker}, D.~B., {Evans}, N.~W., {Wilkinson}, M.~I., {Irwin},
851: M.~J., {Hodgkin}, S., {Bramich}, D.~M., {Irwin}, J.~M., {Gilmore}, G.,
852: {Willman}, B., {Vidrih}, S., {Newberg}, H.~J., {Wyse}, R.~F.~G., {Fellhauer},
853: M., {Hewett}, P.~C., {Cole}, N., {Bell}, E.~F., {Beers}, T.~C., {Rockosi},
854: C.~M., {Yanny}, B., {Grebel}, E.~K., {Schneider}, D.~P., {Lupton}, R.,
855: {Barentine}, J.~C., {Brewington}, H., {Brinkmann}, J., {Harvanek}, M.,
856: {Kleinman}, S.~J., {Krzesinski}, J., {Long}, D., {Nitta}, A., {Smith}, J.~A.,
857: \& {Snedden}, S.~A. 2006, \apjl, 647, L111
858:
859: \bibitem[{{Bullock} {et~al.}(2001){Bullock}, {Kolatt}, {Sigad}, {Somerville},
860: {Kravtsov}, {Klypin}, {Primack}, \& {Dekel}}]{bullock2001}
861: {Bullock}, J.~S., {Kolatt}, T.~S., {Sigad}, Y., {Somerville}, R.~S.,
862: {Kravtsov}, A.~V., {Klypin}, A.~A., {Primack}, J.~R., \& {Dekel}, A. 2001,
863: \mnras, 321, 559
864:
865: \bibitem[{{Bullock} {et~al.}(2000){Bullock}, {Kravtsov}, \&
866: {Weinberg}}]{bullock2000}
867: {Bullock}, J.~S., {Kravtsov}, A.~V., \& {Weinberg}, D.~H. 2000, \apj, 539, 517
868:
869: \bibitem[{{Chapman} {et~al.}(2006){Chapman}, {Ibata}, {Lewis}, {Ferguson},
870: {Irwin}, {McConnachie}, \& {Tanvir}}]{chapman2006}
871: {Chapman}, S.~C., {Ibata}, R., {Lewis}, G.~F., {Ferguson}, A.~M.~N., {Irwin},
872: M., {McConnachie}, A., \& {Tanvir}, N. 2006, \apj, 653, 255
873:
874: \bibitem[{{Chapman} {et~al.}(2007){Chapman}, {Pe{\~n}arrubia}, {Ibata},
875: {McConnachie}, {Martin}, {Irwin}, {Blain}, {Lewis}, {Letarte}, {Lo},
876: {Ludlow}, \& {O'neil}}]{chapman2007}
877: {Chapman}, S.~C., {Pe{\~n}arrubia}, J., {Ibata}, R., {McConnachie}, A.,
878: {Martin}, N., {Irwin}, M., {Blain}, A., {Lewis}, G.~F., {Letarte}, B., {Lo},
879: K., {Ludlow}, A., \& {O'neil}, K. 2007, \apjl, 662, L79
880:
881: \bibitem[{{Dekel} \& {Silk}(1986)}]{dekel1986}
882: {Dekel}, A. \& {Silk}, J. 1986, \apj, 303, 39
883:
884: \bibitem[{{Dekel} \& {Woo}(2003)}]{dekel2003}
885: {Dekel}, A. \& {Woo}, J. 2003, \mnras, 344, 1131
886:
887: \bibitem[{{Diemand} {et~al.}(2007){Diemand}, {Kuhlen}, \&
888: {Madau}}]{diemand2007}
889: {Diemand}, J., {Kuhlen}, M., \& {Madau}, P. 2007, \apj, 667, 859
890:
891: \bibitem[{{Einasto} {et~al.}(1974){Einasto}, {Saar}, {Kaasik}, \&
892: {Chernin}}]{einasto1974}
893: {Einasto}, J., {Saar}, E., {Kaasik}, A., \& {Chernin}, A.~D. 1974, \nat, 252,
894: 111
895:
896: \bibitem[{{Ferguson} {et~al.}(2000){Ferguson}, {Gallagher}, \&
897: {Wyse}}]{ferguson2000}
898: {Ferguson}, A.~M.~N., {Gallagher}, J.~S., \& {Wyse}, R.~F.~G. 2000, \aj, 120,
899: 821
900:
901: \bibitem[{{Ferguson} {et~al.}(2002){Ferguson}, {Irwin}, {Ibata}, {Lewis}, \&
902: {Tanvir}}]{ferguson2002}
903: {Ferguson}, A.~M.~N., {Irwin}, M.~J., {Ibata}, R.~A., {Lewis}, G.~F., \&
904: {Tanvir}, N.~R. 2002, \aj, 124, 1452
905:
906: \bibitem[{{Gilmore} {et~al.}(2007){Gilmore}, {Wilkinson}, {Wyse}, {Kleyna},
907: {Koch}, {Evans}, \& {Grebel}}]{gilmore2007}
908: {Gilmore}, G., {Wilkinson}, M.~I., {Wyse}, R.~F.~G., {Kleyna}, J.~T., {Koch},
909: A., {Evans}, N.~W., \& {Grebel}, E.~K. 2007, \apj, 663, 948
910:
911: \bibitem[{{Huxor} {et~al.}(2008){Huxor}, {Tanvir}, {Ferguson}, {Irwin},
912: {Ibata}, {Bridges}, \& {Lewis}}]{huxor2008}
913: {Huxor}, A.~P., {Tanvir}, N.~R., {Ferguson}, A.~M.~N., {Irwin}, M.~J., {Ibata},
914: R., {Bridges}, T., \& {Lewis}, G.~F. 2008, ArXiv: 0801.0002, 801
915:
916: \bibitem[{{Huxor} {et~al.}(2005){Huxor}, {Tanvir}, {Irwin}, {Ibata}, {Collett},
917: {Ferguson}, {Bridges}, \& {Lewis}}]{huxor2005}
918: {Huxor}, A.~P., {Tanvir}, N.~R., {Irwin}, M.~J., {Ibata}, R., {Collett}, J.~L.,
919: {Ferguson}, A.~M.~N., {Bridges}, T., \& {Lewis}, G.~F. 2005, \mnras, 360,
920: 1007
921:
922: \bibitem[{{Ibata} {et~al.}(2007){Ibata}, {Martin}, {Irwin}, {Chapman},
923: {Ferguson}, {Lewis}, \& {McConnachie}}]{ibata2007}
924: {Ibata}, R., {Martin}, N.~F., {Irwin}, M., {Chapman}, S., {Ferguson}, A.~M.~N.,
925: {Lewis}, G.~F., \& {McConnachie}, A.~W. 2007, \apj, 671, 1591
926:
927: \bibitem[{{Ibata} {et~al.}(1994){Ibata}, {Gilmore}, \& {Irwin}}]{ibata1994}
928: {Ibata}, R.~A., {Gilmore}, G., \& {Irwin}, M.~J. 1994, \nat, 370, 194
929:
930: \bibitem[{{Irwin} \& {Hatzidimitriou}(1995)}]{irwin1995}
931: {Irwin}, M. \& {Hatzidimitriou}, D. 1995, \mnras, 277, 1354
932:
933: \bibitem[{{Irwin} \& {Lewis}(2001)}]{irwin2001}
934: {Irwin}, M. \& {Lewis}, J. 2001, New Astronomy Review, 45, 105
935:
936: \bibitem[{{Irwin}(1994)}]{irwin1994}
937: {Irwin}, M.~J. 1994, in ESO/OHP workshop No. 49 on Dwarf Galaxies, ed. G
938: Meylan. and P. Prugniel., 27
939:
940: \bibitem[{{Irwin} {et~al.}(2007){Irwin}, {Belokurov}, {Evans}, {Ryan-Weber},
941: {de Jong}, {Koposov}, {Zucker}, {Hodgkin}, {Gilmore}, {Prema}, {Hebb},
942: {Begum}, {Fellhauer}, {Hewett}, {Kennicutt}, {Wilkinson}, {Bramich},
943: {Vidrih}, {Rix}, {Beers}, {Barentine}, {Brewington}, {Harvanek},
944: {Krzesinski}, {Long}, {Nitta}, \& {Snedden}}]{irwin2007}
945: {Irwin}, M.~J., {Belokurov}, V., {Evans}, N.~W., {Ryan-Weber}, E.~V., {de
946: Jong}, J.~T.~A., {Koposov}, S., {Zucker}, D.~B., {Hodgkin}, S.~T., {Gilmore},
947: G., {Prema}, P., {Hebb}, L., {Begum}, A., {Fellhauer}, M., {Hewett}, P.~C.,
948: {Kennicutt}, Jr., R.~C., {Wilkinson}, M.~I., {Bramich}, D.~M., {Vidrih}, S.,
949: {Rix}, H.-W., {Beers}, T.~C., {Barentine}, J.~C., {Brewington}, H.,
950: {Harvanek}, M., {Krzesinski}, J., {Long}, D., {Nitta}, A., \& {Snedden},
951: S.~A. 2007, \apjl, 656, L13
952:
953: \bibitem[{{Irwin} {et~al.}(2008){Irwin}, {Ferguson}, {Huxor}, {Tanvir},
954: {Ibata}, \& {Lewis}}]{irwin2008}
955: {Irwin}, M.~J., {Ferguson}, A.~M.~N., {Huxor}, A.~P., {Tanvir}, N.~R., {Ibata},
956: R.~A., \& {Lewis}, G.~F. 2008, \apjl, 676, L17
957:
958: \bibitem[{{Kalirai} {et~al.}(2006){Kalirai}, {Gilbert}, {Guhathakurta},
959: {Majewski}, {Ostheimer}, {Rich}, {Cooper}, {Reitzel}, \&
960: {Patterson}}]{kalirai2006b}
961: {Kalirai}, J.~S., {Gilbert}, K.~M., {Guhathakurta}, P., {Majewski}, S.~R.,
962: {Ostheimer}, J.~C., {Rich}, R.~M., {Cooper}, M.~C., {Reitzel}, D.~B., \&
963: {Patterson}, R.~J. 2006, \apj, 648, 389
964:
965: \bibitem[{{Karachentsev} \& {Karachentseva}(1999)}]{karachentsev1999}
966: {Karachentsev}, I.~D. \& {Karachentseva}, V.~E. 1999, \aap, 341, 355
967:
968: \bibitem[{{Klypin} {et~al.}(1999){Klypin}, {Kravtsov}, {Valenzuela}, \&
969: {Prada}}]{klypin1999}
970: {Klypin}, A., {Kravtsov}, A.~V., {Valenzuela}, O., \& {Prada}, F. 1999, \apj,
971: 522, 82
972:
973: \bibitem[{{Koch} \& {Grebel}(2006)}]{koch2006}
974: {Koch}, A. \& {Grebel}, E.~K. 2006, \aj, 131, 1405
975:
976: \bibitem[{{Kopylov} {et~al.}(2008){Kopylov}, {Tikhonov}, {Fabrika},
977: {Drozdovsky}, \& {Valeev}}]{kopylov2008}
978: {Kopylov}, A.~I., {Tikhonov}, N.~A., {Fabrika}, S., {Drozdovsky}, I., \&
979: {Valeev}, A.~F. 2008, ArXiv e-prints, 803
980:
981: \bibitem[{{Kravtsov} {et~al.}(2004){Kravtsov}, {Gnedin}, \&
982: {Klypin}}]{kravtsov2004}
983: {Kravtsov}, A.~V., {Gnedin}, O.~Y., \& {Klypin}, A.~A. 2004, \apj, 609, 482
984:
985: \bibitem[{{Krismer} {et~al.}(1995){Krismer}, {Tully}, \& {Gioia}}]{krismer1995}
986: {Krismer}, M., {Tully}, R.~B., \& {Gioia}, I.~M. 1995, \aj, 110, 1584
987:
988: \bibitem[{{Kroupa} {et~al.}(2005){Kroupa}, {Theis}, \& {Boily}}]{kroupa2005}
989: {Kroupa}, P., {Theis}, C., \& {Boily}, C.~M. 2005, \aap, 431, 517
990:
991: \bibitem[{{Lee} {et~al.}(1993){Lee}, {Freedman}, \& {Madore}}]{lee1993a}
992: {Lee}, M.~G., {Freedman}, W.~L., \& {Madore}, B.~F. 1993, \apj, 417, 553
993:
994: \bibitem[{{Lynden-Bell}(1976)}]{lyndenbell1976}
995: {Lynden-Bell}, D. 1976, \mnras, 174, 695
996:
997: \bibitem[{{Lynden-Bell}(1982)}]{lyndenbell1982}
998: ---. 1982, The Observatory, 102, 202
999:
1000: \bibitem[{{Majewski} {et~al.}(2007){Majewski}, {Beaton}, {Patterson},
1001: {Kalirai}, {Geha}, {Mu{\~n}oz}, {Seigar}, {Guhathakurta}, {Gilbert}, {Rich},
1002: {Bullock}, \& {Reitzel}}]{majewski2007}
1003: {Majewski}, S.~R., {Beaton}, R.~L., {Patterson}, R.~J., {Kalirai}, J.~S.,
1004: {Geha}, M.~C., {Mu{\~n}oz}, R.~R., {Seigar}, M.~S., {Guhathakurta}, P.,
1005: {Gilbert}, K.~M., {Rich}, R.~M., {Bullock}, J.~S., \& {Reitzel}, D.~B. 2007,
1006: \apjl, 670, L9
1007:
1008: \bibitem[{{Majewski} {et~al.}(2003){Majewski}, {Skrutskie}, {Weinberg}, \&
1009: {Ostheimer}}]{majewski2003}
1010: {Majewski}, S.~R., {Skrutskie}, M.~F., {Weinberg}, M.~D., \& {Ostheimer}, J.~C.
1011: 2003, \apj, 599, 1082
1012:
1013: \bibitem[{{Martin} {et~al.}(2008){Martin}, {de Jong}, \& {Rix}}]{martin2008}
1014: {Martin}, N.~F., {de Jong}, J.~T.~A., \& {Rix}, H.-W. 2008, ArXiv:0805.2945,
1015: 805
1016:
1017: \bibitem[{{Martin} {et~al.}(2007){Martin}, {Ibata}, \& {Irwin}}]{martin2007}
1018: {Martin}, N.~F., {Ibata}, R.~A., \& {Irwin}, M. 2007, \apjl, 668, L123
1019:
1020: \bibitem[{{Martin} {et~al.}(2006){Martin}, {Ibata}, {Irwin}, {Chapman},
1021: {Lewis}, {Ferguson}, {Tanvir}, \& {McConnachie}}]{martin2006}
1022: {Martin}, N.~F., {Ibata}, R.~A., {Irwin}, M.~J., {Chapman}, S., {Lewis}, G.~F.,
1023: {Ferguson}, A.~M.~N., {Tanvir}, N., \& {McConnachie}, A.~W. 2006, \mnras,
1024: 371, 1983
1025:
1026: \bibitem[{{Mateo}(1998)}]{mateo1998a}
1027: {Mateo}, M.~L. 1998, \araa, 36, 435
1028:
1029: \bibitem[{{Mayer} {et~al.}(2006){Mayer}, {Mastropietro}, {Wadsley}, {Stadel},
1030: \& {Moore}}]{mayer2006}
1031: {Mayer}, L., {Mastropietro}, C., {Wadsley}, J., {Stadel}, J., \& {Moore}, B.
1032: 2006, \mnras, 369, 1021
1033:
1034: \bibitem[{{McConnachie} {et~al.}(2007{\natexlab{a}}){McConnachie}, {Arimoto},
1035: \& {Irwin}}]{mcconnachie2007a}
1036: {McConnachie}, A.~W., {Arimoto}, N., \& {Irwin}, M. 2007{\natexlab{a}}, \mnras,
1037: 379, 379
1038:
1039: \bibitem[{{McConnachie} \& {Irwin}(2006{\natexlab{a}})}]{mcconnachie2006b}
1040: {McConnachie}, A.~W. \& {Irwin}, M.~J. 2006{\natexlab{a}}, \mnras, 365, 1263
1041:
1042: \bibitem[{{McConnachie} \& {Irwin}(2006{\natexlab{b}})}]{mcconnachie2006a}
1043: ---. 2006{\natexlab{b}}, \mnras, 365, 902
1044:
1045: \bibitem[{{McConnachie} {et~al.}(2004){McConnachie}, {Irwin}, {Ferguson},
1046: {Ibata}, {Lewis}, \& {Tanvir}}]{mcconnachie2004a}
1047: {McConnachie}, A.~W., {Irwin}, M.~J., {Ferguson}, A.~M.~N., {Ibata}, R.~A.,
1048: {Lewis}, G.~F., \& {Tanvir}, N. 2004, \mnras, 350, 243
1049:
1050: \bibitem[{{McConnachie} {et~al.}(2005){McConnachie}, {Irwin}, {Ferguson},
1051: {Ibata}, {Lewis}, \& {Tanvir}}]{mcconnachie2005a}
1052: ---. 2005, \mnras, 356, 979
1053:
1054: \bibitem[{{McConnachie} {et~al.}(2007{\natexlab{b}}){McConnachie}, {Venn},
1055: {Irwin}, {Young}, \& {Geehan}}]{mcconnachie2007c}
1056: {McConnachie}, A.~W., {Venn}, K.~A., {Irwin}, M.~J., {Young}, L.~M., \&
1057: {Geehan}, J.~J. 2007{\natexlab{b}}, \apjl, 671, L33
1058:
1059: \bibitem[{{Merrett} {et~al.}(2006){Merrett}, {Merrifield}, {Douglas},
1060: {Kuijken}, {Romanowsky}, {Napolitano}, {Arnaboldi}, {Capaccioli}, {Freeman},
1061: {Gerhard}, {Coccato}, {Carter}, {Evans}, {Wilkinson}, {Halliday}, \&
1062: {Bridges}}]{merrett2006}
1063: {Merrett}, H.~R., {Merrifield}, M.~R., {Douglas}, N.~G., {Kuijken}, K.,
1064: {Romanowsky}, A.~J., {Napolitano}, N.~R., {Arnaboldi}, M., {Capaccioli}, M.,
1065: {Freeman}, K.~C., {Gerhard}, O., {Coccato}, L., {Carter}, D., {Evans}, N.~W.,
1066: {Wilkinson}, M.~I., {Halliday}, C., \& {Bridges}, T.~J. 2006, \mnras, 369,
1067: 120
1068:
1069: \bibitem[{{Metz} {et~al.}(2007){Metz}, {Kroupa}, \& {Jerjen}}]{metz2007}
1070: {Metz}, M., {Kroupa}, P., \& {Jerjen}, H. 2007, \mnras, 374, 1125
1071:
1072: \bibitem[{{Moore} {et~al.}(1999){Moore}, {Quinn}, {Governato}, {Stadel}, \&
1073: {Lake}}]{moore1999}
1074: {Moore}, B., {Quinn}, T., {Governato}, F., {Stadel}, J., \& {Lake}, G. 1999,
1075: \mnras, 310, 1147
1076:
1077: \bibitem[{{Morrison} {et~al.}(2003){Morrison}, {Harding}, {Hurley-Keller}, \&
1078: {Jacoby}}]{morrison2003}
1079: {Morrison}, H.~L., {Harding}, P., {Hurley-Keller}, D., \& {Jacoby}, G. 2003,
1080: \apjl, 596, L183
1081:
1082: \bibitem[{{Pe{\~n}arrubia} {et~al.}(2008{\natexlab{a}}){Pe{\~n}arrubia},
1083: {McConnachie}, \& {Navarro}}]{penarrubia2008a}
1084: {Pe{\~n}arrubia}, J., {McConnachie}, A.~W., \& {Navarro}, J.~F.
1085: 2008{\natexlab{a}}, \apj, 672, 904
1086:
1087: \bibitem[{{Pe{\~n}arrubia} {et~al.}(2008{\natexlab{b}}){Pe{\~n}arrubia},
1088: {Navarro}, \& {McConnachie}}]{penarrubia2008b}
1089: {Pe{\~n}arrubia}, J., {Navarro}, J.~F., \& {McConnachie}, A.~W.
1090: 2008{\natexlab{b}}, \apj, 673, 226
1091:
1092: \bibitem[{{Sakai} {et~al.}(1999){Sakai}, {Madore}, \& {Freedman}}]{sakai1999}
1093: {Sakai}, S., {Madore}, B.~F., \& {Freedman}, W.~L. 1999, \apj, 511, 671
1094:
1095: \bibitem[{{Salaris} \& {Cassisi}(1997)}]{salaris1997}
1096: {Salaris}, M. \& {Cassisi}, S. 1997, \mnras, 289, 406
1097:
1098: \bibitem[{{Sales} {et~al.}(2007){Sales}, {Navarro}, {Abadi}, \&
1099: {Steinmetz}}]{sales2007}
1100: {Sales}, L.~V., {Navarro}, J.~F., {Abadi}, M.~G., \& {Steinmetz}, M. 2007,
1101: \mnras, 379, 1475
1102:
1103: \bibitem[{{Schlegel} {et~al.}(1998){Schlegel}, {Finkbeiner}, \&
1104: {Davis}}]{schlegel1998}
1105: {Schlegel}, D.~J., {Finkbeiner}, D.~P., \& {Davis}, M. 1998, \apj, 500, 525
1106:
1107: \bibitem[{{Sharina} {et~al.}(1997){Sharina}, {Karachentsev}, \&
1108: {Tikhonov}}]{sharina1997}
1109: {Sharina}, M.~E., {Karachentsev}, I.~D., \& {Tikhonov}, N.~A. 1997, Astronomy
1110: Letters, 23, 373
1111:
1112: \bibitem[{{Skillman}(2005)}]{skillman2005}
1113: {Skillman}, E.~D. 2005, New Astronomy Review, 49, 453
1114:
1115: \bibitem[{{Smecker-Hane} {et~al.}(1994){Smecker-Hane}, {Stetson}, {Hesser}, \&
1116: {Lehnert}}]{smeckerhane1994}
1117: {Smecker-Hane}, T.~A., {Stetson}, P.~B., {Hesser}, J.~E., \& {Lehnert}, M.~D.
1118: 1994, \aj, 108, 507
1119:
1120: \bibitem[{{Strigari} {et~al.}(2007{\natexlab{a}}){Strigari}, {Bullock},
1121: {Kaplinghat}, {Diemand}, {Kuhlen}, \& {Madau}}]{strigari2007b}
1122: {Strigari}, L.~E., {Bullock}, J.~S., {Kaplinghat}, M., {Diemand}, J., {Kuhlen},
1123: M., \& {Madau}, P. 2007{\natexlab{a}}, \apj, 669, 676
1124:
1125: \bibitem[{{Strigari} {et~al.}(2007{\natexlab{b}}){Strigari}, {Koushiappas},
1126: {Bullock}, \& {Kaplinghat}}]{strigari2007a}
1127: {Strigari}, L.~E., {Koushiappas}, S.~M., {Bullock}, J.~S., \& {Kaplinghat}, M.
1128: 2007{\natexlab{b}}, \prd, 75, 083526
1129:
1130: \bibitem[{{VandenBerg} {et~al.}(2006){VandenBerg}, {Bergbusch}, \&
1131: {Dowler}}]{vandenberg2006}
1132: {VandenBerg}, D.~A., {Bergbusch}, P.~A., \& {Dowler}, P.~D. 2006, \apjs, 162,
1133: 375
1134:
1135: \bibitem[{{VandenBerg} \& {Clem}(2003)}]{vandenberg2003}
1136: {VandenBerg}, D.~A. \& {Clem}, J.~L. 2003, \aj, 126, 778
1137:
1138: \bibitem[{{Walsh} {et~al.}(2007){Walsh}, {Jerjen}, \& {Willman}}]{walsh2007}
1139: {Walsh}, S.~M., {Jerjen}, H., \& {Willman}, B. 2007, \apjl, 662, L83
1140:
1141: \bibitem[{{Whiting} {et~al.}(1999){Whiting}, {Hau}, \& {Irwin}}]{whiting1999}
1142: {Whiting}, A.~B., {Hau}, G.~K.~T., \& {Irwin}, M. 1999, \aj, 118, 2767
1143:
1144: \bibitem[{{Whiting} {et~al.}(1997){Whiting}, {Irwin}, \& {Hau}}]{whiting1997}
1145: {Whiting}, A.~B., {Irwin}, M.~J., \& {Hau}, G.~K.~T. 1997, \aj, 114, 996
1146:
1147: \bibitem[{{Willman} {et~al.}(2005){Willman}, {Dalcanton}, {Martinez-Delgado},
1148: {West}, {Blanton}, {Hogg}, {Barentine}, {Brewington}, {Harvanek}, {Kleinman},
1149: {Krzesinski}, {Long}, {Neilsen}, {Nitta}, \& {Snedden}}]{willman2005}
1150: {Willman}, B., {Dalcanton}, J.~J., {Martinez-Delgado}, D., {West}, A.~A.,
1151: {Blanton}, M.~R., {Hogg}, D.~W., {Barentine}, J.~C., {Brewington}, H.~J.,
1152: {Harvanek}, M., {Kleinman}, S.~J., {Krzesinski}, J., {Long}, D., {Neilsen},
1153: Jr., E.~H., {Nitta}, A., \& {Snedden}, S.~A. 2005, \apjl, 626, L85
1154:
1155: \bibitem[{{Willman} {et~al.}(2004){Willman}, {Governato}, {Dalcanton}, {Reed},
1156: \& {Quinn}}]{willman2004}
1157: {Willman}, B., {Governato}, F., {Dalcanton}, J.~J., {Reed}, D., \& {Quinn}, T.
1158: 2004, \mnras, 353, 639
1159:
1160: \bibitem[{{Willman} {et~al.}(2006){Willman}, {Masjedi}, {Hogg}, {Dalcanton},
1161: {Martinez-Delgado}, {Blanton}, {West}, {Dotter}, \& {Chaboyer}}]{willman2006}
1162: {Willman}, B., {Masjedi}, M., {Hogg}, D.~W., {Dalcanton}, J.~J.,
1163: {Martinez-Delgado}, D., {Blanton}, M., {West}, A.~A., {Dotter}, A., \&
1164: {Chaboyer}, B. 2006, astro-ph/0603486
1165:
1166: \bibitem[{{Zucker} {et~al.}(2006){Zucker}, {Belokurov}, {Evans}, {Wilkinson},
1167: {Irwin}, {Sivarani}, {Hodgkin}, {Bramich}, {Irwin}, {Gilmore}, {Willman},
1168: {Vidrih}, {Fellhauer}, {Hewett}, {Beers}, {Bell}, {Grebel}, {Schneider},
1169: {Newberg}, {Wyse}, {Rockosi}, {Yanny}, {Lupton}, {Smith}, {Barentine},
1170: {Brewington}, {Brinkmann}, {Harvanek}, {Kleinman}, {Krzesinski}, {Long},
1171: {Nitta}, \& {Snedden}}]{zucker2006b}
1172: {Zucker}, D.~B., {Belokurov}, V., {Evans}, N.~W., {Wilkinson}, M.~I., {Irwin},
1173: M.~J., {Sivarani}, T., {Hodgkin}, S., {Bramich}, D.~M., {Irwin}, J.~M.,
1174: {Gilmore}, G., {Willman}, B., {Vidrih}, S., {Fellhauer}, M., {Hewett}, P.~C.,
1175: {Beers}, T.~C., {Bell}, E.~F., {Grebel}, E.~K., {Schneider}, D.~P.,
1176: {Newberg}, H.~J., {Wyse}, R.~F.~G., {Rockosi}, C.~M., {Yanny}, B., {Lupton},
1177: R., {Smith}, J.~A., {Barentine}, J.~C., {Brewington}, H., {Brinkmann}, J.,
1178: {Harvanek}, M., {Kleinman}, S.~J., {Krzesinski}, J., {Long}, D., {Nitta}, A.,
1179: \& {Snedden}, S.~A. 2006, \apjl, 643, L103
1180:
1181: \bibitem[{{Zucker} {et~al.}(2004){Zucker}, {Kniazev}, {Bell},
1182: {Mart{\'{\i}}nez-Delgado}, {Grebel}, {Rix}, {Rockosi}, {Holtzman},
1183: {Walterbos}, {Annis}, {York}, {Ivezi{\' c}}, {Brinkmann}, {Brewington},
1184: {Harvanek}, {Hennessy}, {Kleinman}, {Krzesinski}, {Long}, {Newman}, {Nitta},
1185: \& {Snedden}}]{zucker2004a}
1186: {Zucker}, D.~B., {Kniazev}, A.~Y., {Bell}, E.~F., {Mart{\'{\i}}nez-Delgado},
1187: D., {Grebel}, E.~K., {Rix}, H., {Rockosi}, C.~M., {Holtzman}, J.~A.,
1188: {Walterbos}, R.~A.~M., {Annis}, J., {York}, D.~G., {Ivezi{\' c}}, {\v Z}.,
1189: {Brinkmann}, J., {Brewington}, H., {Harvanek}, M., {Hennessy}, G.,
1190: {Kleinman}, S.~J., {Krzesinski}, J., {Long}, D., {Newman}, P.~R., {Nitta},
1191: A., \& {Snedden}, S.~A. 2004, \apjl, 612, L121
1192:
1193: \bibitem[{{Zucker} {et~al.}(2007){Zucker}, {Kniazev},
1194: {Mart{\'{\i}}nez-Delgado}, {Bell}, {Rix}, {Grebel}, {Holtzman}, {Walterbos},
1195: {Rockosi}, {York}, {Barentine}, {Brewington}, {Brinkmann}, {Harvanek},
1196: {Kleinman}, {Krzesinski}, {Long}, {Neilsen}, {Nitta}, \&
1197: {Snedden}}]{zucker2007}
1198: {Zucker}, D.~B., {Kniazev}, A.~Y., {Mart{\'{\i}}nez-Delgado}, D., {Bell},
1199: E.~F., {Rix}, H.-W., {Grebel}, E.~K., {Holtzman}, J.~A., {Walterbos},
1200: R.~A.~M., {Rockosi}, C.~M., {York}, D.~G., {Barentine}, J.~C., {Brewington},
1201: H., {Brinkmann}, J., {Harvanek}, M., {Kleinman}, S.~J., {Krzesinski}, J.,
1202: {Long}, D., {Neilsen}, Jr., E.~H., {Nitta}, A., \& {Snedden}, S.~A. 2007,
1203: \apjl, 659, L21
1204:
1205: \end{thebibliography}
1206:
1207: \clearpage
1208:
1209: \begin{table*}
1210: \begin{center}
1211: \begin{tabular}{lccc}
1212: \hline
1213: & Andromeda XVIII & Andromeda XIX & Andromeda XX\\
1214: \hline
1215: $\alpha$ (J2000) & 00\,h\,02\,m\,14.5\,s~~($\pm 10''$) & 00\,h\,19\,m\,32.1\,s~~($\pm 10''$) & 00\,h\,07\,m\,30.7\,s~~($\pm 15''$) \\
1216: $\delta$ (J2000) & +45$^\circ$\,05$^\prime$\,20$''$ ($\pm 10''$) & +35$^\circ$\,02$^\prime$\,37.1$''$~~($\pm 10''$) & +35$^\circ$\,07$^\prime$\,56.4$''$~~($\pm 15''$) \\
1217: $(l, b)$ & $(113.9^\circ, -16.9^\circ)$ & $(115.6^\circ, -27.4^\circ)$ & $(112.9^\circ, -26.9^\circ)$\\
1218: $E(B - V)$ & 0.104 & 0.062 & 0.058\\
1219: $I_{o,trgb}$ & $21.62 \pm 0.05$ & $20.81 \pm 0.05$ & $20.48^{+0.73}_{-0.20}$\\
1220: $(m - M)_o$ & $25.66 \pm 0.13$ & $24.85 \pm 0.13$ & $24.52^{+0.74}_{-0.24}$\\
1221: Distance & $1355 \pm 88$\,kpc & $933 \pm 61$\,kpc & $802^{+297}_{-96}$\,kpc\\
1222: $r_{M31}$ & $\sim 589$\,kpc & $\sim 187$\,kpc & $\sim 129$\,kpc\\
1223: ${\rm [Fe/H]}$ & $-1.8 \pm 0.1$ & $-1.9 \pm 0.1$ & $-1.5 \pm 0.1$\\
1224: $IQR$ & $0.5$ & $0.4$ & $0.5$\\
1225: $r_h$ (arcmins) & $0.92^{+0.05}_{-0.06}$ & $6.2 \pm 0.1$ & $0.53^{+0.14}_{-0.04}$\\
1226: $r_h$ (pc) & $363^{+31}_{-33}$ & $1683 \pm 113$ & $124^{+56}_{-18}$\\
1227: PA (N to E) & $0$ & $(37^{+4}_{-8})^\circ$ & $(80 \pm 20)^\circ$\\
1228: $\epsilon = 1 - b/a$ & $0$ & $0.17 \pm 0.02$ & $0.3 \pm 0.15$\\
1229: $m_v$ & $\le 16.0$ & $15.6 \pm 0.6$ & $18.2 \pm 0.8$ \\
1230: $M_V$ & $\le -9.7$ & $-9.3 \pm 0.6$ & $-6.3^{+1.0}_{-0.8}$\\
1231: $S_o$ & $\le 25.6$ & $29.3 \pm 0.7$ & $26.2 \pm 0.8$\\
1232: \hline
1233: \end{tabular}
1234: \end{center}
1235: \caption{Properties of Andromeda XVIII, XIX and XX}
1236: \end{table*}
1237:
1238: \clearpage
1239:
1240:
1241:
1242: \begin{figure}
1243: \begin{center}
1244: \includegraphics[angle=270, width=14.cm]{f1.ps}
1245: \caption{A tangent plane projection of the CFHT/MegaPrime survey
1246: area around M31. The inner ellipse represents a disk of
1247: inclination 77 degrees and radius 2 degrees (27\,kpc), the
1248: approximate edge of the regular M31 disk. The outer ellipse shows
1249: a 55\,kpc radius ellipse flattened to $c/a = 0.6$, the limit of
1250: the original INT/WFC survey (\citealt{ferguson2002}). Major and
1251: minor axes of M31 are indicated. The inner and outer blue dashed
1252: circles show maximum projected distances of 100\,kpc and 150\,kpc
1253: from the center of M31, respectively. Red hatched fields show the
1254: location of our extant imaging of the south-east quadrant of M31
1255: (\citealt{ibata2007}). Magenta fields show the location of fields
1256: for our ongoing survey of the south-west quadrant of M31 (solid
1257: lines denote observed fields, dotted lines denote fields still to
1258: be observed). Black stars show the locations of various known M31
1259: satellite galaxies, and open stars show the positions of the three
1260: new dwarf galaxies presented herein.}
1261: \end{center}
1262: \end{figure}
1263:
1264: \clearpage
1265:
1266: \begin{figure*}
1267: \begin{center}
1268: \includegraphics[angle=270, width=5.cm]{f2a.ps}
1269: \includegraphics[angle=270, width=5.cm]{f2b.ps}
1270: \includegraphics[angle=270, width=5.cm]{f2c.ps}
1271: \includegraphics[angle=270, width=5.cm]{f2d.ps}
1272: \includegraphics[angle=270, width=5.cm]{f2e.ps}
1273: \includegraphics[angle=270, width=5.cm]{f2f.ps}
1274: \caption{The top panels show a comparison of the $i_o$ versus $(g
1275: - i)_o$ colour-magnitude diagrams of the three newly discovered
1276: Local Group galaxies, Andromeda~XVIII (left), XIX (middle) and XX
1277: (right), where all stars lying within 2 half-light radii from the
1278: center of each galaxy have been plotted (corresponding to 1.8,
1279: 12.4 and 1 arcmins, respectively). The bottom panels are reference
1280: fields probing an equivalent area offset from each galaxy by
1281: several half light radii. A red giant branch (RGB) is visible in
1282: each galaxy, although in the case of Andromeda~XX it is very
1283: sparsely populated. None of the galaxies display any evidence for
1284: bright blue stars (either bright main sequence or blue-loop),
1285: indicative of a young population, and in this respect they
1286: resemble the typical stellar populations of dSph galaxies. The
1287: faint blue objects centered around $i_o \sim 25.2$ with a mean
1288: colour of $(g - i)_o \sim 0.5$ in the Andromeda~XIX CMD may be a
1289: horizontal branch component, although contamination from
1290: misidentified galaxies is considerable in this region of
1291: colour-magnitude space.}
1292: \end{center}
1293: \end{figure*}
1294:
1295: \clearpage
1296:
1297: \begin{figure*}
1298: \begin{center}
1299: \includegraphics[angle=270, width=16.cm]{f3a.ps}\vspace{0.5cm}
1300: \includegraphics[angle=270, width=16.cm]{f3b.ps}\vspace{0.5cm}
1301: \includegraphics[angle=270, width=16.cm]{f3c.ps}
1302: \end{center}
1303: \end{figure*}
1304:
1305: \clearpage
1306:
1307: \begin{figure*}
1308: \begin{center}
1309: \caption{Each set of panels shows various properties of
1310: Andromeda~XVIII (top row), Andromeda~XIX (middle row) and
1311: Andromeda~XX (bottom row). Left panels: $I_o$ versus $(V - I)_o$
1312: colour magnitude diagram for each galaxy. Dashed lines define a
1313: colour-cut used to preferentially select stars associated with the
1314: dwarf. A 13\,Gyr isochrone with the representative metallicity of
1315: the dwarf from \cite{vandenberg2006}, shifted to the appropriate
1316: distance modulus, is overlaid on each CMD. Only stars within the
1317: dotted ellipses shown in the middle panels are plotted. Middle
1318: panels: tangent plane projections of the spatial distribution of
1319: stars in the vicinity of each dwarf. Only stars satisfying the
1320: colour cuts shown in the CMDs are plotted. Dashed lines show the
1321: edges of the CFHT/MegaPrime CCDs. Dashed ellipses mark two
1322: half-light radii from the center of each galaxy. For
1323: Andromeda~XVIII and XX, the dwarf galaxies are clearly visible as
1324: overdensities in the centers of each field, whereas Andromeda~XIX
1325: is more extended and diffuse and contours have been overlaid to
1326: more clearly define its structure. The first contour is set $3 -
1327: \sigma$ above the background, and subsequent contour levels
1328: increase by $1.5\,\sigma$ over the previous level. Right top
1329: panels: foreground-corrected, de-reddened, $I$-band luminosity
1330: functions of stars in each galaxy satisfying our colour and
1331: spatial cuts. Scaled reference field luminosity functions are
1332: shown as dotted lines. The estimated luminosity of the tip of the
1333: red giant branch is highlighted. Right bottom panels:
1334: foreground-corrected observed photometric metallicity distribution
1335: function derived using the technique detailed in
1336: \cite{mcconnachie2005a} using 13\,Gyr isochrones from
1337: \cite{vandenberg2006} with [$\alpha$/Fe] $= 0$]. Scaled reference
1338: field metallicity distribution functions are shown as dotted
1339: lines. The mean metallicity and metallicity spread (quantified
1340: using the inter-quartile range, IQR) for each galaxy is
1341: highlighted.}
1342: \end{center}
1343: \end{figure*}
1344:
1345: \clearpage
1346:
1347: \begin{figure}
1348: \begin{center}
1349: \includegraphics[angle=270, width=14.cm]{f4.ps}
1350: \caption{De-reddened $g$-band luminosity functions of stars within
1351: 2 half-light radii of Andromeda~XVIII, XIX and XX (top, middle and
1352: bottom panels, respectively). These luminosity functions go deeper
1353: than the CMDs shown previously since only detection in the
1354: $g$-band is required. Nearby reference fields, scaled by area, are
1355: shown as dotted lines in each panel. For reference, the horizontal
1356: branch of M31 has a magnitude of $g_o \sim 25.2$. In Andromeda~XIX
1357: and XX, we attribute the peak of stars at $g_o \sim 25.3$ and $g_o
1358: \sim 25.6$, respectively, to a detection of horizontal branch
1359: stars (indicated by the dashed lines). In Andromeda~XVIII, no
1360: feature attributable to the horizontal branch is visible, as
1361: expected from its larger distance. Thus these detections (and
1362: non-detection) are consistent with the distances derived for these
1363: galaxies via the TRGB analysis.}
1364: \end{center}
1365: \end{figure}
1366:
1367: \clearpage
1368:
1369: \begin{figure}
1370: \begin{center}
1371: \includegraphics[angle=180, width=9.cm]{f5a.ps}\vspace{1cm}
1372: \includegraphics[angle=180, width=9.cm]{f5b.ps}
1373: \caption{Top panel: The CFHT/MegaPrime $i-$band image of
1374: Andromeda~XVIII with linear scaling. Approximately $2.5 \times
1375: 1.2$\,arcmins in the vicinity of Andromeda~XVIII is shown. This
1376: galaxy lies in the south-west corner of one of the CCDs, and some
1377: of it remains hidden behind the large gap between the second and
1378: first rows of CFHT/MegaPrime mosaic. Unlike the majority of recent
1379: discoveries in the Local Group, Andromeda~XVIII is clearly visible
1380: based upon its resolved light. Bottom panel: A $10 \times
1381: 10$\,arcmins image centered on the coordinates of Andromeda~XVIII,
1382: with linear scaling, from the POSS2/UKSTU (Blue) survey, taken
1383: from the Digitized Sky Survey. Andromeda~XVIII is visible at the
1384: center. Some Galactic nebulosity is also present in this
1385: region. This galaxy is also visible on the original POSS1 (Blue)
1386: survey plates, and suggests that there may be other comparably
1387: bright galaxies within the Local Group which have so far eluded
1388: detection. In each panel, North is to the top and East is to the
1389: left.}
1390: \end{center}
1391: \end{figure}
1392:
1393: \clearpage
1394:
1395: \begin{figure*}
1396: \begin{center}
1397: \includegraphics[angle=270, width=5.cm]{f6a.ps}
1398: \includegraphics[angle=270, width=5.cm]{f6b.ps}
1399: \includegraphics[angle=270, width=5.cm]{f6c.ps}
1400: \caption{Radial profiles of Andromeda~XVIII, XIX and XX (left to
1401: right, respectively), derived in the same way as in
1402: \cite{mcconnachie2006b} using elliptical annuli with the position
1403: angles, ellipticities and centroids listed in Table~1. Overlaid on
1404: these profiles are the most probable exponential profiles derived
1405: using the same maximum likelihood technique.}
1406: \end{center}
1407: \end{figure*}
1408:
1409: \clearpage
1410:
1411: \begin{figure}
1412: \begin{center}
1413: \includegraphics[angle=270, width=14.cm]{f7.ps}
1414: \caption{The surroundings of Andromeda~XIX, shown as a tangent
1415: plane projection of the stellar density. The first two contour
1416: levels are $2$ and $3 - \sigma$ above the background, and then
1417: they increase by $1.5\,\sigma$ over the previous level. The major
1418: axis of M31 is shown as the solid black line, and the blue dashed
1419: line shows part of the circle which marks the $100$\,kpc boundary
1420: from the center of M31, as in Figure 1. Andromeda~XIX lies close
1421: to the major axis substructure identified in \cite{ibata2007},
1422: some of which can be seen in this plot. There is also some
1423: evidence of tidal features in the outskirts of Andromeda~XIX.}
1424: \end{center}
1425: \end{figure}
1426:
1427:
1428: \end{document}
1429:
1430: