1: \documentclass[12pt,preprint]{aastex}
2:
3: \def\plotfiddle#1#2#3#4#5#6#7{\centering \leavevmode
4: reference to#2{\rule{0pt}{#2}}
5: \special{psfile=#1 voffset=#7 hoffset=#6 vscale=#5 hscale=#4 angle=#3}}
6:
7: \def\sun{\hbox{$\odot$}}
8:
9:
10: \def\etal{{\it et~al.\ }}
11: % \tightenlines
12: % \mathwithsecnums
13: % \received{}
14: % \accepted{}
15: % \journalid{}
16: % \articleid{}{}
17:
18: \def\pasa{\ref@jnl{PASA}} % Publications of the ASA
19:
20: \lefthead{}
21: \righthead{}
22:
23: \begin{document}
24: \title{A Search for Stellar Populations in High Velocity Clouds}
25:
26: \author{M. H. Siegel\altaffilmark{1}, S. R. Majewski\altaffilmark{2,3},
27: C. Gallart\altaffilmark{4}, S. Sohn\altaffilmark{2,5}, W. E. Kunkel\altaffilmark{6}, R. Braun\altaffilmark{7}}
28:
29: \altaffiltext{1}{University of Texas -- McDonald Observatory, Austin, TX, 78712
30: (siegel@astro.as.utexas.edu).}
31: \altaffiltext{2}{University of Virginia, P.O. Box 3818, Charlottesville, VA,
32: 22903 (srm4n@virginia.edu,ss5fb@virginia.edu).}
33: \altaffiltext{3}{David and Lucile Packard Foundation Fellow.}
34: \altaffiltext{4}{Ram\'on y Cajal Fellow, Instituto de Astrof\'\i sica de Canarias
35: 38205 La Laguna, Tenerife, Spain, (carme@iac.es).}
36: \altaffiltext{5}{Present Address:
37: Korea Astronomy and Space Science Institute, 61-1, Whaam-Dong,
38: Youseong-Gu, Daejeon 305-348, Korea (tonysohn@kao.re.kr).}
39: \altaffiltext{6}{Las Campanas Observatory, Casilla 601, Las Serena, Chile (kunkel@jeito.lco.cl).}
40: \altaffiltext{7}{Netherlands Foundation for Research in Astronomy,
41: PO Box 2, 7990 AA Dwingeloo, The Netherlands (rbraun@nfra.nl).}
42:
43: % version date December 30, 2004
44:
45: \begin{abstract}
46: We report the results of a photometric search for giant stars associated with the cores of four
47: high velocity clouds (HVCs) -- two of which are compact HVCs -- using
48: the Las Campanas Du Pont 2.5 meter and Cerro Tololo Blanco 4 meter telescopes in combination with a
49: system of filters (Washington $M$, $T_2$ + $DDO51$) useful for identifying low
50: surface gravity, evolved stars. Identical observations of nearby control
51: fields provide a measure of the ``giant star'' background.
52: Our data reach $M_0=22$ for three of the HVCs and $M_0=21.25$ for the fourth, depths that
53: allow the detection of any giant stars within 600 kpc.
54: Although we identify a number of faint late-type giant star candidates, we find neither a coherent red giant branch
55: structure nor a clear excess of giant candidate counts in any HVC.
56: This indicates that the giant candidates are probably not related to the HVCs and are
57: more likely to be either random Milky Way giant stars or one of several classes of potential survey
58: contaminants. Echelle spectroscopy of the brightest giant
59: candidates in one HVC and its control field reveal radial velocities representative of the canonical Galactic stellar
60: populations. In addition to these null results, no evidence
61: of any young HVC stellar populations --- represented by blue main sequence stars --- is found,
62: a result consistent with previous searches. Our methodology, specifically designed to find faint
63: diffuse stellar populations, places the tightest upper limit yet on
64: the {\it total} stellar mass of HVCs of a few $10^5 M_{\sun}$.
65: \end{abstract}
66: \keywords{Local Group; Dwarf Galaxies; High Velocity Clouds}
67:
68: \section{Introduction}
69:
70: The high velocity clouds (HVCs) are a population of neutral hydrogen clouds with LSR velocities of 80-400
71: km s$^{-1}$ (Muller et al. 1963). Despite nearly four decades of study, the nature of the clouds is
72: still a matter of some debate (see, e.g., Wakker \& van Woerden 1997; Putman \& Gibson 1999) at least partly
73: because of the difficulty in measuring distances to HVCs. Without distances, the nature of the HVCs --
74: e.g., whether they are a Galactic or extragalactic phenomenon --
75: can only be ascertained by indirect means.
76:
77: It is clear that the dominant HVC feature -- the 100$^{\circ}$ long Magellanic Stream -- is the result of
78: interaction between the Magellanic Clouds and the Milky Way, either as a result of tidal disruption or ram-pressure
79: stripping or both (Wannier \& Wrixon 1972; Matthewson, Clearly \& Murray
80: 1974; Putman et al. 1998, 2003; Putman \& Gibson 1999).
81: Some HVCs, particularly Complex C, have been attributed to
82: gas falling onto the Milky Way from similar but perhaps smaller-scale interactions (Larson 1972; Tosi 1988;
83: Gibson et al. 2001;
84: Tripp et al. 2003) while other HVCs are ascribed as the product of Galactic fountains
85: (Shapiro \& Field 1976; Bregman 1980, 1996).
86:
87: The compact high velocity clouds (CHVCs) are possibly the most contentious subset of HVCs. It has
88: been proposed
89: that the CHVCs could be the ``missing satellites'' of the Local Group
90: (see discussion in Braun \& Burton 1999 and 2000, hereafter BB99 and BB00; Blitz et al. 1999; de
91: Heij et al. 2002a; Maloney \& Putman 2002; Maller \& Bullock 2004), whose existence
92: is presumed on the basis of their formation in N-body simulations of structure formation in the presence of
93: Cold Dark Matter (CDM). These simulations predict the mass distribution of the Local Group
94: to follow a power law
95: $N(M) \propto M^{-\alpha}$ where $\alpha$ is at least 2 (see, e.g., Klypin et al. 1999; Moore et al. 1999).
96: Such a spectrum would imply the existence of several hundred low mass dwarfs in the Local Group, an order of
97: magnitude more than detected, despite a number of optical surveys for
98: low surface brightness members (Mateo 1998; Armandroff et al. 1998, 1999;
99: Karachentseva \& Karachentsev 1998, 2000; Karachentsev \& Karachentseva 1999; Karachentsev et al. 2000;
100: Whiting et al. 2002; Willman et al. 2004).
101: This contradiction would resolve if the Local Group were filled with dark galaxies -- objects
102: containing dark matter but few, if any, stars, perhaps because of inhibition of star formation
103: in the early Universe (see, e.g., Bullock et al. 2000).
104:
105: The CHVCs' kinematical preference for the Local Group
106: standard of rest (BB99) and the existence of a small overdensity of
107: clouds toward M31 with high negative velocities (Burton et al. 2002; de Heij et al. 2002a)
108: suggest Local Group scale distances. High resolution
109: mapping (BB00; Burton et al. 2002) shows several CHVCs to have core-halo morphology, similar to that expected
110: for dark matter-dominated mini-halos. Three independent methods yield distance estimates for some HVCs of
111: $\sim$ 0.2-0.7 Mpc (BB99; BB00; Burton et al. 2001; Robishaw et al. 2003). The CHVCs also show velocity
112: gradients along their major axes with implied rotation speeds of $V_c \sim 15-20$ km s$^{-1}$ (BB99; BB00;
113: Burton et al. 2001) as well as high internal velocity dispersions. Both phenomena imply high dark matter
114: content (BB00), assuming the HVCs are virialized.
115:
116: Conversely, the sky distribution of CHVCs appears
117: inconsistent with a Local Group origin (Putman et al. 2002; cf. B99 and de Heij et al. 2002a) and CHVCs are not
118: definitively seen in extragalactic groups (de Blok et al. 2002; Zwann 2001;
119: Banks et al. 1999; Pisano et al. 2004; cf. Braun \& Burton 2001). Maloney \& Putman (2003) evaluated the effect of
120: extragalactic ionizing radiation on the CHVCs and concluded that the properties of CHVCs are more consistent with a
121: Galactic halo origin than a cosmological one. And while anomalous hydrogen clouds have been found in nearby
122: galaxies (M31; Thilker et al. 2004: M53 and M81; Miller \& Bregman 2004), these clouds are closer to their
123: parent galaxies than would be likely for dark galaxies.
124:
125: One puzzling aspect of CHVCs are the chemical abundances. They are about a tenth solar, a value too
126: low for Galactic fountain models (Wakker et al. 1999; BB99; BB00) but too high for protogalactic fragments,
127: given the lack of recent star formation in HVCs (Blitz et al. 1999; BB99; BB00). It might be consistent, however,
128: with a tidal or ram-pressure stripping origin since many Local Group dwarfs have abundances near this range.
129:
130: The detection of stars in HVCs would open a new dimension to their study and quickly resolve many uncertainties,
131: including their distance and chemical enrichment history. Additionally, if the CHVCs were discovered to have
132: stellar populations while the HVCs did not, this would clearly indicate that different mechanisms were
133: responsible for the formation of the compact and non-compact HVCs.
134:
135: If the HVCs were within one Mpc of the Milky Way, they would have
136: red giant stars bright enough to be detected with modest aperture
137: telescopes. A number of recent studies have searched for HVC stars. Willman et al. (2002) surveyed a
138: small number of HVCs
139: that were included in the Sloan Digital Sky Survey and found no indication of stellar content.
140: Hopp et al. (2003) surveyed only CHVCs, combining deep VLT imaging
141: with data from the 2MASS survey and comparing
142: each CHVC to a nearby control field. The $VI$ data rule out significant stellar content to a
143: distance of 2 Mpc, although they cover small regions of each cloud. The 2MASS data cover much
144: larger areas of sky, but would only detect giant stars to a distance of 300 kpc.
145: The most complete survey, by Simon \& Blitz (2002), also focused on CHVCs
146: and failed to detect any stellar content. This study placed a lower limit of 100 kpc
147: on the distance to 250 CHVCs and even greater lower limits (300-1400 kpc) on the distance to 60 CHVCs. The
148: Simon \& Blitz survey searched POSS
149: plates for enhancements of stellar density at the positions of the CHVCs and followed up potential detections with
150: monochromatic observations in the Spinrad $R_s$ filter. Though impressive in scale, the Simon \& Blitz survey was
151: biased toward larger HVC stellar populations with densities that could rise above
152: the statistical noise of the foreground of Milky Way stars.
153: Diffuse, low surface brightness systems are a much greater challenge to identify
154: against the Poissonian noise of dense foreground/background Galactic starcounts. To find such structures
155: requires observational strategies that increase their detectability.
156:
157: In this article, we present the first search for {\it low surface-brightness} stellar populations in HVCs, a search
158: based on the observational strategy of minimizing the effects of Galactic contamination.
159: Our photometric survey of four HVCs and adjacent control fields utilizes a filter
160: system especially sensitive to red giant/supergiant stars --- the primary stellar type sought in
161: distant HVCs (and a stellar type found in stellar populations of all ages and metallicities)
162: --- and that allows a large fraction of Galactic stellar contaminants --- which will be
163: dominated by main sequence stars --- to be identified and eliminated. The substantially
164: reduced background/foreground makes it possible to look for very tenuous CHVC stellar populations
165: that would be swamped within simple starcount analyses.
166: For one of our target HVCs and control fields we supplement our photometric analysis with
167: radial velocities derived from echelle
168: spectroscopy of the brightest giant star candidates. \S2 of this paper details
169: the observation and reduction of our photometric observations. \S3 presents an
170: analysis of these data and attempts to quantify the residual sources of possible ``noise''
171: in our program after elimination of the bulk of the likely Galactic dwarf stars.
172: \S4 details the spectroscopic work on the brighter giant candidates in one of the HVCs, work that
173: confirms the basic analysis of \S3. \S5 defines the limits that our survey places on the distances
174: of any stellar populations in the HVCs.
175:
176: \section{Observations and Reduction}
177:
178: The high velocity clouds HVC 017-25-218, HVC 030-51-119 and HVC 271+29+181 were observed
179: in the Washington $C$, $M$, and $T_2$ filters as well
180: as the MgH+Mgb triplet sensitive $DDO51$ filter on UT 5-8 June 2000. We used the Du Pont 2.5 meter telescope and the
181: Wide Field Camera (WFC) designed by Ray Weymann and collaborators. The WFC has a circular field of view of
182: 23' at a pixel scale of $\sim$0\farcs7 per pixel. The optical defect in the WFC noted by Majewski et al.
183: (2000a) was corrected prior to the observations, providing a useable, but
184: still modest quality point spread function (PSF) over the entire image (see below).
185: Each HVC was paired with a control field displaced
186: by $\sim2$-$3$ degrees at nearly constant Galactic latitude
187: and observations were switched back and forth between the HVC and control,
188: resulting in identical depth and image quality for each target and paired
189: control field. Total exposure times for each HVC and its control field were 11500-13400 seconds in
190: $DDO51$, 10000-11000 seconds in $C$, 1830-2130 seconds in $T_2$ and 2730-2790 seconds in $M$, integration times
191: that provide
192: completeness in all four passbands to beyond $M=22$. HVC 030-51-119 and its control
193: field were only minimally observed in $C$.
194:
195: We observed the high velocity cloud HVC 267+26+216 with the Blanco 4-m
196: Telescope and Mosaic II
197: 8k$\times$8k CCD at CTIO during the nights of UT 26-27 February, 2000. The CCD array has a pixel scale of $0\farcs27$
198: per pixel resulting in a field of view of $37\arcmin$. The HVC and control field
199: were observed with total exposure times of 300 seconds in $M$ and $T_2$, and 2100 seconds in $DDO51$. No
200: data were obtained in $C$ for this HVC.
201:
202: Detailed radio properties of the observed HVCs are listed in B99. Figure 1 shows HIPASS (Barnes
203: et al. 2001, Putman et al. 2002) radio images of the four target HVCs
204: with the test and control fields marked. Coordinates and average reddening coefficients (from
205: Schlegel et al. 1998) for the four HVCs and their control fields are listed in Table 1.
206: We note that the control fields for HVC 017-25-218 and HVC 030-51-119 are not ideally placed. They were
207: positioned before HIPASS (Barnes et al. 2001, Putman et al. 2002) maps became available and unfortunately landed
208: on the extended structure of the target HVCs. However, the locations of the control fields have much lower
209: column densities than the central fields and should still provide a useful contrast to the cores. Analysis
210: of the HIPASS data indicates that HVC 017-25-218 and HVC 271+29+181 meet the definition of {\it compact}
211: high velocity clouds.\footnote[1]{At the time of our optical observations, HVC 030-51-119 and HVC 267+26+216
212: were classified as CHVCs. Later radio observations by HIPASS revealed a less compact structure
213: than previous radio images, so that these two objects would no longer be considered CHVCs. In
214: this paper we have reclassified these two objects as HVCs.}
215:
216: Reductions were carried out through the standard IRAF\footnote[2]{IRAF is distributed by the
217: National Optical Astronomy Observatories, which are operated by the Association of Universities for
218: Research in Astronomy, Inc., under cooperative agreement with the National
219: Science Foundation.} CCDRED and MSCRED pipelines as appropriate to each data set.
220: The WFC produces a roughly circular image on the square CCD chip. The unexposed areas of the chip
221: consist entirely of CCD noise and can render normal pipeline analysis difficult.
222: We produced an image mask from the flat fields to remove these sections from analysis.
223: The location of the unexposed region of the chip, however, varied
224: slightly over the course of the run. As a result, the unexposed regions occasionally extended
225: beyond the nominal mask. These ``noisy edges'' were generally small ($\leq 0.5\%
226: $ of the total chip area) but
227: produced hundreds of ``detections'' during photometry. The detections from the noisy edges were
228: clearly non-stellar and were easily removed by our stringent morphological classification
229: (see \S3). A re-analysis of our data with more restrictive masking that excludes more of the field
230: edges showed no change in the starcount density.
231:
232: We derived photometric measures with the DAOPHOT/ALLFRAME PSF photometry package (Stetson 1987, 1994).
233: The PSFs were fit with a quadratically varying Moffat function. The Du Pont data
234: produce modest quality PSFs, with typical rms residuals of 5-7\% per pixel. The Blanco images
235: exhibit excellent PSFs, with typical rms residuals of 1-2\% per pixel.
236:
237: All photometry is calibrated to the standard stars of Geisler (1990, 1996) using the matrix
238: inversion and iterative techniques to derive and apply transformation equations described in
239: Siegel et al. (2002). For the Du Pont data, both
240: program and standard stars are measured in identical $3 \times FWHM$ circular apertures.
241: The Blanco MOSAIC data are calibrated using curve-of-growth analysis from DAOGROW
242: (Stetson 1990). The uncertainty in the photometric transformations is approximately
243: 0.01 magnitudes in all filters.
244:
245: We derived astrometry for our stars from the reference frame of the USNO-A2.0 catalogue
246: (Monet et al. 1996) using the STSDAS TFINDER program in IRAF. Reddening values were generated for each
247: individual star based on interpolation of the reddening maps of Schlegel et al. (1998).
248:
249: Final photometric catalogues are available in electronic form from the Astrophysical Journal. The
250: catalogues list the photometric measures for stars (as selected in \S3) and include $(\alpha, \delta)
251: _{J2000.0}$, photometric measures and $E_{B-V}$ reddening measures. We have also included measures of a
252: modified Welch-Stetson (1993) variability index in which non-variable stars would have values $\leq1.0$.
253: The 4-meter data only have one observation in each filter and the variability indices are listed as 0.0.
254: The stars from the 2.5-meter data generally have indices $\leq 1.0$ indicating little variability over the three
255: days of our observing program and agreement between the formal errors and empirical scatter of the
256: photometric measures. Bright stars are occasionally
257: saturated on the CCD images, resulting in elevated variability indices.
258:
259: \section{Analysis}
260:
261: The $M$, $T_2$, $DDO51$ filter system is now well-established for its ability to distinguish late-type
262: high surface gravity dwarf stars from late-type, low surface gravity, evolved giant stars through the use
263: of two color diagrams (Majewski et al. 2000a,b; Siegel et al.
264: 2000; Palma et al. 2003; Sohn et al. in prep.; Westfall et al. in prep). In particular, we reference Figure 5 of
265: Palma et al. (2003) and Figure 5 of Majewski et al. (2000a), which demonstrate
266: expectations for the appearance of a red giant branch (RGB) in the two-color diagram.
267:
268: Our data reach a magnitude or more deeper than $M\sim 22$. However,
269: data fainter than $M\sim22$ have larger uncertainties and less reliable morphological classification. We can reliably
270: classify objects brighter than $M_0=22.0$ for the Du Pont data and $M_0=21.25$ for the Mayall data
271: \footnote[3]{We classify stars based on the DAOPHOT SHARP
272: parameter. For the 2.5-meter data, objects with $-0.5 < SHARP < 0.11$ are classified as stars; for the 4-meter
273: data, the criterion was $-0.2<SHARP<0.11$. See Siegel et al. (2002) for a discussion of the efficacy of
274: DAOPHOT parameters for object classification.}. Figure 2 shows the run of photometric error with magnitude for our
275: data. We apply
276: magnitude error limits to remove obviously poor measures but retain the bulk of the stars. The adopted magnitude
277: error limits
278: are $\sigma_{T_2}$, $\sigma_M$ and $\sigma_{DDO51} <$ 0.06 for the Du Pont data and $\sigma_{T_2}$, $\sigma_M$ and
279: $\sigma_{DDO51} <$ 0.035 for the Mayall data.
280:
281: Figure 3 shows color magnitude diagrams (CMDs) of our four HVCs and their corresponding
282: control fields. Because they are at similar Galactic latitudes, are dominated by Milky Way stars, and have identical
283: quality data, the HVC and control field CMDs look similar.
284:
285: The CMDs clearly show
286: that the HVCs fields do not contain young stars (at least if they are closer than a few hundred kpc, a limit
287: that depends on the age of the main sequence turnoff [MSTO]) since no obvious main sequences
288: project blueward of the ``blue edge'' of the old, metal-poor halo stars at $(M-T_2)\sim 0.75$.
289: (The slight blueward shift of the blue edge at faint magnitudes is a result of the transition to the
290: more metal-poor populations of the inner halo and thick disk.)
291: This is not altogether surprising, since the previous optical searches of
292: HVCs (B99; B00; Simon \& Blitz 2002) have already placed tighter constraints
293: on the absence of young stars in HVCs than our data by virtue of the greater number of clouds surveyed,
294: deeper photometry and/or larger spatial area.
295:
296: The principal advantage of our program is the degree to which we can detect old or
297: intermediate age stellar populations --- particularly very diffuse ones.
298: Such populations would have RGBs well redward of the
299: blue edge. Ordinarily, with simple starcount techniques, a weak RGB might easily be swamped by
300: the foreground of red Galactic dwarf stars and
301: fall below the detection threshold of the previous surveys. Our program overcomes this problem by
302: severely reducing the Galactic foreground.
303:
304: The basic premise of the three-filter selection method is that late-type giants and dwarfs occupy different
305: loci in color-color space (Fig.\ 4). The dwarf stars lie along the arc of stars at negative $M-DDO51$
306: that traces the strength of magnesium absorption as a function of effective temperature in high
307: surface-gravity, main sequence stars. Low surface-gravity giants lie above this curve in general, offset to a
308: degree dependent upon metallicity (Majewski et al. 2000b).
309:
310: We interactively fit a sixth order polynomial to $(M-DDO51)_0$ as a function of $(M-T_2)_0$ over the
311: limit $0.7 < (M-T_2)_0 < 3.0$. We gradually removed outliers from this function so that it eventually was
312: entirely defined by stars along the dwarf sequence. Once this
313: system was defined, we derived each star's distance from the nominal dwarf function as:
314:
315: \begin{center}
316: $\Delta = (M-DDO51)_0 - f (M-T_2)_0$
317: \end{center}
318:
319: \noindent where $f(M-T_2)_0$ is the ``best-fitting" locus to the dwarf stars in the two-color diagram.
320: We then flagged any star
321: with $1.0 < (M-T_2)_0 < 2.5$ and a positive $\Delta$
322: value greater than three times its photometric error
323: ($\sigma = \sqrt{\sigma_M^2 + \sigma_{T_2}^2 + \sigma_{DDO51}^2}$) as a potential giant.
324: \footnote[4]{Since the shape of the dwarf star curve will
325: change with age/abundance (see Paltoglou \& Bell 1984),
326: using a single fit for all of the dwarf stars is too simplistic.
327: To take out first order effects in the variation of the dwarf locus,
328: we actually measure $\Delta$ with respect to the locus as defined by
329: stars within one magnitude of each star's $M$. }
330:
331: This approach provides some distinct advantages over previous efforts that used rigidly defined selection
332: regions in color-color space
333: to select giant candidates
334: (see Majewski et al. 2000a,b; Morrison et al. 2001; Palma et al. 2003). First, if the errors are Gaussian, each
335: giant candidate selected by the above method
336: is 99.7\% likely to have photometry placing it genuinely outside the nominal dwarf locus. Second,
337: we obtain an immediate analytical estimate for the contamination level as 0.3\% of the total number of stars.
338: Third, even if the errors are not Gaussian or are underestimated, the
339: number of stars with {\it negative} $\Delta$ values greater than $3\sigma$ gives a second, robust estimate of the
340: photometric contamination level. Finally, by removing the constraints of rigidly defined selection limits, a star
341: with exceptional photometry
342: is flagged as a giant even if only a short distance from the nominal dwarf locus while a
343: star with poor photometry must pass correspondingly more stringent requirements to be selected as a potential giant.
344:
345: Figure 5 shows the $\Delta$ versus $M$ distribution of the stars in our eight HVC and control fields. The trend is
346: roughly linear, with growing dispersion as the faint end. Stars flagged as giants and negative outliers are marked.
347: Figure 6 shows the color-color diagram of our giant and negative outlier stars. Note that the reddest giant candidates
348: are nearly within the apparent dwarf locus. These are stars with exceptionally small photometric uncertainties.
349:
350: Table 2 lists the number of giant candidates in each HVC and its corresponding control field, while Figure
351: 7 shows the CMDs of the giant star candidates in the HVCs. The advantage of our method becomes immediately
352: obvious from comparing the raw starcounts to giant star candidate counts in Table 2 or by comparing Figure 7
353: to Figure 3. The screening for giant star candidates
354: reduces the foreground dramatically -- by more than an order of magnitude in all eight
355: cases. A $3\sigma$ surface brightness enhancement would have to be 60-200 stars strong to show up in the
356: raw starcounts or uncleaned CMD. However, Figure 7 and the last columns of Table 2 would easily reveal the signature
357: of even a handful of HVC RGB stars, with a $3\sigma$ surface brightness enhancement corresponding to only 5-25 stars,
358: an order of magnitude improvement in sensitivity.
359:
360: The HVC fields have a significant number of giant candidates that, on first blush, would appear to show
361: a stellar population in the HVCs. However, the control fields provide a critical second constraint.
362: The HVC fields show no excess of giant candidates over the
363: control fields. In fact, three of the HVC fields have {\it fewer} giant candidates than their control fields.
364:
365: The color-magnitude distribution of giant candidates in the HVCs and their corresponding control
366: fields are similar, although they show a number of interesting features.
367: Most of the giant candidates are toward the fainter end of the data, where
368: contamination by extragalactic objects and photometric errors is most likely. However, CHVC 017-25-218 and its
369: control field also show a diffuse clump of bright stars around $M_0\sim15-17$. If bona fide giants, these
370: bright stars would be at distances of 40-100 kpc and therefore well within the bounds of the
371: outer Milky Way halo.
372:
373: There are few giant candidates with colors near the tip of the red giant branch (TRGB) -- typically found
374: at colors
375: of $1.8 \leq M-T_2 \leq 2.5$ for the metallicity range $-2.0 \leq [Fe/H] \leq -0.5$ (see Table 5) and no
376: obvious RGB in any of our fields. The only
377: potential TRGB stars are in the control field of HVC 271+29+181. This field shows a clump of stars near
378: ($M_0, (M-T_2)_0) \sim (21.5, 1.9$) as well as a sprinkling of brighter red stars. The latter we have found
379: to be field star contaminants (see \S4) while the latter could be a very
380: distant ($> 450$ kpc), very dispersed RGB tip.
381:
382: In any case, it is clear that the HVCs are not host to any significant stellar population --- and our discussions
383: and analysis below suggest that they are not likely even to host {\it meager} stellar populations.
384: The most likely explanation for the presence of ``giant star candidates" in our fields is that
385: these candidates are either field giants from the Galactic halo, or represent contamination of the
386: giant sample by non-giants.
387: We now quantify these contributions and in \S4 show that
388: the radial velocity distribution in one of our fields is consistent with that of
389: Galactic field stars.
390:
391: \subsection{Photometric Errors}
392:
393: As discussed above, our new method of giant star selection should provide an immediate evaluation of the photometric
394: contamination of the giant star region in the two-color diagram.
395: Table 3 lists, for each field, the number of contaminants expected given the
396: 3$\sigma$ selection criterion and the number of stars with negative $\Delta$ values greater than 3$\sigma$.
397: For several fields, the latter number is significantly greater than the former. This is a result
398: of the modest quality of the WFC PSF. A disproportionate number of the negative $\Delta$ stars are located near
399: the chip edges where the quadratic PSF is not quite adequate.
400: We use the greater of the two numbers to represent the level
401: of photometric contamination, thus providing a conservative estimate for the number of
402: {\it bona fide} giant stars within our candidate sample.
403:
404: \subsection{Subdwarf Contamination}
405:
406: Another potential source of sample contamination is subdwarf stars.
407: Our candidate selection is based on the presumption that the observed
408: weaker magnesium absorption in these stars is due to surface gravity effects.
409: However, very metal-poor ($[Fe/H] \leq -2.0$) halo subdwarfs may contain so little magnesium that their
410: absorption lines are intrinsically weak and therefore insensitive to surface gravity. Such stars would also
411: fall into the giant
412: region of the two-color diagram. Morrison et al. (2001) estimated this contamination level
413: to be small at bright magnitudes but to grow rapidly at faint magnitudes, depending on the $(l,b)$ of
414: the field. Using
415: their Figure 2, we estimate that our fields should each contain approximately ten subdwarfs, with the
416: exception of HVC 017-25-218, which should contain approximately 20 subdwarfs. This could provide a substantial
417: fraction of the excess ``giant candidates'' in Table 2.
418:
419: \subsection{Compact Galaxies}
420:
421: The light from distant galaxies has a significant, perhaps even dominant
422: contribution, from giant stars (see, e.g. Bershady 1994 and references therein). In addition, at high enough
423: redshifts, the magnesium features can shift out of the $DDO51$ passband, resulting in bluer $M-DDO51$ colors.
424: An examination of the non-compact objects in our photometry shows that they primarily occupy the giant region,
425: a result we have seen in other deep photometry samples that use this filter system.
426: Our structural parameters should limit galaxy contamination. However,
427: {\it compact, star-like} extragalactic objects, which are
428: expected to contaminate photometric samples at a level of $\sim200$ degree$^{-2}$ to $V\sim21.5$
429: (Kron et al. 1991), could remain in the sample. Of course, only a small fraction of compact
430: galaxies have colors red enough to fall into our selection region in the two-color diagram.
431:
432: To estimate the contribution of compact galaxies to the giant star selection region, we took quasar counts
433: from a recent analysis of the Sloan Digital Sky Survey (Richards et al. 2001). We
434: transformed the $M-T_2$ color limits of our giant selection region ($1.0 < (M-T_2)_0 < 2.5$) to approximate $g-r$
435: colors ($0.4 < g-r < 1.5$) using transformations in Majewski et al. (2000b) and Fukugita et al. (1996).
436: Approximately 10-15\% of the Richards et al. sample fall into
437: this color range resulting in a compact galaxy contamination level of $20-30$ degree$^{-2}$ in
438: each of our fields, which would be three galaxies for each WFC field
439: and nine for the MOSAIC fields.
440:
441: This estimate would probably be a lower bound for the 2.5-meter data, which reaches a depth
442: of $M_0=22.0$ and may suffer from more extragalactic contamination owing to the mediocre
443: PSF and commensurately lower quality morphological classification. On the other hand, it is probably an upper bound
444: for the 4-meter data given its brighter magnitude limit and superior imaging capability. We choose
445: to split the difference and apply a uniform compact galaxy contamination level of five objects per field.
446:
447: The $C$ band photometry in three of our HVCs provides an observational test of this approximation.
448: These data were obtained because $C-M$ colors
449: would allow us, had we found a coherent RGB in any of our fields, to place constraints on the
450: metallicity of the giant stars using the
451: iso-metallicity lines derived by Geisler et al. (1991).
452: However, the two-color diagram (Figure 8) shows something
453: unexpected. Many of the giant candidates have far bluer $C-M$ colors than even very metal-poor populations.
454: We find that many of the non-compact galaxies in our photometry also occupy this region of color-color space.
455: Moreover, the star-like objects most displaced from the iso-metallicity lines are
456: preferentially nearer to the faint limit of the data -- where galaxy contamination is
457: expected to be strongest.
458:
459: Our interpretation is that some of our faintest giant star candidates are compact galaxies and fall into
460: the giant region as a result of their composite spectral energy distribution being dominated by giants, or
461: by being at high enough redshift to shift their magnesium features out of the $DDO51$ band.
462: The giant candidates that fall along the iso-metallicity lines in figure 8
463: are far more likely to be bona fide giants or subdwarfs. Significantly, the number of giant candidates
464: substantially removed
465: from the iso-metallicity lines appears to be approximately 1-10 per field, consistent with
466: the number of extragalactic contaminants estimated above.
467:
468: \subsection{Foreground Giants}
469:
470: As mentioned above, the brighter giant candidates are likely from the Galaxy itself. We estimate
471: the number of Galactic field giants
472: for each field by numerically integrating the Fundamental Equation of Stellar Statistics (von Seeliger 1898) using the
473: density laws derived by Siegel et al. (2002, their table 6), and color-absolute magnitude and giant star luminosity
474: functions from
475: Bergbusch \& Vandenberg (2001) converted to Washington colors using the formulae in Majewski et al. (2000b). The
476: number of total estimated contributed Galactic giant stars is listed in the fourth column of Table 3. This includes
477: contributions from the inner halo, thick disk and bulge.
478:
479: Column five of table 3 lists the total combined predicted contamination level by subdwarfs, compact galaxies,
480: photometric errors as well as the contribution of foreground Milky Way giants.
481: Column six lists the difference between the predicted contamination
482: level and the actual number of giant candidates.
483: Although these numbers come with substantial uncertainties, the comparison shows
484: some patterns. HVC 030-51-119 and HVC 267+26+216
485: appear to have fewer giant candidates than predicted; their
486: control fields match the predictions almost exactly. On the other hand, the two compact HVCs
487: have a clear excess of giant candidates beyond the expected contamination level, an excess matched in their
488: control fields. These discrepancies are larger than the statistical noise of the sample. The
489: potential implications are discussed in \S5.
490:
491: \section{Spectroscopy}
492:
493: In order to test the nature of the giant star candidates, we obtained spectra
494: of twelve giant star candidates in CHVC 271+29+181 and its control field on
495: UT January 27-28 2004 using the 6.5-m Clay telescope at Las Campanas with the MIKE spectrograph. CHVC 271+29+181
496: is one of the two fields in which both the HVC and control field have a clear excess of giant candidates beyond
497: the predicted contamination level.
498:
499: The resolution of the echelle spectrograph is approximately $R\sim19,000$ in the red orders.
500: To derive radial velocities, we used three orders extending from the atmospheric bands near
501: 8300 \AA\ to the Calcium IR triplet.
502: Details of the radial velocity measurement technique can be found in Majewski et al. (2004a). In brief,
503: radial velocities were derived using cross-correlation against a ``universal" template masked down
504: to a series of lines that confer the most power in the cross-correlation, and eliminating the
505: vast portions of the spectrum that contain mostly continuum and therefore contribute little more
506: than noise in the cross-correlation. Residual systematic offsets in the velocities were calibrated by
507: observations of K giant velocity standards. The HVC radial velocities thus obtained
508: were subsequently corrected for slit-centering errors by a second cross-correlation focused on the atmospheric
509: absorption bands. The resultant precision on the radial velocities from this multi-step process
510: is better than 1 km s$^{-1}$.
511:
512: Table 4 presents the heliocentric radial velocities for the brightest giant star candidates in CHVC 271+29+181
513: and its control field.
514: Also listed in the table is the strength of each star's correlation
515: peak, a quality indicator from 0 (worst) to 7 (best) of the spectral quality as well as the de-reddened
516: $M$ magnitude and $(M-T_2)$ color.
517:
518: The most important aspect of the radial velocity distribution is the lack of stars at the radial velocity of
519: CHVC 271+29+181 ($v_{r,LSR}$=+181 km s$^{-1}$; $v_{r,helio}$=+191 km s$^{-1}$). The star nearest this
520: radial velocity is 25 km s$^{-1}$ away and is likely to be a star from the Galactic stellar halo.
521: {\it We believe that none of our spectroscopically observed giant candidates is associated with the HVC.}
522:
523: \S3 and Table 3 indicate that approximately half of the giant candidates in CHVC 271+29+181 are Galactic
524: field stars (subdwarfs, stars with photometric errors and {\it bona fide} Galactic giants). The origin
525: of the remaining half is unclear.
526: However, our spectroscopic sample is bright and therefore expected to be heavily
527: dominated by Galactic field stars.
528: The excess of giant candidates is more prevalent at fainter magnitudes.
529:
530: Figure 9 shows the radial velocity distribution of the stars and confirms
531: that these are indeed Galactic field stars. At the $(l,b)$ of CHVC 271+29+181
532: the heliocentric radial velocities of any
533: Galactic field stars will primarily reflect circular velocity differences. One would expect roughly three
534: groupings of stars: a thin disk contribution near 0 km s$^{-1}$, a thick disk with a contribution near
535: 50 km s$^{-1}$ and a halo contribution near 220 km s$^{-1}$ (see, e.g. Casertano et al. 1990).
536: Figure 9 shows that the radial velocity distribution is roughly consistent with this. The spectroscopically
537: observed stars lie
538: principally between -5 and 115 km s$^{-1}$ with a couple of stars at higher, halo-like radial velocities,
539: including star 146, which is in retrograde rotation.
540:
541: It is not surprising that these stars are Galactic field stars.
542: Inspection of Figure 7 shows that the brightest giants in these two fields are
543: either very red or very blue, where the contamination from photometric error and subdwarfs is expected to
544: be the worst. Indeed, five of the six stars in the control field are at the red end and only slightly removed from
545: the dwarf locus in the two-color diagram,
546: which would suggest that using a more rigid cut-off at the red end of the data may be more
547: appropriate in the future.
548:
549: Of course, the most interesting stars in CHVC 271+29+181 are the faint, moderately red stars that
550: represent the true excess above the expected contamination level, particularly the
551: clump at ($M_0, (M-T_2)_0) \sim (21.5, 1.9$) in the control field. Future spectroscopy of these stars
552: will help to determine their nature.
553:
554: \section{Discussion}
555:
556: \subsection{Limits on the Stellar Content and Distance of the HVCs}
557:
558: Previous searches for stellar populations in HVCs have concluded that there are not significant
559: populations of stars in HVCs. Our survey of four HVCs here, meant to identify diffuse populations of giant stars,
560: suggests that there may not even be {\it insignificant}
561: populations of stars in HVCs. In \S3 and \S4 we show that the small number of giant candidates
562: we have identified along the line of sight to each of the HVCs we studied are likely to be field
563: stars from the standard Galactic stellar populations.
564: Our survey cannot, however, rule out the possibility of a stellar counterpart to the HVCs
565: that might be {\it displaced} from our probes. Our survey probes consist of a single HVC and control field
566: pair for each target and, these imaged fields actually cover a small fraction of each HVC (see figure 1).
567: Were an HVC stellar population {\it both} spatially concentrated in its radial distribution {\it and}
568: centered away from the center of the HI gas, we could erroneously infer a lack of stars.
569:
570: Ram-pressure stripping, tidal stripping and/or
571: supernovae can potentially separate gas from stars and the halo abounds with examples. The most
572: famous high velocity
573: cloud -- the Magellanic stream -- is substantially separated from the stellar population of the LMC.
574: Gallart et al. (2001) argue for association of the Phoenix dwarf with an HI cloud that only
575: partly overlaps the stellar populations. The
576: Sculptor dSph has disconnected clumps of gas at similar velocities (Carignan et al. 1998), although this may
577: be foreground contamination by the Magellanic stream (Putman et al. 2003).
578: It is possible that some of the excess stars in our data, particularly in the control fields,
579: could represent a stellar population displaced from the center of the HI. The
580: aforementioned clump of stars at ($M_0, (M-T_2)_0) \sim (21.5, 1.9$) in the control field of
581: CHVC 271+29+181 could be the RGB tip of a displaced stellar population.
582: Only further spectroscopy of the giant candidates can eliminate this possibility.
583:
584: Although ours is the most sensitive probe so far for diffuse stellar populations, it is
585: important to be clear about the upper limits on HVC stellar populations that our null results place.
586: Though the distance, spatial distribution
587: and mass function of any hypothetical HVC stellar populations are unknown, a few basic
588: assumptions allow us to derive useful order-of-magnitude limits.
589: We first define the distance to which we would be able to detect TRGB stars.
590: We estimated TRGB colors and magnitudes from the synthetic photometry of
591: Ostheimer (2003) for a variety of old stellar populations. The
592: TRGB $M-T_2$ color and $M_M$ absolute magnitudes are given in Table 5 for various metallicities.
593: Table 5 also lists minimum distances of a TRGB at which it would be fainter than $M_0=22$ and
594: $M_0=21.25$.\footnote[5]{We do not expect any of our giant candidates to be red supergiants given the lack
595: of a young main sequence in the CMDs.
596: Post-Asymptotic Giant Branch supergiants could be produced by an old population but are short-lived
597: and therefore rare (see Alves et al. 2001).}
598:
599: Beyond 550-760 kpc, RGB stars would be too faint for our program. For an HVC at a hypothetical distance less
600: than these limits, however, we can make a rough analytical approximation of the HVC stellar mass implied by any
601: detected giant stars. At distance $D$, the absolute magnitude of the TRGB ($M_{abs,TRGB}$)
602: corresponds to an observed apparent magnitude
603: $M_{app,TRGB} = M_{abs,TRGB} + 5 log_{10} \frac{D}{10 pc}$.
604: The limiting magnitude of our sample $M_{app,lim}$, in turn, corresponds to an absolute magnitude at distance
605: $D$ of $M_{abs,lim}=M_{app,lim} - 5 log_{10} \frac{D}{10 pc}$. An estimate of the
606: number of HVC giant candidates between $M_{app,TRGB}$ and $M_{app,lim}$ allows us
607: to integrate over the luminosity function and spatial distribution of the
608: hypothetical HVC stellar population to estimate a total stellar mass.
609:
610: The number of stars used for this exercise would optimally be taken from the excess of
611: HVC field stars against control field stars. However, the {\it deficit} of
612: giant candidates in the HVCs compared with their control fields precludes this.
613: As noted above, however, the two CHVC fields and their control fields have approximately
614: 3 and 7 $\sigma$ more giant candidates than the predicted contamination level.
615: We therefore calculate implied stellar masses for these two CHVCs by taking the number
616: of giant stars as the lower of either the excess of giant candidates or the total
617: number of stars fainter than $M_{app,TRGB}$. We also run calculations for all four HVCs
618: assuming that only {\it one} star in the entire sample represented the HVC stellar population. The latter
619: calculation is a useful benchmark of the upper limit to the HVC stellar mass.
620:
621: We integrate the number of giants over the complete global luminosity function of M3
622: (Rood et al. 1999), assuming that any local group object would have an LF similar to that of
623: globular clusters (see, e.g., Feltzing et al. 1999). We convert the LF to Washington colors via
624: the transformations in Majewski et al. 2000b), and assume an exponential spatial distribution with
625: scalelength equal to the HWHM of the HI gas and axial ratio similar to that of the HI gas.
626:
627: Figure 10 shows the upper limit on stellar mass in the HVCs plotted against distance.
628: The solid lines show the fundamental limitation of our survey. A single HVC star in our sample would imply
629: stellar masses of a few $10^5 M_{\sun}$. CHVC 017-25-218 is close to its fundamental limit, owing to the
630: comparatively small size of the HVC and the brightness of its excess giant candidates. CHVC 271+29+181 has a
631: higher potential stellar mass ($10^6$ $M_{\sun}$). However, this higher mass limit is
632: a result of the large number of faint giant candidates in and large spatial extent of the HVC .
633: We emphasize that these
634: are upper limits and that the CHVC stellar content is likely well below this limit, if it exists at all.
635:
636: \subsection{Halo Streams?}
637:
638: Our predicted giant star candidate counts match the observed counts reasonably well (see table 3). However,
639: the two compact HVCs show an interesting statistical signature.
640: CHVC 017-25-218 and its control field both have approximately
641: 3$\sigma$ more giant candidates than predicted while CHVC 271+29+181 and its control field have 3$\sigma$ and
642: 7$\sigma$ more giant candidates, respectively.
643:
644: {\it It is unlikely that these stars are associated with the clouds themselves}, as show in the lack of excess
645: in the CHVC central fields over the controls. However,
646: these stars may still be interesting objects in their own right.
647: Our interest is particularly drawn to the clump of stars at ($M_0, (M-T_2)_0) \sim (21.5, 1.9$) in the
648: control field of CHVC 271+29+191, which could be the tip of a very distant RGB.
649:
650: One important caveat in the Galactic model used in \S3.4, as noted in Siegel et al. (2002),
651: is that the halo often defies conventional density laws. Siegel et al. argued for two halo populations --
652: a smooth inner halo (which we model) and a highly structured outer halo (which we can't and don't),
653: possibly comprised of coherent streams of stellar debris from disrupted dwarf galaxies and/or
654: globular clusters.
655:
656: Several recent studies using our Washington system have found groups of
657: distant giant stars that are not clearly associated with any known dwarf galaxies or globular clusters.
658: Mu\~noz et al. (2004) have recently identified a velocity coherent group of halo giants in the line of sight
659: of the Carina dSph galaxy, but which are not associated with the core of that object.
660: In a study around the Magellanic Clouds,
661: Majewski et al. (in preparation; see also Majewski 2004)
662: show that, at large Galactocentric distances, the Galactic
663: stellar halo may be {\it dominated} by coherent
664: streams of stars, presumably tidally stripped by the Milky Way from dwarf galaxies or globular
665: clusters. The Majewski et al. fields do {\it not} show clear RGB structure in the CMD because the streams
666: are diffuse -- a result eerily similar to what we have found in our two CHVC fields. Additionally, Ostheimer's
667: (2002) deep Washington photometry survey of M31 uncovered numerous examples of potentially {\it intergalactic}
668: giant star candidates, as well as one coherent halo substructure at a distance of about 20 kpc
669: (Majewski et al. 2004b), now confirmed by a strong MSTO feature in his survey data.
670: Interestingly, even this latter, ``Triangulum-Andromeda" structure does not exhibit a strongly identifiable
671: RGB in the CMD, but one that is more subtle (Rocha-Pinto et al. 2004).
672: These precedents make it possible to infer that the excesses in the HVC
673: fields may represent distant diffuse star streams in the outer halo that just happen to cross our survey fields.
674:
675: CHVC 017-25-218 is particularly interesting because its candidate giants are bright.
676: While the HVC and control fields each have
677: approximately twenty giant candidates fainter than $M\sim19$ --- a number consistent with subdwarf, extragalactic
678: and photometric contamination --- each of these fields also has approximately forty giant candidates
679: brighter than $M\sim19$, which is nearly five times the expected contribution from the canonical Galactic models.
680: This field star overdensity could be a signature of halo substructure in the field.
681:
682: CHVC 017-25-218 is close to the extended stellar stream of the tidally disrupted Sagittarius
683: dSph galaxy (Majewski et al. 2003) and the distance modulus of the Sagittarius stream near this location ($m-M\sim
684: 18$) would put Sgr RGB stars close to the apparent magnitude of the bright clump in CHVC 017-25-218 and its
685: control field. It is highly probably that the excess of bright giants in CHVC 017-25-218 and its control field are
686: giants
687: from the Sagittarius debris stream, although spectroscopy will be needed to confirm this interpretation. If these
688: stars are Sagittarius debris, it validates the basic strategy of our survey since in this case we {\it would} have
689: successfully identified a diffuse,
690: but distinct stellar population in our fields. It should be noted that even if the stars are Sagittarius debris, this
691: says nothing about the CHVC itself. The radial velocity of CHVC 017-25-218 is a few hundred km s$^{-1}$ discrepant
692: from the Sgr radial velocity near this field.
693:
694: In \S4, we showed that the brighter giants in CHVC 271+29+181 and its control field are Galactic field stars.
695: However, the bulk of the giant candidates are fainter than the limit of our spectroscopy (and fainter than the excess
696: in CHVC 017-25-218 and its control field). It is possible that spectroscopy of the fainter giants ---
697: particularly the ($M_0, (M-T_2)_0) \sim (21.5, 1.9$) group in the control field --- might
698: reveal another moving group of giant stars even further into the halo.
699:
700: \section{Summary}
701:
702: Our multifilter examination of four HVCs has failed to detect evidence for a
703: stellar population in four HVCs, which, by virtue of our search strategy for diffuse populations
704: of giant stars, places the strongest limits so far
705: on the existence of stellar populations in the cores of HVCs.
706: Although a number of stars in our survey fields can be classified as distant giants based on their location in
707: two-color space, there is no excess of giant candidates
708: in any HVC compared with its control field. In addition, the total number of giant candidates in our fields
709: roughly corresponds to the predicted contribution from Galactic field stars.
710: Echelle spectroscopy of the brightest giants in one
711: HVC and its control field confirm that they are indeed likely to represent stars from canonical Galactic populations.
712:
713: The two compact HVCs show a significant excess of giant candidates beyond the predicted contamination level. However,
714: it is likely that these excesses are unrelated to the CHVCs and are, in fact, distant
715: overdensities (perhaps streams) of halo giants unrelated to
716: the HVCs themselves, particularly in CHVC 017-25-218, which nearly overlaps the Sagittarius debris stream.
717: Only further spectroscopy of the giant candidates should definitively resolve these issues.
718:
719: If none of our stars are associated with their component HVCs, this places an upper limit of a
720: few $10^5 M_{\sun}$ on the stellar content of the HVCs out to a distance of 0.6 Mpc.
721:
722: \acknowledgements
723:
724: The authors would like to thank J.D. Crane and R.J. Patterson for taking the Mosaic data,
725: J. Rhee and J. Ostheimer for sharing their synthetic photometry
726: for TRGB absolute magnitude calibrations and A. Bernacchi for help reducing the MOSAIC data.
727: This manuscript was much improved by the patient and deligent work of the anonymous referee.
728: The authors also thank M. Putman for helpful discussions. Support for this program was provided to SRM by National
729: Science Foundation (NSF) CAREER Award AST-9702521, NSF grant AST-0307851,
730: the David and Lucile Packard Foundation, a Cottrell Scholar Award from the Research Corporation and a Space
731: Interferometry Mission Key Project grant, NASA/JPL contract 1228235.
732: CG was partially supported by the Spanish Ministry of Science
733: and Technology (Plan Nacional de Investigacion Cientfica, Desarrollo e
734: Investigacion Tecnolgica, AYA2002-01939), and by the European Structural Funds.
735:
736: \begin{references}
737: \reference{ABO} Alves, D. R., Bond, H. E. \& Onken, C. 2001, \aj, 121, 318
738: \reference{A1} Armandroff, T. E., Davies, J. E. \& Jacoby, G .H. 1998, \aj, 116, 2287
739: \reference{A2} Armandroff, T. E., Jacoby, G. H. \& Davies, J. E., 1999, \aj, 118, 1220
740: \reference{B99} Banks, et al. 2002, \apj, 524, 612
741: \reference{BSdB} Barnes, et al. 2001, \mnras 322, 486
742: \reference{B94} Bershady, M. A. {\it The Optical and Near-Infrared Colors of Galaxies}, Ph. D. Dissertation, University
743: of Chicago
744: \reference{B99} Blitz, L., Spergel, D. N., Teuben, P. J., Hartmann, D. \& Burton, W. B. 1999, \apj, 514, 818
745: \reference{BB99} Braun, R. \& Burton, W. B. 1999, \aap, 341, 437 [BB99]
746: \reference{BB00} Braun, R. \& Burton, W. B. 2000, \aap, 354, 853 [BB00]
747: \reference{BB01} Braun, R. \& Burton, W. B. 2001, \aap, 375, 219
748: \reference{B80} Bregman, J. N. 1980, \apj, 236, 577
749: \reference{B96} Bregman, J. N. 1996, in {\it The Interplay Between Massive Star Formation, the
750: ISM and Galaxy Evolution}, 11th IAP Meeting. Paris: Inst. d'Astrophys. Paris
751: \reference{BM04} Miller, E. D. \& Bregman, J. N. 2004, {\it astro-ph/0410238}s
752: \reference{B00} Bullock, J. S., Kravtsov, A. V., \& Weinberg, D. H. 2000, \apj, 539, 517
753: \reference{BBC1} Burton, W. B., Braun, R., \& Chengalur, J. N.\ 2001, \aap, 369, 616
754: \reference{BBDH02} Burton, W. B., Braun, R. \& de Heij, V., 2002, in {\it High Velocity Clouds},
755: eds. Wakker, Van Woerden, Schwarz \& De Boer, {\it astro-ph/0206359}
756: \reference{CBCDM} Carignan, C., Beaulieu, S., C{\^ o}t{\' e}, S., Demers, S., \& Mateo, M.\ 1998, \aj,
757: 116, 1690
758: \reference{CRB} Casertano, S., Ratnatunga, K. U. \& Bahcall, J. N. 1990, \apj, 357, 435
759: \reference{dBk} de Block, W. J. G., Zwann, M. A., Dijkstra, M., Briggs, F. H. \& Freeman, K. C., 2002, \aap, 382, 43
760: \reference{dBB2} de Heij, V., Braun, R. \& Burton, W. B., 2002b, \aap, 391, 67
761: \reference{dBB} de Heij, V., Braun, R. \& Burton, W. B., 2002a, \aap,392,417
762: \reference{FGW} Feltzing, S., Gilmore, G., \& Wyse, R.~F.~G.\ 1999, \apj, 516, L17
763: \reference{F96} Fukugita, M., Ichikawa, T., Gunn, J. E., Doi, M., Shimasaku, K., \& Schneider, D. P. 1996, \aj, 111, 1748
764: \reference{GMDGFM} Gallart, C., Martinez-Delgado, D., Gomez-Flechoso, M. A. \& Mateo, M. 2001, \aj, 121, 2572
765: \reference{1990PASP..102..344G} Geisler, D.\ 1990, \pasp, 102, 344
766: \reference{G91} Geisler, D., Claria, J. J. \& Minniti, D. 1991, \aj, 102, 1836
767: \reference{1996AJ....111..480G} Geisler, D.\ 1996, \aj, 111, 480
768: \reference{G01} Gibson, B. K., Giroux, M. L., Penton, S. V., Stocke, J. T., Shull, J. M., \& Tumlinson, J. 2001,
769: \aj, 122, 3280
770: \reference{H} Hopp, U., Schulte-Ladbeck, R. E. \& Kerp, J. 2003, \mnras, 339, 33
771: \reference{KK1} Karachentsev, I. D. \& Karachentseva, V. E. 1999; \aap, 341, 355
772: \reference{KK2} Karachentsev, I. D., Karachentseva, V. E., Suchkov, A .A. \& Grebel, E. K., 2000; \aap, 145,415
773: \reference{KK3} Karachentseva, V. E. \& Karachentsev, I. D., 1998, \aap, 127, 409
774: \reference{KK4} Karachentseva, V. E. \& Karachentsev, I. D., 2000, \aap, 146, 359
775: \reference{K99} Klypin, A., Kravtsov, A. V., Valenzuela, O. \& Prada, F., 1999, \apj, 522, 82
776: \reference{K91} Kron, R. G., Bershady, M. A., Munn, J. A., Smetanka, J. J., Majewski, S. \& Koo,
777: D. C. 1991, in The Space Distribution of Quasars, ed. Crampton, D., ASP Conf. Ser. Vol 21,
778: (San Francisco: ASP), p. 32
779: \reference{L72} Larson, R. B. 1972, Nature, 236, 21
780: \reference{M00a} Majewski, S. R., Ostheimer, J. C., Patterson, R. J., Kunkel, W. E., Johnston, K. V. \&
781: Geisler, D. 2000a, \apj, 119, 760
782: \reference{M00b} Majewski, S. R., Ostheimer, J. C., Kunkel, W. E., \& Patterson, R. J. 2000b, \aj, 120, 2550
783: \reference{M03} Majewski, S. R., Skrutskie, M. F., Weinberg, M. D., \& Ostheimer, J. C. 2003, \apj, 599, 1082
784: \reference{Magst1} Majewski, S. R., Ostheimer, J. C., Crane, J. D., Siegel, M. H., Palma, C. P.,
785: Rhee, J., Kunkel, W. E., Johnston, K. V. \& Patterson, R. J. 2004, {\it in preparation}
786: \reference{M04a} Majewski, S. R. et al., 2004, \aj, 128, 245
787: \reference{M04b} Majewski, S. R. et al., 2004, \apj, {\it in press, astro-ph/0406221}
788: \reference{M04-II} Majewski, S. R. 2004, \pasa, 21, 197
789: \reference{MB04} Maller, A. H. \& Bullock, J. S., 2004, \mnras, {\it in press, astro-ph/0406632}
790: \reference{MP03} Maloney, P. R. \& Putman, M. E. 2003, \apj, 589, 270
791: \reference{M98} Mateo, M. 1998, \araa, 36, 435
792: \reference{Magst} Mathewson, D. S., Cleary, M. N. \& Murray, J. D. 1974, \apj, 190, 291
793: \reference{USNO} Monet, D. G., et al., 1996, USNO-SA2.0, (Washington: US Naval Obs.)
794: \reference{MOMNHDF} Morrison, H. L., Olszewski, E. W., Mateo, M., Norris, J. E., Harding, P.,
795: Dohm-Palmer, R. C., \& Freeman, K. C. 2001, \aj, 121, 283
796: \reference{BM} Moore, B., Ghigna, S., Governato, F., Lake, G., Quinn, T., Stadel, J. \& Tozzi, P., 1999, \apj, 524, 19L
797: \reference{M63} Muller, C. A., Oort, J. H. \& Raimond, E. 1963, C. R. Acad. Sci. Paris, 257, 1661
798: \reference{JCO} Ostheimer, J. C., 2003, {\it Exploring the Halo of M31}, Ph. D. Dissertation, University of Virginia
799: \reference{P} Palma, C., Majewski, S. R., Siegel, M. H., Patterson, R. J., Ostheimer, J. C. \& Link, R. P. 2003, apj, 125, 1352
800: \reference{haloIII} Palma, C., Link, R., Majewski, S. R., Ostheimer, J. C., Frinchaboy, P. M., Patterson,
801: R. J., Kunkel, W. E., Johnston, K. V., \& Geisler, D. 2004, \aj, {\it in prep}
802: \reference{PB84} Paltoglou, G. \ Bell, R. A. 1994, \mnras, 268, 793
803: \reference{P04} Pisano, D. J., Barnes, D. G., Gibson, B. K., Staveley-Smith, L., Freeman, K. C. \& Kilborn, V. A.
804: 2004, \apj, 610, L17
805: \reference{P98} Putman, M. E., et al., Nature, 394, 752
806: \reference{PG} Putman, M. E. \& Gibson, B. K.\ 1999, Publications of the Astronomical Society of Australia, 16, 70
807: \reference{P02} Putman, M. E., et al., 2002, \aj, 123, 873
808: \reference{P03} Putman, M. E., Staveley-Smith, L., Freeman, K. C., Gibson, B. K. \& Barnes, D. G. 2003, \apj, 586, 170
809: \reference{R01} Richards, G. T. et al. 2001, \aj, 121, 2308
810: \reference{R02} Robishaw, T., Simon, J. D. \& Blitz, L. 2003, \apj, 580, 129L
811: \reference{R04} Rocha-Pinto, H. J. et al., 2004, {\it in press, astro-ph/0405437}
812: \reference{M3} Rood, R. T., Carretta, E., Paltrinieri, B., Ferraro, F. R., Fusi-Pecci, F.,
813: Dorman, B., Chieffi, A., Straniero, O. \& Buonanno R., 1999, \apj, 523, 752
814: \reference{S2} Siegel, M., Majewski, S., \& Patterson, R. 2000, BAAS, 197, 134.02
815: \reference{S1} Siegel, M. H., Majewski, S. R., Reid, I. N. \& Thompson, I. B. 2002, \apj, 578, 151
816: \reference{Sch98} Schlegel, D. J., Finkbeiner, D. P. \& Davis, M. 1998, \apj, 1998, 500, 525
817: \reference{SF76} Shapiro, P. R. \& Field G. B. 1976, \apj, 205, 762
818: \reference{SMLP} Sohn, S., Majewski, S. R., Siegel, M. H. \& Patterson, R. J. 2003, {\it in preparation}
819: \reference{SB02} Simon, J.D. \& Blitz, L., 2002, \apj, 574, 726
820: \reference{PBS87} Stetson, P. B. 1987, \pasp, 99, 191
821: \reference{PBS90} Stetson, P. B. 1990, \pasp, 102, 932
822: \reference{PBS94} Stetson, P. B. 1994, \pasp, 106, 250
823: \reference{T04} Thilker, D. A., Braun, R., Walterbos, R. A. M., Corbelli, E., Lockman, F. J., Murphy, E. \&
824: Maddalena, R. 2004, \apj, 601, L39
825: \reference{T03} Tripp, T. M. et al. 2003, \aj, 125, 3122
826: \reference{T88} Tosi, M. 1988, \aap, 197, 47
827: \reference{WW} Wakker, B. P. \& van Woerden, H. 1997, \araa, 35, 217
828: \reference{WW72} Wanner, P. \& Wrixon, G. T. 1972, \apj, 173, L119
829: \reference{WS} Welch, D. L. \& Stetson, P. B. 1993, \aj, 105, 1813
830: \reference{W} Whiting, A. B., Hau, G. K. T. \& Irwin, M., 2002, \apj, 141, 123
831: \reference{W02} Willman, B., Dalcanton, J., Ivezic, Z., Schneider, D. P. \& York, D. G. 2002,
832: \reference{W04} Willman, B., Governato, F., Delcanton, J. J., Reed, D. \& Quinn, T. 2004, \mnras, 353, 639
833: \aj, 124, 2600
834: \reference{Z} Zwann, M. A. 2001, \mnras, 325, 1142
835: \end{references}
836: \clearpage
837:
838: {\bf Figure Captions}
839:
840: \figcaption[f1a.eps]{HI column density distribution in the four HVC fields. The HI
841: emission at about $15\arcmin$ resolution as observed with HIPASS (Barnes
842: et al. 2001, Putman et al. 2002) from a 7$\times$7 degree field integrated
843: over the velocity extent of each object and scaled to column density
844: assuming negligible HI opacity. The greyscale and contour values are in
845: units of 10$^{18}$cm$^{-2}$. The greyscale ranges are indicated by the
846: labeled wedge at the top of each plot. Contours are drawn at 1, 2, 5, 10,
847: 20, 50 and 100 $\times$ 10$^{18}$cm$^{-2}$. The overlaid boxes indicate
848: the positions and approximate sizes of the central and control fields
849: observed for each object.}
850:
851: \figcaption[f2.eps]{Magnitude error as a function of magnitude for all our passbands. DuPont
852: data are on the left, Mayall data on the right. The horizontal lines mark our error limit while the
853: vertical dashed lines indicate the $M$ magnitude limits. The larger uncertainties for the 2.5-meter data reflect
854: the poorer PSF of the WFC more than the smaller aperture.}
855:
856: \figcaption[f3.eps]{Color-magnitude diagrams of (left, top to bottom) HVC 017-25-218,
857: 030-51-119, 267+26+216, and 271+29+181, along with their corresponding
858: control fields (right, top to bottom). Note the lack of any
859: obvious differences between the HVC and control fields. Note also that any
860: sparse RGB would be swamped by the sheer number of stars in each field.}
861:
862: \figcaption[f4.eps]{Color-color diagrams of the four HVC fields and control
863: fields. Panels are as in Figure 3. Note the clear separation of the stars into
864: two populations beyond $M-T\sim1$ -- the extended ``swoosh'' of high-surface gravity dwarfs
865: and the sprinkling of low-surface gravity giant candidates.}
866:
867: \figcaption[f5.eps]{Giant star selection. The plots show $M$ magnitude against $\Delta$ -- a measure
868: of the separation of each star from the dwarf locus. Star are selected as giants (large dots) if they
869: are more than 3$\sigma$ away from the mean $\Delta$ at that magnitude. Starred points are those with
870: $\Delta \leq -3\sigma$ and provide a measure of the photometric contamination level.}
871:
872: \figcaption[f6.eps]{Color-color diagrams of the four HVC fields and control
873: fields. Panels are as in Figure 2. Large dots are objects selected as potential giant stars based
874: on their $\Delta$ values while starred points have $\Delta \leq -3\sigma$ and provide a measure of the photometric contamination level.
875: HVC 267+26+216 shows far less scatter in the two-color diagram because of the shallower limiting
876: magnitude and the superior imaging of the 4-meter telescope.}
877:
878: \figcaption[f7.eps]{Color-magnitude diagrams of giant candidates selected from
879: the four HVC fields and their control fields. Panels are as in Figure 2. Note
880: the lack of any obvious RGB in the distribution of giant candidates. Note also
881: that most of the candidates are at the faint end of the data, with the exception of
882: HVC 017-225-218.}
883:
884: \figcaption[f8.eps]{($C-M$, $M-T_2$) two color diagrams of giant candidates (filled circles) with
885: $\sigma_C \le 0.08$ from the HVCs and control fields. The dots are dwarf stars. The solids lines
886: are iso-metallicity lines from Geisler et al. 1991. These lines
887: only extend to $M-T_2=1.8$. The lines represent (bottom to top) [Fe/H] = -0.5, -1.0, -1.5 and -2.0.
888: Note that this two-color diagram compresses surface gravity information. Both
889: dwarf and giant stars will lie along these lines. The gaps between the lines narrows at red colors because the $C$ filter becomes
890: less sensitive to abundance at low temperatures. Many of our giant candidates are well blueward of
891: these lines, indicating that they are not stellar objects. These two-color diagrams can be
892: contrasted with the coherent RGBs shown in Figure 7 of Geisler et al.. The paucity of stars
893: in HVC 030-51-119 and its control field reflects the minimal
894: data obtained in the $C$ filter for those fields.}
895:
896: \figcaption[f9.eps]{The heliocentric radial velocity distribution of stars in CHVC 271+29+181 and its
897: control field. Overlayed are the approximate radial velocity ranges expected for the canonical thin
898: disk, thick disk and halo velocity distributions from Casertano et al. (1990).
899: Note that the data are roughly consistent with the canonical
900: Galactic stellar populations. Note also the dearth of stars near +191 km s$^{-1}$, the heliocentric radial
901: velocity of the CHVC.}
902:
903: \figcaption[f10.eps]{The upper limit of stellar mass in each of our four HVCs as a
904: function of distance along the line of sight. The rollover at large distances is
905: the magnitude limit of the data.}
906:
907: \clearpage
908:
909: \begin{deluxetable}{cccc}
910: \tablewidth{0 pt}
911: \tablecaption{Target Fields}
912: \tablehead{
913: \colhead{Field} &
914: \colhead{$(\alpha,\delta)_{2000.0}$} &
915: \colhead{$(l,b)$} &
916: \colhead{$E_{B-V}$}}
917: \startdata
918: CHVC 017-25-218 & 19:59:06.0,-24:55:00 & 16.6,-25.3 & 0.107\nl
919: Control 017-25-218 & 20:00:18.5,-23:04:03 & 18.6,-24.9 & 0.148\nl
920: \hline
921: HVC 030-51-119 & 21:58:18.0,-22:42:00 & 29.6,-50.7 & 0.044\nl
922: Control 030-51-119 & 21:59:55.9,-21:24:01 & 31.7,-50.7 & 0.037\nl
923: \hline
924: HVC 267+26+216 & 10:28:06.15,-26:40:44.0 & 267.2,26.2 & 0.077\nl
925: Control 267+26+216 & 10:36:26.18,-27:56:49.5 & 270.5,26.2 & 0.061\nl
926: \hline
927: CHVC 271+29+181 & 10:48:54.0,-26:23:00 & 271.5,28.9 & 0.051\nl
928: Control 271+29+181 & 10:57:22.2,-27:29:12 & 274.0,28.9 & 0.068\nl
929: \enddata
930: \end{deluxetable}
931:
932: \begin{deluxetable}{lcccc}
933: \tablewidth{0 pt}
934: \tablecaption{Number Counts}
935: \tablehead{
936: \colhead{Field} &
937: \colhead{$N_{stars,HVC}$} &
938: \colhead{$N_{stars,Control}$} &
939: \colhead{$N_{giants,HVC}$} &
940: \colhead{$N_{giants,Control}$}}
941: \startdata
942: CHVC 017-25-218 & 3179 & 2973 & 63 & 60 \\
943: HVC 030-51-119 & 555 & 550 & 6 & 22 \\
944: HVC 267+26+216 & 1867 & 1885 & 10 & 28 \\
945: CHVC 271+29+181 & 1032 & 1262 & 35 & 67 \\
946: \enddata
947: \end{deluxetable}
948:
949: \begin{deluxetable}{lccccc}
950: \tablewidth{0 pt}
951: \tablecaption{Giant Region Contamination}
952: \tablehead{
953: \colhead{Field} &
954: \colhead{$N_{expected}$} &
955: \colhead{$N_{\Delta \geq 3\sigma}$} &
956: \colhead{$N_{Galactic}$} &
957: \colhead{$N_{Total}$} &
958: \colhead{$N_{Excess}$}}
959: \startdata
960: CHVC 017-25-218 & 10 & 4 & 8 & 43 & 20\\
961: Control 017-25-218 & 9 & 6 & 8 & 42 & 18\\
962: \hline
963: HVC 030-51-119 & 2 & 0 & 2 & 19 & -13\\
964: Control 030-51-119 & 2 & 3 & 2 & 20 & -2\\
965: \hline
966: HVC 267+26+216 & 6 & 1 & 6 & 27 & -17\\
967: Control 267+26+216 & 6 & 6 & 7 & 28 & 0\\
968: \hline
969: CHVC 271+29+181 & 3 & 2 & 2 & 20 & 15\\
970: Control 271+29+181 & 4 & 11 & 2 & 28 & 39\\
971: \enddata
972: \end{deluxetable}
973:
974: \begin{deluxetable}{lcrcccc}
975: \tablewidth{0 pt}
976: \tablecaption{Giant Candidate Radial Velocities}
977: \tablehead{
978: \colhead{Field} &
979: \colhead{Star ID} &
980: \colhead{$v_{r,helio}$} &
981: \colhead{Correlation} &
982: \colhead{Q} &
983: \colhead{$M_0$} &
984: \colhead{$(M-T_2)_0$} \\
985: \colhead{} &
986: \colhead{} &
987: \colhead{km s$^{-1}$} &
988: \colhead{Peak} &
989: \colhead{(0-7)} &
990: \colhead{} &
991: \colhead{}}
992: \startdata
993: CHVC 271+29+181 & 240 & 147.01 & 0.67 & 4 & 17.49 & 1.09 \\
994: CHVC 271+29+181 & 231 & 85.21 & 1.01 & 7 & 17.27 & 1.24 \\
995: CHVC 271+29+181 & 206 & 40.55 & 0.71 & 6 & 17.04 & 1.15 \\
996: CHVC 271+29+181 & 160 & 115.01 & 1.10 & 7 & 16.41 & 1.26 \\
997: CHVC 271+29+181 & 146 & 349.69 & 0.82 & 7 & 16.53 & 1.04 \\
998: CHVC 271+29+181 & 142 & 66.76 & 0.65 & 7 & 16.36 & 1.18 \\
999: CHVC 271+29+181 & 126 & 106.95 & 0.84 & 7 & 16.15 & 1.32 \\
1000: CHVC 271+29+181 & 87 & -44.42 & 1.10 & 7 & 15.54 & 1.29 \\
1001: Control 271+29+181 & 517 & 24.67 & 0.66 & 6 & 18.10 & 2.49 \\
1002: Control 271+29+181 & 445 & -8.33 & 0.83 & 7 & 17.52 & 2.29 \\
1003: Control 271+29+181 & 134 & 215.08 & 0.79 & 7 & 15.99 & 1.06 \\
1004: Control 271+29+181 & 93 & -4.63 & 1.15 & 7 & 16.26 & 2.38 \\
1005: \enddata
1006: \end{deluxetable}
1007:
1008: \begin{deluxetable}{cccccc}
1009: \tablewidth{0 pt}
1010: \tablecaption{HVC Distance Limits}
1011: \tablehead{
1012: \colhead{[Fe/H]} &
1013: \colhead{Age} &
1014: \colhead{$(M-T_2)_{TRGB}$} &
1015: \colhead{$M_{M,TRGB}$} &
1016: \colhead{$r_{\sun,M=22}$} &
1017: \colhead{$r_{\sun,M=21.25}$}\\
1018: \colhead{} &
1019: \colhead{(Gyr)} &
1020: \colhead{} &
1021: \colhead{} &
1022: \colhead{(kpc)} &
1023: \colhead{(kpc)}}
1024: \startdata
1025: -2.0 & 10-15 & 1.78 & -2.41 & 760 & 540\nl
1026: -1.5 & 10-15 & 1.94 & -2.28 & 720 & 510\nl
1027: -1.0 & 10-15 & 2.22 & -1.99 & 630 & 450\nl
1028: -0.5 & 10-15 & 2.49 & -1.70 & 550 & 390\nl
1029: \enddata
1030: \end{deluxetable}
1031:
1032: \clearpage
1033: %\begin{figure}[h]
1034: %\plotone{f1a.jpg}
1035: %\end{figure}
1036: %\begin{figure}[h]
1037: %\plotone{f1b.jpg}
1038: %\end{figure}
1039: %\begin{figure}[h]
1040: %\plotone{f1c.jpg}
1041: %\end{figure}
1042: %\begin{figure}[h]
1043: %\plotone{f1d.jpg}
1044: %\end{figure}
1045: %\clearpage
1046: %\begin{figure}[h]
1047: %\plotone{f2.jpg}
1048: %\end{figure}
1049: %\begin{figure}[h]
1050: %\plotone{f3.jpg}
1051: %\end{figure}
1052: %\begin{figure}[h]
1053: %\plotone{f4.jpg}
1054: %\end{figure}
1055: %\begin{figure}[h]
1056: %\plotone{f5.jpg}
1057: %\end{figure}
1058: %\clearpage
1059: \begin{figure}[h]
1060: \plotone{f6.eps}
1061: \end{figure}
1062: \begin{figure}[h]
1063: \plotone{f7.eps}
1064: \end{figure}
1065: \begin{figure}[h]
1066: \plotone{f8.eps}
1067: \end{figure}
1068: \begin{figure}[h]
1069: \plotone{f9.eps}
1070: \end{figure}
1071: \begin{figure}[h]
1072: \plotone{f10.eps}
1073: \end{figure}
1074:
1075: \end{document}
1076: