1: \documentstyle[preprint,12pt]{aastex}
2: \def\arcs{$''$}
3: \def\arcm{$'$}
4: \newcommand{\mathR}{\hbox{$\mathcal{R}$}}
5: \newcommand{\royalsociety}{Phil.\ Trans.\ R.\ Soc.\ Lond.\ A}
6: \begin{document}
7: \title{Galaxies at $z\sim7-8$: $z_{850}$-dropouts in the
8: Hubble Ultra Deep Field}
9: \author{R.J. Bouwens$^{2}$, R.I. Thompson$^{3}$,
10: G.D. Illingworth$^{2}$, M. Franx$^{4}$,P. van Dokkum$^{5}$,
11: X. Fan$^{3}$, M.E. Dickinson$^{6}$, D.J. Eisenstein$^{3}$,
12: M.J. Rieke$^{3}$}
13: % }
14: \affil{1 Based on observations made with the NASA/ESA Hubble Space
15: Telescope, which is operated by the Association of Universities for
16: Research in Astronomy, Inc., under NASA contract NAS 5-26555.}
17: \affil{2 Astronomy Department, University of California, Santa Cruz,
18: CA 95064}
19: \affil{3 Steward Observatory, University of Arizona, Tucson, AZ 85721.}
20: \affil{4 Leiden Observatory, Postbus 9513, 2300 RA
21: Leiden, Netherlands.}
22: \affil{5 Department of Astronomy, Yale University, New Haven, CT 06520.}
23: \affil{6 National Optical Astronomy Obs., P.O. Box 26732, Tucson, AZ 85726.}
24:
25: \begin{abstract}
26:
27: We have detected likely $z\sim7-8$ galaxies in the $144''\times144''$
28: NICMOS observations of the Hubble Ultra Deep Field. Objects are
29: required to be $\geq3\sigma$ detections in both NICMOS bands,
30: $J_{110}$ and $H_{160}$. The selection criteria for this sample are
31: $(z_{850}-J_{110})_{AB}>0.8$, $(z_{850}-J_{110})_{AB} > 0.66
32: (J_{110}-H_{160})_{AB}+0.8$, $(J_{110}-H_{160})_{AB}<1.2$, and no
33: detection at $<8500\AA$. The 5 selected sources have total magnitudes
34: $H_{160,AB} \sim 27$. Four of the five sources are quite blue
35: compared to typical lower--redshift dropout galaxies and are clustered
36: within a $1\,\sq '$ region. Because all 5 sources are near the limit
37: of the NICMOS data, we have carefully evaluated their reality. Each
38: of the candidates is visible in different splits of the data and a
39: median stack. We analyzed several noise images and estimate the
40: number of spurious sources to be $1\pm1$. A search using an
41: independent reduction of this same data set clearly revealed 3 of the
42: 5 candidates and weakly detected a 4th candidate, suggesting the
43: contamination could be higher. For comparison with predictions from
44: lower redshift samples we take a conservative approach and adopt four
45: $z\sim7-8$ galaxies as our sample. With the same detection criteria
46: on simulated datasets, assuming no-evolution from $z\sim3.8$, we
47: predict 10 sources at $z\sim7-8$, or 14 if we use a more realistic
48: $(1+z)^{-1}$ size scaling. We estimate that the rest-frame continuum
49: $UV$ ($\sim1800\AA$) luminosity density at $z\sim7.5$ (integrated down
50: to $0.3L_{z=3} ^{*}$) is just $0.20_{-0.08}^{+0.12}\times$ that found
51: at $z\sim3.8$ (or $0.20_{-0.12}^{+0.23}\times$ including cosmic
52: variance). Effectively this sets an upper limit on the luminosity
53: density down to $0.3L_{z=3} ^{*}$. This result is consistent with
54: significant evolution at the bright end of the luminosity function
55: from $z\sim7.5$ to $z\sim3.8$. Even with the lower UV luminosity
56: density at $z\sim7.5$, it appears that galaxies could still play an
57: important role in reionization at these redshifts, though definitive
58: measurements remain to be made.
59:
60: \end{abstract}
61:
62: \keywords{galaxies: evolution --- galaxies: high-redshift}
63:
64: \section{Introduction}
65:
66: From the spectroscopic identification of a population of $z\sim3$
67: dropouts (Steidel et al.\ 1996) to recent work on $i$-dropouts (Yan et
68: al.\ 2003; Stanway et al.\ 2003; Bouwens et al.\ 2003b; Dickinson et
69: al.\ 2004), the frontier for high redshift galaxy studies is
70: continually being redefined. In this paper, we extend this frontier
71: to $z\sim7$ and beyond by performing a $z_{850}$-dropout search over
72: the area of the Hubble Ultra Deep Field (Beckwith et al.\ 2004) with
73: deep NICMOS coverage (Thompson et al.\ 2004a). The exceptional depth
74: of both the optical and infrared data makes this area ideal for
75: carrying out such a search, reaching to 29.5, 29.7, 29.4, 28.8, 27.6,
76: and 27.4 ($5\sigma$, 0.6\arcsec-diameter apertures) in the $F435W$,
77: $F606W$, $F775W$, $F850LP$, $F110W$, and $F160W$ bands (hereinafter,
78: $B_{435}$, $V_{606}$, $i_{775}$, $z_{850}$, $J_{110}$ and $H_{160}$,
79: respectively.) Previously, this redshift range had been probed by an
80: $I_{814}$-dropout search in the HDF-North (Dickinson 2000) and similar
81: dropout searches around lensing clusters (Kneib et al.\ 2004; Pell{\'
82: o} et al.\ 2004). All magnitudes are expressed in the AB system. We
83: assume $(\Omega_M,\Omega_{\Lambda},h) = (0.3,0.7,0.7)$ (Bennett et
84: al.\ 2003).
85:
86: \section{Analysis}
87:
88: Our search area was the 0.09$''$ pixel $144''\times 144''$ NICMOS
89: mosaic (Thompson et al.\ 2004a). Sources were identified in the
90: summed $J+H$ image (RT) and the $\chi^2$ (Szalay et al.\ 1999) image
91: (RB) using the SExtractor code (Bertin \& Arnouts 1996). Colors were
92: calculated using a scaled aperture Kron magnitude (1980) with the Kron
93: factor equal to 1.2. Total magnitudes were then derived using the
94: $\chi^2$ image to correct these fluxes to a much larger aperture
95: (where the Kron factor was equal to 2.5) (see Bouwens et al.\ 2003a).
96: Typical corrections were $\sim0.8$ mag for each object.
97:
98: \textit{(a) $z_{850}$-dropout selection.} Objects were required to be
99: null detections ($<2\sigma$) in the deepest ($V_{606}$ and $i_{775}$)
100: optical bands (in 0.6\arcsec$\,$-diameter apertures), and lie in the
101: expected place [$(z_{850}-J_{110})>0.8, (z_{850}-J_{110})_{AB} > 0.66
102: (J_{110}-H_{160})_{AB} + 0.8$, $(J_{110}-H_{160})_{AB}<1.2$] in the
103: standard two-colour $z_{850}-J_{110}/J_{110}-H_{160}$ diagram. To
104: clean our catalog of possible spurious detections, objects were
105: required to be $3\sigma$ detections (0.6\arcsec$\,$-diameter aperture)
106: in both the $J_{110}$ and $H_{160}$ bands. These procedures
107: identified a set of 8 sources to a limiting magnitude of
108: $H_{160,AB}\sim28$. A separate selection by RT identified a similar
109: set of objects. After identification, each source was located in the
110: original exposures (16 in each band) to ensure that they did not arise
111: from a small subset of the exposures (e.g., from a pre-integration
112: cosmic ray hit). Three of our 8 sources were rejected, being visible
113: in only a couple of exposures. The 5 real sources are shown in Figure
114: 1 and Table 1.
115:
116: Table 1 includes the photometric information for our 5 candidates plus
117: one red galaxy which nearly met our criteria (this latter object was
118: also found by Yan \& Windhorst 2004). Candidates had $H_{160,AB}$
119: magnitudes ranging from 26.0 to 27.3, or 0.5-1.5 times the
120: characteristic rest-frame UV luminosity $(L^{*})$ for Lyman break
121: galaxies at $z\sim3$ (Steidel et al.\ 1999). Candidates appear to be
122: rather clustered, with 4 of the 5 candidates falling within a
123: $\sim1\,\sq '$ area. Figure 2 displays postage stamp images of each
124: candidate along with its position in color-color space and an SED fit
125: to the broadband fluxes.
126:
127: \begin{figure}[h]
128: \epsscale{0.95}
129: \plotone{f1.eps}
130: \caption{$(z_{850}-J_{110})_{AB}/(J_{110}-H_{160})_{AB}$ color-color
131: diagram showing the position of our $z_{850}$-dropouts (selection
132: region is shaded gray) relative to the UDF photometric sample (cyan
133: squares). Objects included in the source list (Table 1) are shown as
134: black squares ($2\sigma$ lower limits are indicated by vertical
135: arrows). These objects are not detected in the optical $V_{606}$ and
136: $i_{606}$ bands. The cyan squares that lie in the selection region
137: have clear $V_{606}$ and $i_{775}$ detections ($>2\sigma$) and so are
138: not candidate $z_{850}$-dropouts; representative error bars for these
139: objects are shown at the right of this diagram. The color-color
140: tracks of both lower redshift interlopers (\textit{red lines}) and
141: $10^8$ yr starburst SEDs with different reddenings (\textit{blue
142: lines}) are plotted as a function of redshift. The position of M, L,
143: and T dwarfs are also shown (\textit{green cross hatched region})
144: (Knapp et al.\ 2004). Error bars on the $z_{850}-J_{110}$ and
145: $J_{110}-H_{160}$ colors are $1\sigma$.}
146: \end{figure}
147:
148: \textit{(b) Testing Source Reality.} Our 5 candidates were then
149: subjected to several additional tests. Each source was verified to
150: exist at the $>2.5\sigma$ level in the $J+H$ image for each of the two
151: epochs (taken two months apart and at a 90 degree angle to each
152: other). Each source was also evident ($>2.4\sigma$) in a median
153: stacking of the 16 overlapping exposures for each band. This is
154: useful since the median process should eliminate sources with flux in
155: only a few exposures. After performing the above sanity checks on our
156: candidates, we repeated our selection procedure on three different
157: images sets to examine the likelihood that our candidates are simply
158: spurious detections. These three images include the ``negative''
159: images, the first epoch images subtracted from the second epoch
160: images, and the second epoch images subtracted from the first. These
161: images should have similar noise characteristics to the data, but
162: contain no real sources. Only 1, 2, and 0 objects, respectively, were
163: found on each of the above three image sets (5.76 arcmin$^2$) using an
164: identical selection procedure. This suggested a small level of
165: contamination from spurious sources ($1\pm1$ object) in the current
166: sample.
167:
168: \textit{(c) An Independent Check on Source Reality.} An independent
169: reduction of the NICMOS images was kindly made available to one of us
170: (RT) by Robberto et al.\ (2004). The image was inspected by RT and 3
171: of our 5 candidate sources clearly appear in those images. However,
172: no signal is evident at the position of UDF-818-886, while UDF-491-880
173: is only weakly detected. Until this is resolved, the contamination may
174: be higher than estimated above (\S2b).
175:
176: \begin{figure}
177: \epsscale{1.0}
178: \plotone{f2.eps}
179: \caption{Postage stamps images ($V_{606}i_{775}z_{850}J_{110}H_{160}$
180: bands) of our 5 $z_{850}$-dropout candidates. Also shown
181: (UDF-640-1417: bottom row) is one very red
182: $(z_{850}-J_{110})_{AB}=1.1$ object which nearly met our selection
183: criteria and could be a reddened starburst at $z\sim6.5$ (or a
184: reddened early type at $z\sim1.6$) (also found by Yan \& Windhorst
185: 2004). While our best-fit to UDF-387-1125 is a $z\sim6.8$ starburst
186: spectrum, this object is also consistent with being a compact
187: $z\sim1$, $0.01L^{*}$ dust-reddened early type galaxy. Magnitudes are
188: those measured in a $0.6''$-diameter aperture. The three rightmost
189: panels show the combined $J_{110}+H_{160}$ image for each object, its
190: position in color-color space, and an SED fit to the broadband fluxes.
191: The derived redshift is also provided in the rightmost panel. The ACS
192: cutouts here are shown at a much higher contrast than the NICMOS
193: cutouts, demonstrating the significance of the optical non-detections.
194: The postage stamps are 2.9\arcsec$\times$2.9\arcsec$\,$ in size. A
195: linear stretch is used for scaling the pixel fluxes.}
196: \end{figure}
197:
198: \textit{(d) Low-Redshift Contamination.} To test for possible
199: contamination from low-redshift interlopers, we randomly assigned the
200: colors of bright ($23.5<H_{160,AB}<25$) objects from the UDF to faint
201: objects in our field, added photometric scatter, and then repeated our
202: selection. No objects were found, suggesting minimal contamination
203: from low-redshift interlopers. Possible contamination from T dwarfs
204: was also considered, given their position in color-color space (Figure
205: 1) and predicted numbers (0.04-0.3 objects) over our field of view
206: (Burgasser et al.\ 2004). However, this proved not to be a concern
207: for our sample, since T dwarfs would appear as $\gtrsim5\sigma$ point
208: sources in the deep $z_{850}$-band images, and none were found.
209:
210: \textit{(e) Expected Numbers/Incompleteness Tests.} It is interesting
211: to compare the number of candidates against that predicted assuming
212: no-evolution from lower redshift. As in other recent work, we adopt a
213: $z\sim3.8$ $B$-dropout sample from the GOODS fields (Bouwens et al.\
214: 2004, hereinafter B04) as our reference point and project it to
215: $z\sim6-10$ using our well-established cloning machinery (Bouwens et
216: al.\ 1998a,b; Bouwens et al.\ 2003a; B04). Such simulations are
217: important for establishing the incompleteness, which can be as high as
218: 75\% for these $z\sim7-8$ objects (this includes the effect of
219: possible blending with foreground galaxies). Adding the cloned
220: galaxies directly to the data, we repeat our selection procedure and
221: thereby derive a no-evolution prediction; this yields 10 dropouts.
222: However, we know that galaxies evolve in size (a $(1+z)^{-1}$ size
223: scaling for fixed luminosity: Bouwens et al.\ 2004a,c; Ferguson et
224: al.\ 2004) and hence surface brightness. Including this effect, 14
225: objects are found. Steeper size scalings (e.g., $(1+z)^{-2}$) yield
226: still larger values ($\sim18$ objects) while using bluer colors (e.g.,
227: UV slopes $\beta\sim-2.5$) has little effect on the predictions.
228:
229: \textit{(f) Source Characteristics / Possible Concerns.} Given the
230: depth of the UDF $z_{850}$-band imaging, it was somewhat surprising
231: that only 1 of our 5 candidates is detected in this band. We used the
232: simulations described above (\S2e) to quantify this and found that
233: $58\%$, or 2.9 of our 5 candidates, should be detected at $>2\sigma$
234: in the $z_{850}$-band. A single detection in $z_{850}$ has only a
235: 10-22\% likelihood of occurrence, the larger number for significant
236: clustering. Four of the objects are spatially clustered, falling
237: within a $1\,\sq '$ area. Such clustering is not unexpected, and the
238: lack of $z_{850}$-band flux would result if they are also at $z>7.5$.
239:
240: These 4 clustered objects are also quite blue, relative to the
241: fiducial $10^8$ yr starburst, though the significance of this result
242: is modest ($<2\sigma$). Their rest-frame UV colors (with $\beta\sim
243: -3$) are bluer than typical dropouts at both $z\sim3$ (Steidel et al.\
244: 1999) and $z\sim6$ (Stanway et al.\ 2004b) where $\beta\gtrsim-2.2$
245: (the $E(B-V)\gtrsim0$ track in Figure 1). If real, they are also
246: bluer than one would expect from models which give $\beta\gtrsim -2.5$
247: regardless of age, dust, or metallicity content, including population
248: III objects (Schaerer 2003; Venkatesan et al.\ 2003). However, such
249: blue colors are not completely unknown at $z\sim3$ (e.g., Adelberger
250: \& Steidel 2000). They also might arise from a significant
251: contribution of Ly-$\alpha$ emission (rest-frame EWs $\gtrsim200\AA$)
252: to the $J_{110}$-band flux.
253:
254: The one red object in our $z\sim 7-8$ sample has colors that are also
255: consistent with a $z\sim 1$ galaxy, though it would be rather unusual,
256: needing to be a compact ($\sim 1$kpc), significantly reddened, old
257: stellar population dwarf galaxy, $>6$ magnitudes fainter that L$^*$;
258: we consider it more likely to be at high redshift.
259:
260: \section{Luminosity Density and Implications}
261:
262: We are now prepared to compare the observations with the predictions
263: made earlier (\S2e). This will permit us to set important constraints
264: on the evolution at the bright end of the luminosity function in
265: rest-frame continuum-$UV$ ($\sim1800\AA$) and therefore make
266: inferences about changes in the luminosity density. To be
267: conservative, we shall assume the number of $z_{850}$-dropouts is
268: four. Given possible concerns about their validity (\S2b;\S2c;\S2f), we
269: will also consider the implications if there are even fewer sources.
270: For the expected number of $z_{850}$-dropouts, we use 14, the
271: prediction from the $(1+z)^{-1}$ size scaling (\S2e).
272:
273: Comparing our 4 fiducial candidates with the 14 objects predicted
274: suggests that the number of objects at the bright end of the LF at
275: $z\sim7.5$ is just $0.29\times$ that at $z\sim3.8$ (Figure 3). This
276: decreases to $0.14\times$ and $<0.13\times$ ($1\sigma$) if only two or
277: none of our candidates are real, respectively. Obviously, there are
278: substantial uncertainties in the estimated shortfall, both as a result
279: of the small number statistics and the expected cosmic variance
280: (factor of 2: assuming a CDM power spectrum normalized to high
281: redshift observations and a redshift selection window of unit width,
282: e.g., Somerville et al.\ 2004). Therefore, even no evolution is
283: consistent with the present result at the $1.5\sigma$ level.
284:
285: While a number of options are open, the most likely case is a drop of
286: at least $3.5\times$ in the number of objects at the bright end of the
287: LF. Since this is similar to what is found at $z\sim6$ (Stanway et
288: al.\ 2003; Dickinson et al.\ 2004; Stanway et al.\ 2004a), it is
289: likely a continuation of the same effect. A key question is whether
290: the observed deficit continues all the way down the luminosity
291: function or if it is due to evolution in the characteristic luminosity
292: ($L^{*}$) at high redshift. This whole issue is pivotal for questions
293: about reionization since it is at faint magnitudes that the bulk of
294: the flux arises (assuming a steep $\lesssim-1.5$ faint-end slope
295: $\alpha$). Fortunately, the fainter $i$-dropouts from the UDF are
296: beginning to provide us with some clues, and some early studies are
297: already suggesting that the principal form of the evolution is in
298: luminosity or a steepening of the faint-end slope (Dickinson et al.\
299: 2004; Yan \& Windhorst 2004; Bouwens et al.\ 2004d; cf. Bunker et al.\
300: 2004). If true, this would provide a natural explanation for our
301: shortfall and may allow for substantial star formation at higher
302: redshifts as suggested by recent measurements from WMAP (Kogut et al.\
303: 2003) or the large stellar masses found in the $z\sim6.5$ Kneib et
304: al.\ (2004) object (Egami et al.\ 2004). It would also suggest that
305: for a proper census of these objects the present surveys need to be
306: extended to considerably fainter magnitudes (with WFC3 and ultimately
307: with JWST).
308:
309: In light of the uncertainties regarding the form of the evolution, we
310: have chosen simply to quote the evolution in luminosity density down
311: to the total magnitude limit of our survey ($H_{160,AB}\sim27.5$, or
312: $\sim0.3L_{z=3} ^{*}$). Conversions to star formation rate density
313: (uncorrected for extinction) are made using the now canonical
314: conversion factors for the Salpeter IMF (Madau et al.\ 1998). For
315: both quantities (the luminosity density and the star formation rate),
316: we infer a larger drop than above (due to luminosity-weighting). To
317: the faint end limit and including the Poissonian variations quoted
318: above, we find that $\rho(UV,z=7.5)/\rho(UV,z=3.8) =
319: 0.20_{-0.08}^{+0.12}$ using our fiducial list of candidates and
320: $0.10_{-0.05}^{+0.09}$ and $<0.05$ ($1\sigma$) assuming only two or
321: none of our candidates are real, respectively. Uncertainties on these
322: quoted factors increase to $0.20_{-0.12}^{+0.23}$,
323: $0.10_{-0.07}^{+0.20}$, and $<0.11$ ($1\sigma$), respectively,
324: including the expected field-to-field variations (cosmic variance)
325: quoted above. Figure 3 shows a comparison of these results with those
326: at lower redshift.
327:
328: This is the first such deep sample ever compiled at $z\sim7-8$ and
329: allowed us to set some constraints on the bright end of the rest-frame
330: $UV$-continuum luminosity function at $z\sim7.5$, during the epoch of
331: reionization. The similarity of the present result with that at
332: $z\sim6$ (Stiavelli et al.\ 2004; Yan \& Windhorst 2004) suggests that
333: galaxies could have been an important contributor to reionization at
334: these early times, though a characterization of their role warrants
335: more definitive measurements.
336:
337:
338: \begin{figure}
339: \plotone{f3.eps}
340: \caption{Top Panel: Rest-frame continuum $UV$ ($\sim1800\AA$)
341: luminosity density (integrated down to 0.3 $L_{z=3} ^{*}$)
342: vs. redshift. The observed luminosity density is converted to a star
343: formation rate (uncorrected for extinction) assuming a Salpeter IMF
344: (e.g., Madau et al.\ 1998). The present determination (assuming 4
345: candidates) is shown as the large red circle, with an upper limit
346: shown to acknowledge possible concerns regarding several of our
347: candidates. Previous determinations from Lilly et al.\ (1996)
348: (\textit{open squares}), Steidel et al.\ (1999) (\textit{green
349: crosses}), Giavalisco et al.\ (2004a) (\textit{solid black diamonds}),
350: Bunker et al.\ (2004) (\textit{solid blue square}), B04 (\textit{solid
351: magenta circles}), and Bouwens et al.\ 2004a (\textit{solid red
352: triangle}) are also shown. The uncertainty expected from large scale
353: structure (cosmic variance) is $\pm$20\% for many of the lower
354: redshift points (e.g., Somerville et al.\ 2004) and $\pm$50\% for the
355: $z\sim7.5$ point. The top horizontal axis provides the corresponding
356: age of the universe. Lower Left Panel: The surface density vs. total
357: magnitude of the observed $z_{850}$-dropouts (\textit{histogram}) and
358: that predicted from a $(1+z)^{-1}$ size scaling of our GOODS
359: $B$-dropout sample (B04) (\textit{red}, see \S2e). Lower Right Panel:
360: The expected redshift distribution for $z_{850}$-dropouts derived from
361: these same simulations. These results suggest a modest to significant
362: decline in the star formation rate density (uncorrected for
363: extinction: see Thompson et al.\ 2004b for the extinction-corrected
364: star formation history).}
365: \end{figure}
366:
367: \acknowledgements
368:
369: We are appreciative to Andy Bunker, Dave Golimowski, Sandy Leggett,
370: Piero Madau, and Daniel Schaerer for useful conversations, Adam
371: Burgasser for important estimates of T dwarf surface densities, Sune
372: Toft for help with the PSFs, and our referee Haojing Yan for comments
373: which significantly improved this manuscript. This research was
374: supported under NASA grant HST-GO09803.05-A and NAG5-7697.
375:
376: \begin{thebibliography}{}
377: \bibitem[Adelberger \& Steidel(2000)]{2000ApJ...544..218A} Adelberger,
378: K.~L.~\& Steidel, C.~C.\ 2000, \apj, 544, 218.
379: \bibitem[Beckwidth 2004]{beckwith2004} Beckwith, S. V. W. et al.\
380: 2004, in preparation.
381: \bibitem[Bennett et al.(2003)]{2003ApJ...583....1B} Bennett, C.~L.~et
382: al.\ 2003, \apj, 583, 1.
383: \bibitem[Bertin and Arnouts (1996)]{1996A&AS..117..393B} Bertin, E.\ and
384: Arnouts, S.\ 1996, \aaps, 117, 393.
385: \bibitem[Bouwens, Broadhurst and Silk (1998)]{1998ApJ...506..557B} Bouwens,
386: R., Broadhurst, T.\ and Silk, J.\ 1998a, \apj, 506, 557.
387: \bibitem[Bouwens, Broadhurst and Silk (1998)]{1998ApJ...506..579B}
388: Bouwens, R., Broadhurst, T.\ and Silk, J.\ 1998b, \apj, 506, 579.
389: \bibitem[Bouwens, Broadhurst, \&
390: Illingworth(2003)]{2003ApJ...593..640B} Bouwens, R., Broadhurst, T.,
391: \& Illingworth, G.\ 2003a, \apj, 593, 640.
392: \bibitem[Bouwens et al.(2003)]{2003ApJ...595..589B} Bouwens, R.~J.~et
393: al.\ 2003b, \apj, 595, 589.
394: \bibitem[Bouwens et al.(2004)]{2004ApJ...606L..25B} Bouwens, R.~J.~et al.\
395: 2004a, \apjl, 606, L25
396: \bibitem[Bouwens et al.\ 2004]{bbi2004} Bouwens, R.J., Broadhurst,
397: T.J., Illingworth, G.D., Meurer, G.R., Blakeslee, J.P., Franx, M., \&
398: Ford, H.C. 2004b, \apj, submitted (B04).
399: \bibitem[Bouwens et al.\ 2004]{b2004c} Bouwens, R.J., Illingworth,
400: G.D., Blakeslee, J.P., Broadhurst, T.J., \& Franx, M. 2004c, \apjl,
401: 611, L1.
402: \bibitem[Bouwens et al.\ 2004]{b2004c} Bouwens, R.J. et al.\ 2004d, in
403: preparation.
404: \bibitem[Bunker et al.\ 2004]{bunk2004} Bunker, A.J., Stanway, E.R.,
405: Ellis, R.S., McMahon, R.G. 2004, \mnras, submitted.
406: \bibitem[Burgasser et al.\ 2004]{bunk2004} Burgasser, A., et al.\
407: 2004, in preparation.
408: \bibitem[Dickinson(2000)]{2000RSPTA.358.2001D} Dickinson, M.\ 2000, Royal
409: Society of London Philosophical Transactions Series A, 358, 2001.
410: \bibitem[Dickinson et al.(2004)]{2004ApJ...600L..99D} Dickinson, M.~et al.\
411: 2004, \apjl, 600, L99.
412: \bibitem[Egami et al.\ 2004]{bunk2004} Egami, E.~et al.\ 2004, in
413: preparation.
414: \bibitem[Ferguson et al.(2004)]{2004ApJ...600L.107F} Ferguson, H.~C.~et
415: al.\ 2004, \apjl, 600, L107.
416: \bibitem[Fern{\' a}ndez-Soto, Lanzetta, \& Chen(2003)]{2003MNRAS.342.1215F}
417: Fern{\' a}ndez-Soto, A., Lanzetta, K.~M., \& Chen, H.-W.\ 2003, \mnras,
418: 342, 1215.
419: \bibitem[Giallongo, Cristiani, D'Odorico, \&
420: Fontana(2002)]{2002ApJ...568L...9G} Giallongo, E., Cristiani, S.,
421: D'Odorico, S., \& Fontana, A.\ 2002, \apjl, 568, L9.
422: \bibitem[Giavalisco et al.(2004)]{2004ApJ...600L.103G} Giavalisco, M.~et
423: al.\ 2004, \apjl, 600, L103.
424: \bibitem[Knapp et al.\ 2004]{kna2004} Knapp, G.R., et al.\ 2004, \aj,
425: in press.
426: \bibitem[Kneib, Ellis, Santos, \& Richard(2004)]{2004ApJ...607..697K}
427: Kneib, J., Ellis, R.~S., Santos, M.~R., \& Richard, J.\ 2004, \apj, 607,
428: 697.
429: \bibitem[Kogut et al.(2003)]{2003ApJS..148..161K} Kogut, A.~et al.\ 2003,
430: \apjs, 148, 161.
431: \bibitem[Kron(1980)]{1980ApJS...43..305K} Kron, R.\ G.\ 1980, \apjs, 43,
432: 305.
433: \bibitem[Lilly et al.\ 1996]{lil96} Lilly, S.J., Le Fevre, O., Hammer,
434: F., \& Crampton, D. 1996, \apj, 460, L1.
435: \bibitem[Madau et al.\ 1998]{mad98} Madau, P., Pozzetti, L. \&
436: Dickinson, M. 1998, \apj, 498, 106.
437: \bibitem[Pello et al.\ 2004]{pel04} Pell{\' o}, R., Schaerer, D.,
438: Richard, J., Le Borgne, J.-F., \& Kneib, J.-P. 2004, \aap, 416, L35.
439: \bibitem[Robberto et al.\ 2004]{robb2004} Robberto, M.~et al.\ 2004,
440: in preparation.
441: \bibitem[Schaerer(2003)]{2003A&A...397..527S} Schaerer, D.\ 2003, \aap,
442: 397, 527.
443: \bibitem[Somerville et al.(2004)]{2004ApJ...600L.171S} Somerville, R.~S.,
444: Lee, K., Ferguson, H.~C., Gardner, J.~P., Moustakas, L.~A., \& Giavalisco,
445: M.\ 2004, \apjl, 600, L171.
446: \bibitem[Stanway, Bunker, \& McMahon(2003)]{2003MNRAS.342..439S}
447: Stanway, E.~R., Bunker, A.~J., \& McMahon, R.~G.\ 2003, \mnras, 342,
448: 439.
449: \bibitem[Stanway et al.(2004)]{2004ApJ...607..704S} Stanway, E.~R., Bunker,
450: A.~J., McMahon, R.~G., Ellis, R.~S., Treu, T., \& McCarthy, P.~J.\ 2004a,
451: \apj, 607, 704.
452: \bibitem[Stanway et al.(2004)]{stansub} Stanway, E.~R.,
453: McMahon, R.~G., \& Bunker, A.~J. 2004b, \mnras, submitted.
454: \bibitem[Steidel et al.\ (1999)]{1999ApJ...519....1S} Steidel, C.\ C.,
455: Adelberger, K.\ L., Giavalisco, M., Dickinson, M.\ and Pettini, M.\ 1999,
456: \apj, 519, 1.
457: \bibitem[Steidel, Pettini, \& Adelberger(2001)]{2001ApJ...546..665S}
458: Steidel, C.~C., Pettini, M., \& Adelberger, K.~L.\ 2001, \apj, 546, 665.
459: \bibitem[Stiavelli, Fall, \& Panagia(2004)]{2004ApJ...610L...1S} Stiavelli,
460: M., Fall, S.~M., \& Panagia, N.\ 2004, \apjl, 610, L1.
461: \bibitem[Szalay, Connolly, \& Szokoly 1999]{sza99} Szalay, A.S.,
462: Connolly, A.J., \& Szokoly, G.P. 1999, \aj, 117, 68.
463: \bibitem[Thompson et al.\ 2004a]{thom04b} Thompson, R.I.~et al.\ 2004a,
464: in preparation.
465: \bibitem[Thompson et al.\ 2004b]{thom04b} Thompson, R.I.~et al.\ 2004b,
466: in preparation.
467: \bibitem[Venkatesan, Tumlinson, \& Shull(2003)]{2003ApJ...584..621V}
468: Venkatesan, A., Tumlinson, J., \& Shull, J.~M.\ 2003, \apj, 584, 621.
469: \bibitem[Yan, Windhorst, \& Cohen(2003)]{2003ApJ...585L..93Y} Yan, H.,
470: Windhorst, R.~A., \& Cohen, S.~H.\ 2003, \apjl, 585, L93.
471: \bibitem[Yan \& Windhorst(2004)]{2004ApJ...612L..93Y} Yan, H.~\& Windhorst,
472: R.~A.\ 2004, \apjl, 612, L93.
473: \end{thebibliography}
474:
475: \newpage
476: \begin{deluxetable}{lccccccccc}
477: \tablewidth{480pt} \tabletypesize{\scriptsize}
478: \tablecaption{$z\sim7-8.5$ Sample.\tablenotemark{a}} \tablehead{
479: \colhead{Object ID} & \colhead{Right Ascension} &
480: \colhead{Declination} & \colhead{$H_{160,Cor}$} &
481: \colhead{$H_{160,Ap1}$} & \colhead{$H_{160,Ap2}$} & \colhead{$z - J$}
482: & \colhead{$J - H$} & \colhead{S/G} & \colhead{$r_{hl}$(\arcs)}}
483: \startdata
484: UDF-825-950 & 03:32:39.538 & -27:47:17.41 & $26.1\pm0.3$ & $27.3\pm0.2$ & $26.7\pm0.2$ &
485: $>$2.1 & 0.5$\pm0.3$ & 0.08 & 0.39\\
486:
487: UDF-491-880\tablenotemark{\dagger} & 03:32:40.941 & -27:47:41.83 & $26.6\pm0.3$ &
488: $27.6\pm0.3$ & $26.9\pm0.2$ & $>$2.3 & 0.1$\pm0.3$ & 0.03 & 0.34\\
489: UDF-387-1125 & 03:32:42.565 & -27:47:31.42 & $26.6\pm0.3$ &
490: $27.4\pm0.2$ & $27.0\pm0.2$ & 1.4$\pm$0.4 & 0.8$\pm0.3$ & 0.68 & 0.28\\
491: UDF-983-964 & 03:32:38.794 & -27:47:07.14 & $27.1\pm0.3$ & $27.8\pm0.3$ & $27.1\pm0.2$ &
492: $>$2.1 & 0.0$\pm0.3$ & 0.48 & 0.27\\
493: UDF-818-886\tablenotemark{\dagger} & 03:32:39.292 & -27:47:22.12 & $27.1\pm0.3$
494: & $27.6\pm0.3$ & $27.3\pm0.3$ & $>$2.0 & 0.2$\pm0.3$ & 0.84 & 0.23\\
495: UDF-640-1417\tablenotemark{*} & 03:32:42.562 & -27:46:56.58 &
496: $26.0\pm0.3$ & $27.1\pm0.2$ & $26.5\pm0.2$ & 1.1$\pm$0.2 & 0.7$\pm0.2$ & 0.11 & 0.37\\
497: \enddata
498:
499: \tablenotetext{a}{All magnitudes are AB magnitudes. Right ascension
500: and declination use the J2000 equinox. Errors are $1\sigma$. Limits
501: on $z_{850}-J_{110}$ colors are 2$\sigma$. ``S/G'' denotes the
502: SExtractor stellarity parameter, for which 0 indicates an extended
503: object, and 1 a point source. ``Cor'' refers to a total magnitude
504: estimated using the Kron system (see \S2), ``Ap1'' refers to a
505: 0.6\arcsec-diameter aperture magnitude, and ``Ap2'' refers to a
506: 1.0\arcsec-diameter aperture magnitude. $z-J$ and $J-H$ colors were
507: estimated in a Kron aperture with Kron factor equal to 1.2 (similar to
508: the $0.6''$-diameter apertures used for ``Ap1'').}
509: \tablenotetext{\dagger}{These candidates are not found (UDF-818-886)
510: or seen to lower significance (UDF-491-880) in an independent
511: reduction of the NICMOS field kindly provided to us by Robberto et
512: al.\ (2004).}
513: \tablenotetext{*}{This object was very close to meeting
514: our selection criteria and could be a reddened starburst at
515: $z\sim6.5$. Another possibility is that of a dusty/evolved galaxy at
516: $z\sim1.6$. This object was also found by Yan \& Windhorst (2004).}
517: \end{deluxetable}
518: \end{document}
519: