1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[12pt,preprint2]{aastex}
3: %\documentclass[]{aastex}
4: %\usepackage{emulateapj5}
5:
6: \newcommand{\etal}{{et\thinspace al.} }
7: \newcommand{\Lya}{{$Ly\alpha$} }
8:
9: \begin{document}
10:
11: \title{The Major Sources of the Cosmic Reionizing Background at $z\simeq$ 6}
12:
13: \author{Haojing Yan}
14: \affil{SIRTF Science Center, California Institute of Technology, MS 100-22,
15: Pasadena, CA 91125}
16: \email{yhj@ipac.caltech.edu}
17: \author{Rogier A. Windhorst}
18: \affil{Department of Physics \& Astronomy, Arizona State University, Tempe, AZ
19: 85287}
20: \email{Rogier.Windhorst@asu.edu}
21:
22: \begin{abstract}
23:
24: In this paper, we address which sources contributed most of the reionizing
25: photons. Our argument assumes that the reionization ended around $z\simeq 6$
26: and that it was a relatively quick process, i.e., that there was a
27: non-negligible fraction of neutral hydrogen in the Universe at somewhat earlier
28: epochs. Starting from our earlier estimate of the luminosity function (LF) of
29: galaxies at $z\simeq 6$, we quantitatively show that the major sources of
30: reionization are most likely galaxies with $L<L_*$. Our
31: approach allows us to put stronger constraints to the LF of galaxies at
32: $z\simeq 6$. To have the Universe completely ionized at this redshift, the
33: faint-end slope of the LF should be steeper than $\alpha=-1.6$, which is the
34: value measured at lower redshifts ($z\simeq 3$), unless either the
35: normalization ($\Phi_*$) of the LF or the clumping factor of the ionized
36: hydrogen has been significantly underestimated. If $\Phi_*$ is actually lower
37: than what we assumed by a factor of two, a steep slope close to $\alpha=-2.0$
38: is required. Our LF
39: predicts a total of 50 -- 80 $z\simeq 6$ galaxies in the HST Ultra Deep Field
40: (UDF) to a depth of $AB=28.4$ mag, which can be used to constraint both
41: $\Phi_*$ and $\alpha$.
42: We conclude that the $least$ luminous galaxies existing at this redshift
43: should reach as low as some critical luminosity in order to accumulate the
44: entire reionizing photon budget. On the other hand, the existence of
45: significant amounts of neutral hydrogen at slightly earlier epochs, e.g.
46: $z\simeq 7$, requires that the $least$ luminous galaxies should not be fainter
47: than another critical value (i.e., the LF should cut-off at this point).
48: %The upcoming observation of the Ultra Deep Field (UDF) by the Hubble Space
49: %Telescope (HST) will be able to narrow down the possible range of the
50: %faint-end slope of the LF, and hence constrain these thresholds. Such
51: %quantitative results will thus help us understand the star-forming objects
52: %in early Universe that caused reionization.
53:
54: \end{abstract}
55:
56: \keywords{cosmology: high-redshift --- galaxies: luminosity function, mass function}
57:
58: \section{Introduction}
59:
60: High resolution UV spectra of $z>5$ quasars found in the Sloan
61: Digital Sky Survey (SDSS) provide solid evidence that the intergalactic
62: hydrogen was in a {\it completely ionized} state up to an early epoch of
63: $z\simeq 6$ (e.g., Fan \etal 2001). At slightly higher redshifts, however,
64: a small fraction of neutral hydrogen is being seen, as indicated by the
65: complete Gunn-Peterson trough found in several SDSS quasars at $z=6.28$ --
66: 6.43 (Becker \etal 2001; Fan \etal 2003). This has been used to argue that $z\simeq 6$ marks the
67: end of the reionization era, and that the redshift at which the reionization
68: began may not be much higher than $z\simeq 6$ (Becker \etal 2001; Fan \etal 2002).
69: Recently, Cen (2003) proposed that the Universe might have been
70: reionized twice, where the first epoch was completed at $z\simeq 15$--16 and the
71: second epoch at $z\simeq 6$. The first reionization is consistent with an
72: early beginning of reionization suggested by the recent Wilkinson Microwave
73: Anisotropy Probe (WMAP) results (e.g., Spergel \etal 2003),
74: while the second reionization is consistent with the SDSS results.
75:
76: It is still not certain what kind of objects provided the reionizing
77: background. Here we focus on the end of the reionization era at $z\simeq 6$.
78: There seems to be a consensus that the AGN population is not sufficient to
79: account for the entire required reionizing photon budget (e.g., Fan \etal 2002),
80: which leaves $normal$ star-forming galaxies as the only alternative. The
81: discovery of \Lya emitters at $z\simeq 6.5$ (Hu \etal 2002; Kodaira \etal 2003)
82: suggests that star-forming galaxies did exist at such early epochs, and the
83: significant number of faint Lyman-break galaxy (LBG) candidates at
84: $z\simeq 6$--6.5 found by Yan \etal (2003a) suggest that these objects could
85: indeed be responsible for the completion of the reionization at this epoch. In
86: this letter, we further study the contribution of galaxies and AGN to the
87: reionizing background, and point out
88: that dwarf galaxies may well have contributed the vast majority of the
89: reionizing photons. Our investigation provides significant constraints to the
90: luminosity function (LF) of galaxies at $z\simeq 6$, most importantly the
91: faint-end slope of the LF, and the critical luminosity of the $least$ luminous
92: galaxies that must have existed at $z\simeq 6$.
93:
94: This paper is organized as follows. In \S 2, we summarize our current best
95: knowledge of the LF of galaxies and AGN at $z\simeq 6$, which forms the base of
96: our discussion. The contribution of
97: these two types of objects to the total ionizing background is calculated in
98: \S 3, followed by a discussion on the implications of our result in \S 4. A
99: summary is given in \S 5. Throughout the paper, we adopt a cosmology of
100: $\Omega_M=0.27$, $\Omega_\Lambda=0.73$, and
101: $H_0=71$ km$\,$s$^{-1}\,$Mpc$^{-1}$.
102:
103: \section{The Luminosity Functions of Galaxies and AGN at $z\simeq 6$}
104:
105: Currently, the number of confirmed galaxies at $z\simeq 6$ is still
106: insufficient to derive their LF in a conventional way. Yan \etal (2002)
107: presented an $estimate$ of the LF of galaxies at $z\simeq 6$ in terms of their
108: cumulative surface density as a function of apparent magnitude. This estimate
109: was made by using a Schechter function,
110: \begin{displaymath}
111: \Phi(M)=0.921\cdot \Phi_*\cdot 10^{0.4(\alpha+1)(M_*-M)}\cdot exp\left[-10^{0.4(M_*-M)}\right],
112: \end{displaymath}
113: with the values of its three free parameters adopted as following: the
114: characteristic absolute magnitude $M_*$ (measured around rest-frame 1300\AA)
115: and the faint-end slope $\alpha$ obtained at $z\simeq 3$ (Steidel \etal 1996)
116: were assumed to be the same at $z\simeq 6$, and the scale factor $\Phi_*$ was
117: found by using the cumulative number density of $z\geq 5.5$ galaxies detected
118: in the Hubble Deep Field North (HDF-N) as the normalization. In the
119: cosmological model used in this paper, these three parameters are
120: $M_{AB}=-21.03$ mag, $\Phi_*=4.55\times 10^{-4}$ Mpc$^{-3}$ mag$^{-1}$, and
121: $\alpha=-1.6$.
122: %In the search of $z\simeq 6$ objects with the deep data obtained
123: %by the Advanced Camera for Surveys (ACS) on-board the HST, Yan \etal (2003a)
124: %pointed out that the faint-end slope could be as steep as $\alpha \simeq 2.0$.
125: The predicted cumulative surface density based on this estimated LF is
126: reproduced here in Fig. 1 (red curves) with a few revisions.
127:
128: All observational constraints available to date are also plotted in this
129: figure for comparison. Among them, those based on the SDSS quasar hosts, the
130: Keck \Lya emitters and the deep ACS parallel field were discussed in
131: Yan \etal (2002; 2003a\footnote{Note that the survey depth quoted in Yan \etal
132: (2003a) is incorrect. These ACS parallel data were taken at a gain of
133: 4 $e^-$/ADU, and we corrected for this before applying the magnitude
134: zeropoints that were measured at a gain of 1 $e^-$/ADU. However, this
135: correction turns out to be unnecessary, since the HST on-the-fly pipeline
136: already makes the extra step of dividing the gain value into the flat-fields,
137: which is not documented in the obvious places. As a consequence, the
138: magnitudes of the $z\simeq 6$ candidates reported in Yan \etal (2003a) should
139: be 1.5 mag brighter than their quoted values. Fig. 1 has reflected this change,
140: which makes the ACS parallel upper limit less restrictive. Other than these, the
141: major conclusion in that paper remains unchanged. This change does not affect
142: this current paper.}).
143: The three new upper limits at the bright-end are
144: discussed in Yan (2003b). In short, the NOAO-4m MOSAIC LBG and \Lya limits are
145: derived from a degree-sized intermediate-band survey centered on the HDF-North
146: and South (Yan \etal 2003, in preparation), and the Subaru LBG limit is what
147: we derived based on the deep Subaru data of the HDF-N (Capak \etal 2003). The
148: data point marked as a cross is derived from the Large Area Lyman-Alpha (LALA)
149: survey
150: (Rhoads \& Malhotra 2001; Rhoads \etal 2003). Recently, three more measurements
151: have become available. Stanway \etal (2003a, b) used the public Great
152: Observatories Origins Deep Survey (GOODS) single-epoch data to search for
153: $z\simeq 6$ objects. Bouwens \etal (2003) searched for such objects in their
154: deep ACS Guaranteed Time Obsevation (GTO) data. Most recently, the GOODS team
155: released its
156: $z\simeq 6$ candidate list derived from the stacked three-epoch ACS data
157: (Dickinson \etal 2003). All these new observations agree with our $z\simeq 6$ LF.
158: %(the result of Stanway \etal will not be discussed separately, as it is
159: %incorporated in the GOODS result (see the discussion of Dickinson \etal 2003).
160: %The data point of Bouwens \etal (2003) is shown as the solid triangle.
161: The GOODS candidate catalog has sufficient statistics to allow us to derive not
162: just one, but several constraints at different flux levels from 25.0 to 27.0
163: mag.
164: %Taking these candidates at face value, the inferred GOODS cumulative surface densities from 25.0 to 27.0 mag are shown as open pentagons in the figure.
165: All these values agree very well with our LF from 23.0 to 26.5 mag. At
166: $m_{AB} >26.5$ mag, the GOODS counts are significantly lower than our estimate,
167: which is possibly due to the incompleteness
168: of the catalog at the faint-end. To summarize, our LF broadly agrees with all
169: known constraints, and is about the best that one can get with the available
170: data so far.
171:
172: For our purpose, an analytic form of the AGN LF is preferred. The
173: ``standard" double power-law quasar LF (e.g., Boyle, Fong \& Shanks 1988)
174: is therefore used:
175: \begin{displaymath}
176: \Phi(M) = \frac{\Phi_*}{10^{0.4(\beta_1+1)(M-M_*)}+10^{0.4(\beta_2+1)(M-M_*)}
177: }.
178: \end{displaymath}
179: This form has been proved to be a very good representation for quasars up to at
180: least $z=2.5$. There are four free parameters involved: the turn-over
181: magnitude $M_*$, the scaling factor $\Phi_*$, the bright-end slope $\beta_1$,
182: and the faint-end slope $\beta_2$. The SDSS has discovered 6 quasars at $z>5.7$
183: in 2870 deg$^2$ to $AB=20.2$ mag (see Fan \etal 2003). Using this as the
184: normalization, and adopting $M_*=-23.9$ and $\beta_1=-2.58$ (Fan \etal 2001),
185: we obtain $\Phi_*=1.53\times 10^{-8}\, Mpc^{-3}\ mag^{-1}$.
186: The biggest uncertainty is the faint-end slope $\beta_2$,
187: which is essentially unknown for $z>4$. Therefore three values, $\beta_2=-1.58$,
188: $-2.0$ and $-2.58$, are used. The first one is the value observed at $z\simeq 2$
189: -- 3, the second one is the steepest value that we consider to be reasonable,
190: and the last one is the limiting case ($\beta_2=\beta_1$), which is considered
191: for illustrative purposes only. The surface densities of AGN at $z\simeq 6$
192: calculated based on these parameters are superposed in Fig. 1 as blue curves.
193:
194: \section{Ionizing Background at $z\simeq 6$ and the Contribution of Galaxies and AGN}
195:
196: We first need to answer the following question: in order
197: to keep the intergalactic hydrogen ionized at this redshift, what should be
198: the critical ionizing photon emission rate per unit co-moving volume
199: ($\dot N_{cri}$)? We can then find out if the integrated production rate of
200: photons at $\lambda < 912$\AA\ due to either AGN or galaxies meets this
201: critical value.
202:
203: A recipe of calculating $\dot N_{cri}$ has been given in Madau, Haardt \&
204: Rees (1999, their Eqn. 26; hereafter MHR99):
205: \begin{equation}
206: \dot N_{cri}(z)=10^{51.2}\ \left(\frac{C}{30}\right)\times\left(\frac{1+z}{6}\right)^3\left(\frac{\Omega_b h_{100}^2}{0.02}\right)^2\, s^{-1}\ Mpc^{-3},
207: \end{equation}
208: where $C$ is the ionized hydrogen clumping factor.
209: Choosing $\Omega_b h_{100}^2=0.02$, $C=30$ and $z=6$,
210: one finds $\dot N_{cri}=2.51\times 10^{51}\,s^{-1}\ Mpc^{-3}$. As noted by
211: MHR99, the time-dependent clumping factor in this expression has be written in
212: the form that is scaled to the value inferred at $z=5$ from the numerical
213: simulation of Gnedin \& Ostriker (1997), which gave $C\simeq 30$ at $z=5$.
214: %The actual value of $C$ might be higher, as this simulation was limited by the
215: %finite resolution. The uncertainty in $C$ will of course affect the value of
216: %$\dot N_{cri}$, and we will discuss its effect in the next section.
217:
218: The total ionizing photon production rate, $\dot N_i$, due to a given
219: population of objects, can be related to its differential LF, $\Phi(M)$, via
220: the following form (Yan 2003b) :
221: \begin{equation}
222: \dot N_i = B\int^{M_{max}}_{-\infty} 10^{-0.4(M+25+5logD_L)}\cdot \Phi(M)\ dM\,\,\, s^{-1}\ Mpc^{-3},
223: \end{equation}
224: where $M$ is absolute magnitude (in the AB system) defined around rest-frame
225: 1300\AA, $D_L$ is the luminosity distance (in Mpc) to the objects in question,
226: and $M_{max}$ is the absolute magnitude of the least luminous object that
227: should be used in the integration. $B$ is a constant whose value depends on the
228: shape of the UV-SED of these objects, which can be represented by power laws.
229: To be specific, $B=6.886\times 10^7\cdot D_L^2 \cdot (0.65)^\delta/(k\gamma)$,
230: where $\delta$ and $\gamma$ are power-law indices (in the frequency domain) for
231: $\lambda \leq $912\AA\, and 912\AA $\leq \lambda \leq$ 1400\AA, respectively.
232: The constant $k$ is the continuum discontinuity factor across the Lyman
233: limit at 912\AA.
234:
235: Following MHR99, we assume that the shape of the SED of a quasar has
236: power-law indices of --1.8 at $\lambda \leq$ 1050\AA\, and --0.8 at
237: 1050\AA$\leq\lambda<$2500\AA, respectively. Using these values and the AGN LF
238: described in \S 2, it is found that the contribution from AGN falls far short
239: of the critical value $\dot N_{cri}$ for any reasonable faint-end LF slope
240: (see also Lehnert \& Bremer (2003) for their discussion at $z\simeq 5$). Integrated to
241: $M=-16.0$ mag
242: beyond which the existence of AGN activity becomes implausible,
243: their total contribution are $0.004\times 10^{51}$ and
244: $0.009\times 10^{51}$ s$^{-1}$ Mpc$^{-3}$ for the
245: faint-end slopes of $\beta_2=-1.58$ and $-2.0$, respectively. Even for the
246: limiting case where $\beta_2=-2.58$, the derived $\dot N_{i}$ value is only
247: $0.055\times 10^{51}$ s$^{-1}$ Mpc$^{-3}$.
248:
249: Based on 29 LBG spectra at $z\simeq 3$, Steidel et al. (2001) measured
250: the ratio of the continuum flux at rest-frame 1300\AA\, and at the blue side of
251: the Lyman limit to be $\sim$ 4.6. This is equivalent to taking
252: $(0.65)^\delta/k=0.217$ in our expression for $B$, and is used hereafter. Note
253: that by using this value, we implicitly adopt the Lyman continuum photon
254: escaping fraction ($f_{esc}$) of Steidel et al. (2001), which is 10--13\% in
255: absolute terms. The index
256: $\gamma$ used in the following analysis was estimated by using a set of model
257: spectra of galaxies with ages of 0.1 Gyr (Bruzual \& Charlot 1993), and it
258: was found that this SED slope is close to an equivalent power-law index of $\gamma\simeq 1.8$.
259:
260: We find that the original LF estimate of Yan \etal (2002), which has a
261: relatively shallow slope of $\alpha=-1.6$, cannot produce a large enough
262: $\dot N_i$ that reaches $\dot N_{cri}$. This LF can account for about 67\% of
263: the critical value if Eqn. 2 is integrated to $M=-16.1$ mag, but the integral
264: increases only very slowly towards fainter magnitudes and never meets
265: $\dot N_{cri}$, even when pushed to globular-cluster-type luminosities
266: ($M\simeq -7$ mag). Modifying our LF by adopting a luminosity evolution scheme
267: of $L_*$, such as $L_*(z)=L_*(z$=3$)(1+z)^\rho/4^\rho$, will not mitigate this
268: problem, because the value of $\rho$ is negative in a reasonable hierarchical
269: structure formation model, and thus will make $M*$ fainter and the $\dot N_i$
270: value even less.
271:
272: Thus we explore other steeper $\alpha$ values to see if $\dot N_{cri}$ can be
273: met at a reasonable minimum luminosity. We consider $\alpha= -1.7, -1.8, -1.9$,
274: and $-2.0$, where the steepest one is the critical value at which the
275: integral of a Schechter function diverges if the integration is carried to
276: infinitely low luminosity. For these slopes to meet the normalizing condition
277: (cumulative surface density of 1.37 per arcmin$^2$ to $m_{AB}=27.0$ mag; see
278: Yan \etal 2002), $\Phi_*$ has to be 4.27, 4.00, 3.74, and
279: $3.49\times 10^{-4} Mpc^{-3}$, respectively. We find that $\dot N_{cri}$ can be
280: reached at the following critical absolute magnitudes for the above slopes,
281: respectively: $M=$ --8.8, --14.6, --15.8, and --16.6 mag (at $z\simeq 6$ these
282: corresponds to apparent magnitudes of 37.9, 32.1, 30.9, and 30.1 mag,
283: respectively). In other words, the $least$ luminous galaxies that the Universe
284: should have produced by this redshift should be at these magnitude levels or
285: fainter in order to have a fully ionized Universe at $z\simeq 6$.
286: These results are shown in Fig. 2 (top panel).
287:
288:
289: \section{Discussion}
290:
291: % We have quantitatively shown that AGN can only contribute an insignificant
292: %fraction (less than 10\%) of the total ionizing photon budget at $z\simeq 6$.
293: %This leaves $normal$ galaxies as the only possible population of major
294: %ionizing sources at this redshift. Since the vast majority of their contribution
295: %comes from less luminous, but more numerous objects, $normal$ galaxies can
296: %indeed account for the entire ionizing photon background provided that the
297: %faint-end slope of their LF is sufficiently steep. As discussed in \S 3.4, this
298: %slope should at least be steeper than --1.6.
299:
300: The quantitative results above depend on
301: three factors, namely, the clumpiness of ionized hydrogen ($C$), the
302: scaling factor of the LF of galaxies ($\Phi_*$), and the escaping fraction of
303: Lyman continuum photons ($f_{esc}$). The first one affects the value of
304: $\dot N_{cri}$, while the last two affect $\dot N_{i}$.
305:
306: The value of $\dot N_{cri}$ scales linearly with the clumping factor $C$.
307: In Eqn. 1, this clumping factor is scaled to its value at
308: $z=5$. We followed MHR99 and adopted $C=30$, which is suggested by the
309: numerical simulations of Gnedin \& Ostriker (1997). Should $C$ drop to $\sim$
310: 20, $\dot N_{cri}$ would drop by 1/3, and a faint-end slope steeper than
311: $\alpha=-1.6$ would no longer be required, since now $\dot N_i$ would reach the
312: critical absolute magnitude at $M=-16.1$ mag. Similarly, steeper slopes,
313: $\alpha=$ --1.7, --1.8, --1.9, and --2.0 would make $\dot N_{i}$ reach the
314: critical absolute magnitude at $M=$--17.0, --17.4, --17.7, and --17.9 mag, respectively.
315:
316: The value of $\dot N_i$ scales linearly with the scaling factor $\Phi_*$ of the LF.
317: Since increasing $\dot N_i$ has the same effect as decreasing $\dot N_{cri}$,
318: we get exactly the same answers as above, if the scaling factor increases by a
319: factor of 1.5. Since $\dot N_i$ also scales linearly with $f_{esc}$ (assuming
320: it is not a function of luminosity), the same answers hold if $f_{esc}$
321: increases by the same factor. On the other hand, if either $\Phi_*$ or
322: $f_{esc}$ drops by 1/3, a faint-end slope steeper than --1.8 is necessary to
323: reionize the Universe. In this case, $\dot N_{cri}$ can be reached at $M=-12.5$
324: and --14.7 mag for a slope of --1.9 and --2.0, respectively. Smaller $\Phi_*$
325: or $f_{esc}$ requires the LF extend to fainter luminosity; however, if either
326: of them is smaller by a factor of two, only $\alpha=-2.0$ can meet the
327: reionization requirement.
328:
329: % for example, if
330: %either quantity drops by a factor of two, the LF needs to extend to $M=-7.7$
331: %and $-12.7$ mag for $\alpha=-1.9$ and --2.0, respectively in order to reach
332: %$\dot N_{cri}$. Should either $\Phi_*$ or $f_{esc}$ be even smaller, only
333: %$\alpha=-2.0$ can meet the requirement.
334:
335: These values give the critical luminosity that the least luminous galaxies
336: should have, i.e., a lower limit to $M$. By considering the additional
337: constraint that the intergalactic hydrogen at slightly earlier epoch still had
338: a non-negligible neutral fraction, we can further improve these limits into
339: two-sided luminosity ranges. Cen (2003) gave a detailed modeling of the
340: evolution of intergalactic
341: medium in the double-reionization scheme, and the neutral fraction inferred
342: from this modeling is about 15\% at $z\simeq 7$. The precise neutral fraction
343: at a specific redshift does not matter too much for our purpose; the point is
344: that at an earlier epoch, e.g. at $z\simeq 7$, the total ionizing photon
345: production rate of galaxies should not exceed the critical value such that
346: a significant amount of neutral hydrogen can still exist, otherwise no GP-trough
347: would be seen at $z\simeq 6$.
348:
349: The calculation can be performed similarly to that described in \S 3. To
350: first order, we can assume that there is no evolution in the LF during the
351: relatively short time period from $z=6$ to 7, which is only 0.16 Gyr. At $z=7$,
352: $\dot N_{cri}=3.75\times 10^{51}$ s$^{-1}$ Mpc$^{-3}$ for the nominal $C$ value
353: of 30. Now the requirement is $N_{i}<N_{cri}$. If nothing else changes, a LF
354: faint-end slope of --1.8 or shallower will satisfy this criterion. Steeper
355: slopes of $\alpha=$ --1.9 and --2.0 require that the LF truncates before it
356: reaches $M=$ --12.5 and --14.6 mag, respectively. The bottom panel of Fig. 2
357: demonstrates these results.
358:
359: If $C=20$, or equivalently if either $\Phi_*$ or $f_{esc}$ is increased by
360: a factor of 1.5, it requires that the LF truncates before it reaches $M=$
361: --12.4, --15.8, --16.9, and --17.5 mag for slopes of --1.7, --1.8, --1.9, and
362: --2.0, respectively (a cut-off is not required for a slope of --1.6). On the
363: other hand, if either $\Phi_*$ or $f_{esc}$ is decreased by 1/3, or
364: equivalently if $C=45$, slopes of --1.9 and --2.0 require the LF truncate
365: before it reaches $M=$ --4.4 and --11.8 mag.
366:
367: Table 1 summarizes these results. We note that case a) is what we believe to
368: be the most plausible, given our best knowledge about the relevant parameters.
369: Case b) and c), on the other hand, explore a wider parameter space. However,
370: we point out that case b) is not very likely for several reasons.
371: First, the clumping factor is not likely to be as low as 20, since the nominal
372: value of $C=30$ obtained by Gnedin \& Ostriker (1997) might already have
373: underestimated the true value because of the finite resolution of their
374: simulation. Second, the scaling factor of the LF, $\Phi_*$, is not likely to be
375: significantly higher than the value suggested in Yan \etal (2002), otherwise
376: all the recent searches for $z\simeq 6$ objects would have resulted
377: in more bright ($z$-band $m_{AB}\leq 26.0$ mag) candidates than what have been
378: actually observed. Third, the Lyman photon escaping fraction of star-forming
379: galaxies is not likely to be much higher than $\sim$ 10--13\%, given that such
380: a value already seems to be rather high (cf. Giallongo \etal 2002).
381:
382: The Ultra Deep Field (UDF) campaign of the ACS/HST
383: (http://www.stsci.edu/hst/udf), whose data are now being taken and will be
384: released to the public in February 2004, will greatly narrow down the range of
385: $\alpha$. Depending on its the real value, we predict a total of 50
386: to 80 genuine $z\simeq 6$ objects in the UDF to its designed 10 $\sigma$ depth
387: of 28.4 mag.
388:
389: % We also note that a LF with a faint-end cut-off might not be desirable,
390: %especially if the luminosity range for the least luminous galaxies is narrow,
391: %since this would require the structure formation at the reionization epoch
392: %has a very sensitive self-regulating mechanism. With this additional
393: %consideration in mind, it seems that the faint-end slope of the LF is most
394: %likely between --1.7 and --1.8.
395: %We note that the Ultra Deep Field (UDF)
396: %campaign of the ACS/HST, whose data are now being taken and will be released to
397: %the public in early 2004, will narrow down the range of $\alpha$. Depends on
398: %its real value, our LF predicts a total number of 50 to 80 genuine $z\simeq 6$
399: %objects in the UDF to its designed depth of $m_{AB}(F850LP)=28.4$ mag.
400:
401: \section{Summary}
402:
403: In this paper we investigate the major sources of reionization at
404: $z\simeq 6$ based on the best available observational constraints.
405: Using the LF estimate proposed for galaxies at this redshift
406: (Yan \etal 2002) as the starting point, we find that $normal$ galaxies can
407: account for the entire reionizing background, provided that the faint-end slope
408: of their LF is steep enough and that the LF extends to sufficiently low
409: luminosity. We explore a range of faint-end slopes, from $\alpha=-1.6$, a value
410: that the LF has at $z\simeq 3$, to $\alpha=-2.0$, the critical value at
411: which the integration of a Schechter LF diverges. We find that dwarf galaxies,
412: rather than their more luminous counterparts or AGN, produced the vast
413: majority of the reionizing photons at $z\simeq 6$. Using the best estimate of
414: the relevant parameters, we find that the faint-slope of the LF of galaxies at
415: $z\simeq 6$ should be steeper than --1.6, if the adopted $\Phi_*$ holds, or
416: as steep as $\alpha=-2.0$ if $\Phi_*$ is lower by a factor of two.
417: In addition, we show that the least luminous galaxies should be
418: fainter than a certain critical luminosity, whose exact value depends on the
419: actual faint-end slope. Furthermore, by requiring that there still was a
420: non-negligible neutral hydrogen fraction at slightly earlier epochs
421: ($z\simeq 7$), we point out that the LF should truncate at a certain luminosity
422: threshold, whose value again depends on the actual faint-end slope.
423: The HST UDF data will be able to narrow down the range of this important
424: parameter, whose precise value is obviously important for planning the
425: performance of the James Webb Space Telescope.
426:
427: \acknowledgments
428: Funding from NASA/JWST Grant NAG5-12460 is acknowledged. The authors would like
429: to thank Dr. Mark Dickinson for pointing out the hidden step in the ACS pipeline
430: reduction that led to the improper zeropoint used in Yan \etal (2003a). We
431: also thank Drs. Piero Madau and Joe Silk for helpful discussions. We thank
432: the referee for the helpful comments.
433:
434: \begin{thebibliography}{}
435:
436: \bibitem[]{398} Becker,R. H., et al. 2001, AJ, 122, 2850
437: \bibitem[]{399} Bouwens, R. J., et al. 2003, ApJ, 595, 589
438: \bibitem[]{400} Boyle, B. J., Fong, R., \& Shanks, T. 1988, MNRAS, 231, 897
439: \bibitem[]{401} Bruzual, A. G., \& Charlot, S. 1993, ApJ, 405, 538
440: \bibitem[]{403} Cen, R. 2003, ApJ, 591, 12
441: \bibitem[]{404} Dickinson, M., et al. 2003 (astro-ph/0309070)
442: \bibitem[]{405} Fan, X., et al. 2001, AJ, 122, 2833
443: \bibitem[]{406} Fan, X., et al. 2002, AJ, 123, 1247
444: \bibitem[]{} Fan, X., et al. 2003, AJ, 125, 1649
445: \bibitem[]{408} Gnedin, N. Y., \& Ostriker, J. P. 1997, ApJ, 486, 581
446: \bibitem[]{410} Giallongo, E., Cristiani, S., D'Odorico, S., \& Fontana, A., 2002, ApJ, 568, L9
447: \bibitem[]{413} Hu, E. M., et al. 2002, ApJ, 568, L75
448: \bibitem[]{414} Kodaira, K., et al. 2003, PASJ, 55, L17
449: \bibitem[]{} Lehnet, M. D. \& Bremer, M 2003, ApJ, 593, 630
450: \bibitem[]{416} Madau, P., Haardt, F., \& Rees, M. J. 1999, ApJ, 514, 648
451: \bibitem[]{418} Rhoads, J. E., et al. 2003, AJ, 125, 1006
452: \bibitem[]{} Rhoads, J. E. \& Malhotra, S. 2001, ApJL, 563, 5
453: \bibitem[]{419} Spergel, D. N., et al. 2003, ApJS, 148, 175
454: \bibitem[]{420} Stanway, E. R., Bunker, A. J., \& McMahon, R. G., 2003a, MNRAS, 342, 439
455: \bibitem[]{421} Stanway, E. R., Bunker, A. J., McMahon, R. G., Ellis, R. S., Treu, T., \& McCarthy, P. J., 2003b (astro-ph/0308124)
456: \bibitem[]{422} Steidel, C. C., Pettini, M., \& Adelberger, K. L. 2001, ApJ, 546, 665
457: \bibitem[]{423} Yan, H., Windhorst, R. A., Odewahn, S. C., Cohen, S. H., R\"{o}ttgering, H. J. A., \& Keel, W. C. 2002, ApJ, 580, 725
458: \bibitem[]{424} Yan, H., Windhorst, R. A., \& Cohen, S. H. 2003a, ApJ, 585, L93
459: \bibitem[]{425} Yan, H. 2003b, Ph. D. Dissertation, Arizona State University
460: \end{thebibliography}
461:
462: \clearpage
463: \begin{figure}
464: \epsscale{.90}
465: \plotone{f1.eps}
466: \caption{The best estimate of the LF of galaxies at $z\simeq 6$ (red curves),
467: based on Yan \etal (2002), agrees with all known observational constraints.
468: Two faint-end slopes, $\alpha=-1.6$ (solid line) and $-2.0$ (long dashed line),
469: are shown. Only new constraints are discussed here. The upper limits marked
470: (1) and (2) are derived from a degree-sized intermediate-band survey
471: (Yan \etal 2003, in preparation), while the one marked (3) is derived based on
472: the deep Subaru data of Capak \etal (2003). The cross is derived from the LALA
473: survey (Rhoads \etal 2003). The solid triangle is based on Bouwens \etal (2003).
474: The open pentagons connected by line are based on the results of the GOODS
475: (Dickinson \etal 2003). The upper limit based on the ACS parallel observation
476: of Yan \etal (2003a) has been corrected to $AB=26.5$ mag to account for the
477: proper magnitude zeropoint. This LF predicts that 50--80 $z\simeq 6$ objects
478: will be found by the UDF campaign. For comparison, the blue curves
479: show the expected number density of quasars at $z\simeq 6$, derived by using a
480: double-power law LF and the normalization based on the SDSS $z\simeq 6$ quasar
481: sample. Three different faint-end slopes are assumed, $\beta_2=-1.58$ (solid
482: curve), --2.0 (dashed curve), and --2.58 (dotted line). A truncation at $M=-16$
483: mag is applied, as beyond this point the existence of significant AGN activity
484: is implausible.
485: }
486: \end{figure}
487:
488: \begin{figure}
489: \epsscale{.90}
490: \plotone{f2.eps}
491: \caption{(Top) At $z\simeq 6$, the reionizing photon production rate due to
492: galaxies should meet the critical value in order to keep the Universe fully
493: ionized. The solid, long-dashed and short-dashed horizontal lines are the
494: critical values if the clumping factor (at $z=5$) $C=30$, 45 and 20,
495: respectively. The contribution from galaxies is shown in color, assuming
496: different faint-end slopes of their LF: --1.6 (blue), --1.7 (green), --1.8
497: (cyan), --1.9 (magenta), and --2.0 (red). Galaxies can account for the entire
498: reionizing photon budget provided that their LF is sufficiently steep
499: ($\alpha\leq -1.7$), and extends to sufficiently low luminosity.
500: %i.e., there is a minimum luminosity that the least luminous galaxies should at least reach.
501: Quasars can only contribute
502: an insignificant fraction of the ionizing photons, as the black curves show:
503: the solid, long-dashed and dotted lines are for faint-end slopes of --1.58,
504: --2.0 and --2.58, respectively. (Bottom) The reionizing photon production rate
505: due to galaxies at a slightly earlier epoch should be smaller than the
506: corresponding critical value, as there was still a significant fraction of
507: neutral hydrogen. The case is shown for $z\simeq 7$. Legends are the same as
508: in the top panel. The LF of galaxies should cut-off at a certain
509: luminosity, i.e., there is a luminosity threshold for the $least$ luminous
510: galaxies at this redshift, below which the LF cannot continue or it would leave
511: no neutral hydrogen at $z\simeq 7$ and no GP-trough in SDSS quasars at $z > 6$.
512: }
513: \end{figure}
514:
515: \begin{table}
516: \caption{The luminosity range of the least luminous galaxies that should exist at reionization}
517: \begin{center}
518: \begin{tabular}{ccccc}\tableline \tableline
519:
520: $\alpha$ & $\Phi_*^\dagger$ & $M_{min}$\tablenotemark{a} & $M_{min}$\tablenotemark{b} & $M_{min}$\tablenotemark{c} \\ \tableline
521: --1.6 & 4.55 & N/A & $-16.1\leq M_{min}$ & N/A \\
522: --1.7 & 4.27 & $-8.8\leq M_{min}$ & (--17.0, --9.1) & N/A \\
523: --1.8 & 4.00 & $-14.6\leq M_{min}$ & (--17.4, --14.6) & N/A \\
524: --1.9 & 3.74 & (--15.8, --12.6) & (--17.7, --15.9) & (--12.5, --4.4) \\
525: --2.0 & 3.49 & (--16.6, --14.7) & (--17.9, --16.6) & (--14.7, --11.8) \\
526: \tableline
527:
528: \end{tabular}
529: \end{center}
530: \tablenotetext{\dagger} {Nominal scaling factor of the LF (in unit of $10^{-4} Mpc^{-3}$) by forcing an accumulative number density of 1.37 per arcmin$^2$ to
531: $m_{AB}=27.0$ mag (Yan \etal 2002).}
532: \tablenotetext{a} {For the case where $C=30$, and $\Phi_*$ and $f_{esc}$ have
533: their nominal values (see text). The minimum absolute magnitudes are referred
534: to rest-frame wavelength at around 1300\AA, and are in the AB system.}
535: \tablenotetext{b} {For $C=20$, or equivalently either $\Phi_*$ or $f_{esc}$
536: increased by a factor of 1.5.}
537: \tablenotetext{c} {For $C=45$, or equivalently either $\Phi_*$ or $f_{esc}$
538: decreased by 1/3. If $\Phi_*$ decreases by a factor of two, only $\alpha=-2.0$
539: can meet the reionization requirement.}
540:
541: \end{table}
542:
543: \end{document}
544: