1: \documentstyle[12pt,aasms4]{article}
2:
3: \def\gsim{\;\rlap{\lower 2.5pt
4: \hbox{$\sim$}}\raise 1.5pt\hbox{$>$}\;}
5: \def\lsim{\;\rlap{\lower 2.5pt
6: \hbox{$\sim$}}\raise 1.5pt\hbox{$<$}\;}
7:
8: \def\ge{\;\rlap{\lower 2.5pt
9: \hbox{$-$}}\raise 1.5pt\hbox{$>$}\;}
10: \def\le{\;\rlap{\lower 2.5pt
11: \hbox{$-$}}\raise 1.5pt\hbox{$<$}\;}
12: \def\HI{\ion{H}{1}~}
13: \def\HII{\ion{H}{2}~}
14:
15: \newcommand\beq{\begin{equation}}
16: \newcommand\eeq{\end{equation}}
17: \def\lya{Ly$\alpha$~}
18: \def\v{\vspace{-0.1in}}
19:
20: \begin{document}
21:
22: \Large \centerline{\bf A Size of $\sim10$Mpc for the Ionized Bubbles}
23: \Large \centerline{\bf at the End of Cosmic Reionization}
24:
25: \normalsize
26: \author{\bf J. Stuart B. Wyithe$^\dagger$ and Abraham Loeb$^\star$}
27: \medskip
28: \noindent
29: $\dagger$ School of Physics, The University of Melbourne, Parkville, Vic
30: 3010, Australia \\ $\star$ Astronomy Dept., Harvard University, 60 Garden
31: Street, Cambridge, MA 02138, USA\\
32: %\centerline{submitted to {\it Nature}, October 2003}
33:
34: \vskip 0.2in
35: \hrule
36: \vskip 0.2in
37:
38: {\bf The first galaxies to appear in the universe at redshifts $z\ga 20$
39: created ionized bubbles in the intergalactic medium (IGM) of neutral
40: hydrogen (\HI) left over from the Big-Bang. It is thought that the ionized
41: bubbles grew with time, surrounded clusters of dwarf
42: galaxies\cite{BL1}$^,$\cite{FZH1} and eventually overlapped quickly
43: throughout the universe over a narrow redshift interval near $z\sim
44: 6$. This event signaled the end of the reionization epoch when the universe
45: was a billion years old. Measuring the hitherto unknown size distribution
46: of the bubbles at their final overlap phase is a focus of forthcoming
47: observational programs aimed at highly redshifted 21cm emission from atomic
48: hydrogen. Here we show that the combined constraints of cosmic variance
49: and causality imply an observed bubble size at the end of the overlap epoch
50: of $\sim 10$ physical Mpc, and a scatter in the observed redshift of
51: overlap along different lines-of-sight of $\sim 0.15$. This scatter is
52: consistent with observational constraints from recent spectroscopic data on
53: the farthest known quasars. Our novel result implies that future radio
54: experiments should be tuned to a characteristic angular scale of $\sim
55: 0.5^\circ$ and have a minimum frequency band-width of $\sim 8$ MHz for an
56: optimal detection of 21cm flux fluctuations near the end of reionization.}
57:
58: During the reionization epoch, the characteristic bubble size (defined
59: here as the spherically averaged mean radius of the \HII regions that
60: contain most of the ionized volume\cite{FZH1}) increased with time as
61: smaller bubbles combined until their overlap completed and the diffuse
62: IGM was reionized. However the largest size of isolated bubbles
63: (fully surrounded by \HI boundaries) that can be {\em observed} is
64: finite, because of the combined phenomena of cosmic variance and
65: causality. Figure 1 presents a schematic illustration of the
66: geometry. There is a surface on the sky corresponding to the time
67: along different lines-of-sight when the diffuse (uncollapsed) IGM was
68: {\em most recently neutral}. We refer to it as the Surface of Bubble
69: Overlap (SBO). There are two competing sources for fluctuations in the
70: SBO, each of which is dependent on the characteristic size, $R_{\rm
71: SBO}$, of the ionized regions just before the final overlap. First,
72: the finite speed of light implies that photons observed from different
73: points along the curved boundary of an \HII region must have been
74: emitted at different times during the history of the universe.
75: Second, bubbles on a co-moving scale $R$ achieve reionization over a
76: spread of redshifts due to cosmic variance in the initial conditions
77: of the density field smoothed on that scale. The characteristic scale
78: of \HII bubbles grows with time, leading to a decline in the spread of
79: their formation redshifts\cite{BL1} as the cosmic variance is averaged
80: over an increasing spatial volume. However the light-travel time
81: across a bubble rises concurrently. Suppose a signal photon which
82: encodes the presence of neutral gas (e.g. a 21cm line photon), is emitted
83: from the far edge of the ionizing bubble. If the adjacent region along
84: the line-of-sight has not become ionized by the time this photon
85: reaches the near side of the bubble, then the photon will encounter
86: diffuse neutral gas. Other photons emitted at this lower redshift will
87: therefore also encode the presence of diffuse neutral gas, implying
88: that the first photon was emitted prior to overlap, and not from the
89: SBO. Hence the largest observable scale of \HII regions when their
90: overlap completes, corresponds to the first epoch at which the light
91: crossing time becomes larger than the spread in formation times of
92: ionized regions. Only then will the signal photon leaving the far
93: side of the HII region have the lowest redshift of any signal photon
94: along that line-of-sight.
95:
96: The observed spectra of all quasars beyond $z\sim6.1$ each show a
97: Gunn-Peterson trough\cite{GP}$^,$\cite{f2}, a blank spectral region at
98: wavelengths shorter than \lya at the quasar redshift, indicating the
99: presence of \HI in the diffuse IGM. The detection of Gunn-Peterson
100: troughs indicates a rapid change\cite{f1}$^,$\cite{Pe}$^,$\cite{WB1}
101: in the neutral content of the IGM at $z\sim6$, and hence a rapid
102: change in the intensity of the background ionizing flux. This rapid
103: change implies that overlap, and hence the reionization epoch,
104: concluded near $z\sim6$. The most promising observational
105: probe\cite{ZFH}$^,$\cite{MH1} of the reionization epoch is redshifted
106: 21cm emission from intergalactic \HI. Future observations using low
107: frequency radio arrays (e.g. the Low Frequency Array, LOFAR) will
108: allow a direct determination of the topology and duration of the phase
109: of bubble overlap. Here we determine the expected angular scale and
110: redshift width of the 21cm fluctuations at the SBO theoretically, and
111: show that our determination is consistent with current observational
112: constraints.
113:
114: We start by quantifying the constraints of causality and cosmic
115: variance. First suppose we have an \HII region with a physical radius
116: $R/(1+\langle z\rangle)$. The light crossing time of this radius is
117: \begin{equation}
118: \label{causality}
119: \langle\Delta z^2\rangle^{1/2} =
120: \left|\frac{dz}{dt}\right|_{\langle z\rangle}
121: \frac{R}{c(1+\langle z\rangle)},
122: \end{equation}
123: where at the high-redshifts of interest
124: $(dz/dt)=-(H_0\sqrt{\Omega_m})(1+z)^{5/2}$. Here, $c$ is the speed of
125: light, $H_0$ is the present-day Hubble constant, $\Omega_m$ is the present
126: day matter density parameter, and $\langle z\rangle$ is the mean redshift
127: of the SBO. Note that when discussing this crossing time, we are referring
128: to photons used to probe the ionized bubble (e.g. at 21cm), rather than
129: photons involved in the dynamics of the bubble evolution.
130:
131: Second, overlap would have occurred at different times in different regions
132: of the IGM due to the cosmic scatter in the process of structure formation
133: within finite spatial volumes\cite{BL1}. Reionization should be completed
134: within a region of co-moving radius $R$ when the fraction of mass
135: incorporated into collapsed objects in this region attains a certain
136: critical value, corresponding to a threshold number of ionizing photons
137: emitted per baryon. The ionization state of a region is governed by the
138: enclosed ionizing luminosity, by its over-density, and by dense pockets of
139: neutral gas that are self shielding to ionizing radiation. There is an
140: offset\cite{BL1} $\delta z$ between the redshift when a region of mean
141: over-density $\bar{\delta}_{\rm R}$ achieves this critical collapsed
142: fraction, and the redshift ${\bar z}$ when the universe achieves the same
143: collapsed fraction on average. This offset may be computed\cite{BL1} from
144: the expression for the collapsed fraction\cite{BCE1} $F_{\rm col}$ within a
145: region of over-density $\bar{\delta}_{\rm R}$ on a co-moving scale $R$,
146: \begin{equation}
147: \label{scatter}
148: F_{\rm col}(M_{\rm min})=\mbox{erfc}\left[\frac{\delta_{\rm
149: c}-\bar{\delta}_{\rm R}}{\sqrt{2[\sigma_{\rm R_{\rm min}}^2-\sigma_{\rm
150: R}^2]}}\right],\hspace{5mm}\mbox{yielding}\hspace{5mm} \frac{\delta
151: z}{(1+\bar{z})}=\frac{\bar{\delta}_{\rm R}}{\delta_{\rm
152: c}(\bar{z})}-\left[1-\sqrt{1-\frac{\sigma_{\rm R}^2}{\sigma_{\rm R_{\rm
153: min}}^2}}\right],
154: \end{equation}
155: where $\delta_{\rm c}(\bar{z})\propto (1+\bar{z})$ is the collapse
156: threshold for an over-density at a redshift $\bar{z}$; $\sigma_{\rm R}$ and
157: $\sigma_{R_{\rm min}}$ are the variances in the power-spectrum linearly
158: extrapolated to $z=0$ on co-moving scales corresponding to the region of
159: interest and to the minimum galaxy mass $M_{\rm min}$, respectively. The
160: offset in the ionization redshift of a region depends on its linear
161: over-density, $\bar{\delta}_{\rm R}$. As a result, the distribution of
162: offsets, and therefore the scatter in the SBO may be obtained directly from
163: the power spectrum of primordial inhomogeneities. As can be seen from
164: equation~(\ref{scatter}), larger regions have a smaller scatter due to
165: their smaller cosmic variance.
166:
167: Note that equation~(\ref{scatter}) is independent of the critical
168: value of the collapsed fraction required for reionization. Moreover,
169: our numerical constraints are very weakly dependent on the minimum
170: galaxy mass, which we choose to have a virial temperature of $10^4$K
171: corresponding to the cooling threshold of primordial atomic gas. The
172: growth of an \HII bubble around a cluster of sources requires that the
173: mean-free-path of ionizing photons be of order the bubble radius or
174: larger. Since ionizing photons can be absorbed by dense pockets of
175: neutral gas inside the \HII region, the necessary increase in the
176: mean-free-path with time implies that the critical collapsed fraction
177: required to ionize a region of size $R$ increases as well. This larger
178: collapsed fraction affects the redshift at which the region becomes
179: ionized, but not the scatter in redshifts from place to place which is
180: the focus of this {\em Letter}. Our results are therefore independent
181: of assumptions about unknown quantities such as the star formation
182: efficiency and the escape fraction of ionizing photons from galaxies,
183: as well as unknown processes of feedback in galaxies and clumping of
184: the IGM.
185:
186: Figure~2 displays our two fundamental constraints. The causality
187: constraint (Eq.~\ref{causality}) is shown as the blue line, giving a
188: longer crossing time for a larger bubble size. This contrasts with the
189: constraint of cosmic variance (Eq.~\ref{scatter}), indicated by the
190: red line, which shows how the scatter in formation times decreases
191: with increasing bubble size. The scatter in the SBO redshift and the
192: corresponding fluctuation scale of the SBO are given by the
193: intersection of these curves. We find that the thickness of the SBO is
194: $\langle\Delta z^2\rangle^{1/2}\sim0.13$, and that the bubbles which
195: form the SBO have a characteristic co-moving size of $\sim60$Mpc
196: (equivalent to 8.6 physical Mpc). At $z\sim6$ this size corresponds to
197: angular scales of $\theta_{\rm SBO}\sim0.4$ degrees on the sky. We
198: have also examined a second model for the formation of \HII regions,
199: where ionization is reached when the collapsed fraction divided by the
200: density contrast exceeds a critical value. This model allows for the
201: possibility that the increased recombination rate offsets reionization
202: in overdense regions.
203: The increased range of formation times in this case leads
204: to a slightly larger value for the scale of overlapping \HII regions,
205: and we find $\langle\Delta z^2\rangle^{1/2}\sim0.2$, $R_{\rm
206: SBO}\sim90$Mpc and $\theta_{\rm SBO}\sim0.6$ degrees.
207:
208: A scatter of $\sim0.15$ in the SBO is somewhat larger than the value
209: extracted from existing numerical
210: simulations\cite{g1}$^,$\cite{y1}. The difference is most likely due
211: to the limited size of the simulated volumes; while the simulations
212: appropriately describe the reionization process within limited regions
213: of the universe, they are not sufficiently large to describe the
214: global properties of the overlap phase\cite{BL1}. The scales over
215: which cosmological radiative transfer has been simulated are smaller
216: than the characteristic extent of the SBO, which we find to be $R_{\rm
217: SBO}\sim70$ co-moving Mpc.
218:
219: We can constrain the scatter in the SBO redshift observationally using
220: the spectra of the highest redshift quasars. Since only a trace amount
221: of neutral hydrogen is needed to absorb Ly$\alpha$ photons, the time
222: where the IGM becomes \lya transparent need not coincide with bubble
223: overlap. Following overlap the IGM was exposed to ionizing sources in
224: all directions and the ionizing intensity rose rapidly. After some
225: time the ionizing background flux was sufficiently high that the \HI
226: fraction fell to a level at which the IGM allowed transmission of
227: resonant \lya photons. This is shown schematically in Figure~1. The
228: lower wavelength limit of the Gunn-Peterson trough corresponds to the
229: \lya wavelength at the redshift when the IGM started to allow
230: transmission of \lya photons {\em along that particular
231: line-of-sight}. In addition to the SBO we therefore also define the
232: Surface of \lya Transmission (hereafter SLT) as the redshift along
233: different lines-of-sight when the diffuse IGM became transparent to
234: Ly$\alpha$ photons.
235:
236: The scatter in the SLT redshift is an observable which we would like to
237: compare with the scatter in the SBO redshift. The variance of the density
238: field on large scales results in the biased clustering of
239: sources\cite{BL1}. \HII regions grow in size around these clusters of
240: sources. In order for the ionizing photons produced by a cluster to advance
241: the walls of the ionized bubble around it, the mean-free-path of these
242: photons must be of order the bubble size or larger. After bubble overlap,
243: the ionizing intensity at any point grows until the ionizing photons have
244: time to travel across the scale of the new mean-free-path, which represents
245: the horizon out to which ionizing sources are visible. Since the
246: mean-free-path is larger than $R_{\rm SBO}$, the ionizing intensity at the
247: SLT averages the cosmic scatter over a larger volume than at the SBO. This
248: constraint implies that the cosmic variance in the SLT redshift must be
249: smaller than the scatter in the SBO redshift. However, it is possible that
250: opacity from small-scale structure contributes additional scatter to the
251: SLT redshift.
252:
253: If cosmic variance dominates the observed scatter in the SLT redshift, then
254: based on the spectra of the three $z>6.1$ quasars\cite{f2}$^,$\cite{WB1} we
255: would expect the scatter in the SBO redshift to satisfy $\langle\Delta
256: z^2\rangle^{1/2}_{\rm obs}\ga0.05$. In addition, analysis of the {\it
257: proximity effect} for the size of the \HII regions around the two highest
258: redshift quasars\cite{WL1}$^,$\cite{mh1} implies a neutral fraction that is
259: of order unity (i.e. pre-overlap) at $z\sim6.2-6.3$, while the transmission
260: of Ly$\alpha$ photons at $z\la6$ implies that overlap must have completed
261: by that time. This restricts the scatter in the SBO to be $\langle\Delta
262: z^2\rangle^{1/2}_{\rm obs}\la0.25$. The constraints on values for the
263: scatter in the SBO redshift are shaded gray in Figure~\ref{fig2}. It is
264: reassuring that the theoretical prediction for the SBO scatter of
265: $\langle\Delta z^2\rangle^{1/2}_{\rm obs}\sim0.15$, with a characteristic
266: scale of $\sim70$ co-moving Mpc, is bounded by these constraints.
267:
268: The presence of a significantly neutral IGM just beyond the redshift
269: of overlap\cite{WL1}$^,$\cite{mh1} is encouraging for upcoming 21cm
270: studies of the reionization epoch as it results in emission near an
271: observed frequency of 200 MHz where the signal is most readily
272: detectable. Future observations of redshifted 21cm line emission at
273: $6\la z\la 6.5$ with instruments such as the Low Frequency Array
274: (LOFAR), will be able to map the three-dimensional distribution of HI
275: at the end of reionization. The intergalactic \HII regions will
276: imprint a 'knee' in the power-spectrum of the 21cm anisotropies on a
277: characteristic angular scale corresponding to a typical isolated \HII
278: region\cite{ZFH}. Our results suggest that this characteristic angular
279: scale is large at the end of reionization, $\theta_{\rm SBO}\sim
280: 0.5$ degrees, motivating the construction of compact low
281: frequency arrays. An SBO thickness of $\langle \Delta
282: z^2\rangle^{1/2}\sim0.15$ suggests a minimum frequency
283: band-width of $\sim8$ MHz for experiments aiming to detect
284: anisotropies in 21cm emission just prior to overlap. These results
285: will help guide the design of the next generation of low-frequency
286: radio observatories in the search for 21cm emission at the end of the
287: reionization epoch.
288:
289:
290: %\noindent {\it Correspondence and requests for materials to Stuart Wyithe
291: %or Abraham Loeb.}
292:
293:
294: \small
295: \noindent
296: \begin{thebibliography}{}
297:
298:
299:
300:
301: \v
302: \bibitem[$^{1}$]{BL1}\cite{BL1} Barkana, R., Loeb, A., Unusually large
303: fluctuations in the statistics of galaxy formation at high redshift,
304: {\it Astrophys. J.}, {\bf 609}, 474-481, (2004)
305:
306: \v
307: \bibitem[$^{2}$]{FZH1}\cite{FZH1} Furlanetto, S.R., Zaldarriaga, M.,
308: Hernquist, L., The growth of \HII regions during reionization, {\it
309: Astrophys. J.}, submitted, astro-ph/0403697
310:
311: \v
312: \bibitem[$^{3}$]{GP}\cite{GP} Gunn, J. E., Peterson, B. A., On the
313: density of neutral hydrogen in intergalactic space, {\it Astrophys. J.}
314: {\bf 142}, 1633-1641 (1965)
315:
316: \v
317: \bibitem[$^{4}$]{f2}\cite{f2} Fan, X., \textit{et al.}, A survey of $z>5.7$ quasars in the Sloan Digital Sky Survey III: discovery of five additional quasars, {\it Astron. J.}, in press, astro-ph/0405138
318:
319: \v
320: \bibitem[$^{5}$]{f1}\cite{f1} Fan, X., \textit{et al.}, Evolution of the
321: ionizing background and the epoch of reionization from the spectra of
322: $z\sim 6$ quasars, {\it Astron. J.} {\bf 123}, 1247-1257 (2002)
323:
324: \v
325: \bibitem[$^{6}$]{Pe}\cite{Pe} Pentericci, L.~et al., VLT optical and
326: near-infrared observations of the $z=6.28$ quasar SDSS J1030+0524, {\it
327: Astron. J.}, {\bf 123}, 2151-2158 (2002)
328:
329: \v
330: \bibitem[$^{7}$]{WB1}\cite{WB1} White, R.L., Becker, R.H., Fan, X.,
331: Strauss, M.A., Probing the ionization state of the universe at $z>6$,
332: {\it Astron. J.} {\bf 126}, 1-14, (2003)
333:
334: \v
335: \bibitem[$^{8}$]{ZFH}\cite{ZFH} Zaldarriaga, M., Furlanetto, S.R.,
336: Hernquist, L., 21 Centimeter Fluctuations from Cosmic Gas at High
337: Redshifts, {\it Astrophys. J.}, submitted, astro-ph/0311514
338:
339: \v
340: \bibitem[$^{9}$]{MH1}\cite{MH1}
341: Morales, M.F., Hewitt, J., Toward Epoch of Reionization Measurements with Wide-Field Radio Observations, {\it Astrophys. J.}, accepted, astro-ph/0312437
342:
343: \v
344: \bibitem[$^{10}$]{BCE1}\cite{BCE1}
345: Bond, J. R.; Cole, S.; Efstathiou, G.; Kaiser, N., Excursion set mass functions for hierarchical Gaussian fluctuations, {\it Astrophys. J.}, {\bf 379}, 440-460 (1991)
346:
347: \v
348: \bibitem[$^{11}$]{g1}\cite{g1} Gnedin, N.Y., Cosmological reionization by stellar sources, {\it Astrophys. J.}, {\bf 535}, 530-554, (2000)
349:
350: \v
351: \bibitem[$^{12}$]{y1}\cite{y1} Yoshida, N., Sokasian, A., Hernquist, L., Springel, V., Early structure formation and reionization in a cosmological model with a running primordial power spectrum, {\it Astrophys. J.}, {\bf 598}, 73-85, (2003)
352:
353: \v
354: \bibitem[$^{13}$]{WL1}\cite{WL1} Wyithe, J.S.B., Loeb, A., A large neutral fraction of cosmic hydrogen a billion years after the Big Bang, {\it Nature} {\bf 427}, 815-817, (2004)
355:
356: \v
357: \bibitem[$^{14}$]{mh1}\cite{mh1} Mesinger, A., Haiman, Z., Evidence for a cosmological Stromgren surface and for significant neutral hydrogen surrounding the quasar SDSS J1030+0524, {\it Astrophys. J. Lett.}, in press, astro-ph/0406188
358:
359: \v
360: \bibitem[$^{15}$]{ME1}\cite{ME1} Miralda-Escude, J., On the Evolution of
361: the Ionizing Emissivity of Galaxies and Quasars Required by the Hydrogen
362: Reionization, {\it Astrophys. J.}, {\bf 597}, 66-73, (2003)
363:
364: \v
365: \bibitem[$^{16}$]{BL2}\cite{BL2} Barkana, R., Loeb, A., GRBs verses
366: quasars: Lyman-$\alpha$ signatures of reionization verses cosmological
367: infall, {\it Astrophys. J.}, {\bf 601}, 64-77 (2004)
368:
369: \v
370: \bibitem[$^{17}$]{BL3}\cite{BL3} Bromm, V., Loeb, A., The expected
371: redshift distribution of Gamma-Ray Bursts, {\it Astrophys. J.}, {\bf
372: 575}, 111-116, (2002)
373:
374: \v
375: \bibitem[$^{18}$]{SB1}\cite{SB1} Spergel, D. N, et al., First-year
376: Wilkinson microwave anisotropy probe (WMAP) observations:
377: determination of cosmological parameters, {\it Astrophys. J. Supp.}, {\bf
378: 148}, 175-194 (2003)
379:
380: \end{thebibliography}
381:
382:
383:
384: \normalsize
385: \vskip 0.2in
386: \noindent
387: {ACKNOWLEDGMENTS.} This work was supported in part by grants from ARC,
388: NSF and NASA.
389:
390:
391: \vskip 1in
392:
393: %FIGURE 1
394: \begin{figure*}[htbp]
395: \epsscale{1.}
396: \plotone{fig1.eps}
397: \end{figure*}
398:
399: %FIGURE 1
400: \begin{figure*}[htbp]
401: \caption{\label{fig1} The distances to the observed Surface of Bubble
402: Overlap (SBO) and Surface of Ly$\alpha$ Transmission (SLT) fluctuate
403: on the sky. The SBO corresponds to the first region of diffuse neutral
404: IGM {\em observed} along a random line-of-sight. It fluctuates across
405: a shell with a minimum width dictated by the condition that the light
406: crossing time across the characteristic radius $R_{\rm SBO}$ of
407: ionized bubbles equals the cosmic scatter in their formation
408: times. Thus, {\it causality} and {\it cosmic variance} determine the
409: characteristic scale of bubbles at the completion of bubble overlap.
410: After some time delay the IGM becomes transparent to Ly$\alpha$
411: photons, resulting in a second surface, the SLT. The upper panel
412: illustrates how the lines-of-sight towards two quasars (Q1 in red and
413: Q2 in blue) intersect the SLT with a redshift difference $\delta
414: z$. The resulting variation in the observed spectrum of the two
415: quasars is shown in the lower panel. Observationally, the ensemble of
416: redshifts down to which the Gunn-Peterson troughs are seen in the
417: spectra of $z>6.1$ quasars is drawn from the probability distribution
418: $dP/dz_{\rm SLT}$ for the redshift at which the IGM started to allow
419: \lya transmission along random lines-of-sight. The observed values of
420: $z_{\rm SLT}$ show a small scatter\cite{f2} in the SLT redshift around
421: an average value of $\langle z_{\rm SLT}\rangle\approx 5.95$. Some
422: regions of the IGM may have also become transparent to Ly$\alpha$
423: photons prior to overlap, resulting in windows of transmission inside
424: the Gunn-Peterson trough (one such region may have been seen\cite{WB1}
425: in SDSS J1148+5251). In the existing examples, the portions of the
426: universe probed by the lower end of the Gunn-Peterson trough are
427: located several hundred co-moving Mpc away from the background quasar,
428: and are therefore not correlated with the quasar host galaxy. The
429: distribution $dP/dz_{\rm SLT}$ is also independent of the redshift
430: distribution of the quasars. Moreover, lines-of-sight to these quasars
431: are not causally connected at $z\sim 6$ and may be considered
432: independent. }
433: \end{figure*}
434: \clearpage
435:
436: %FIGURE 2
437: \begin{figure*}[htbp]
438: \epsscale{1.}
439: \plotone{fig2.eps}
440: \end{figure*}
441:
442: %FIGURE 2
443: \begin{figure*}[htbp]
444: \caption{\label{fig2} Constraints on the scatter in the SBO redshift and
445: the characteristic size of isolated bubbles at the final overlap stage,
446: $R_{\rm SBO}$ (see Fig. 1). The characteristic size of \HII regions grows
447: with time. The SBO is observed for the bubble scale at which the light
448: crossing time (blue line) first becomes smaller than the cosmic scatter in
449: bubble formation times (red line). At $z\sim6$, the implied scale $R_{\rm
450: SBO}\sim60$ co-moving Mpc (or $\sim 8.6$ physical Mpc), corresponds to a
451: characteristic angular radius of $\theta_{\rm SLT}\sim 0.4$ degrees on the
452: sky. After bubble overlap, the ionizing intensity grows to a level at
453: which the IGM becomes transparent to Ly$\alpha$ photons. The collapsed
454: fraction required for \lya transmission within a region of a certain size
455: will be larger than required for its ionization. However, the scatter in
456: equation~(\ref{scatter}) is not sensitive to the collapsed fraction, and so
457: may be used for both the SBO and SLT. The scatter in the SLT is smaller
458: than the cosmic scatter in the structure formation time on the scale of the
459: mean-free-path for ionizing photons. This mean-free-path must be longer
460: than $R_{\rm SBO}\sim60$Mpc, an inference which is supported by analysis of
461: the Ly$\alpha$ forest at $z\sim4$ where the mean-free-path is
462: estimated\cite{ME1} to be $\sim 120$ co-moving Mpc at the Lyman limit (and
463: longer at higher frequencies). If it is dominated by cosmic variance, then
464: the scatter in the SLT redshift provides a lower limit to the SBO
465: scatter. The three known quasars at $z>6.1$ have \lya transmission
466: redshifts of\cite{WB1}$^,$\cite{f2} $z_{\rm SLT}=5.9$, 5.95 and 5.98,
467: implying that the scatter in the SBO must be $\ga0.05$ (this scatter may
468: become better known from follow-up spectroscopy of Gamma Ray Burst
469: afterglows at $z>6$ that might be discovered by the {\it SWIFT}
470: satellite\cite{BL2}$^,$\cite{BL3}).
471: The observed scatter in the SLT redshift is somewhat smaller than the
472: predicted SBO scatter, confirming the expectation that cosmic variance is
473: smaller at the SLT. The scatter in the SBO redshift must also be $\la0.25$
474: because the lines-of-sight to the two highest redshift quasars have a
475: redshift of \lya transparency at $z\sim6$, but a neutral fraction that is
476: known from the {\it proximity effect}\cite{WL1} to be substantial at
477: $z\ga6.2-6.3$. The excluded regions of scatter for the SBO are shown in
478: gray. Throughout this {\it Letter}, we adopt the latest values for the
479: cosmological parameters as inferred from the {\em Wilkinson Microwave
480: Anisotropy Probe} data\cite{SB1}.}
481: \end{figure*}
482:
483: \end{document}
484:
485:
486:
487:
488:
489:
490:
491:
492: