1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3:
4: \newcommand{\myemail}{dr@caltech.edu}
5: %\usepackage{amssymb}
6:
7: %% -------Author macros----------------------
8: %% DO NOT USE IN MATH MODE!
9: \def\ts {\thinspace}
10: \def\kms {\ts km\ts s$^{-1}$}
11: \def\etal {{\rm et\ts al.}}
12: \def\msol {$M_{\odot}$}
13: \def\lsol {$L_{\odot}$}
14: \def\ci {C\ts {\scriptsize I}}
15: \def\hi {H\ts {\scriptsize I}}
16: \def\aco {{\rm CO}($J$=1$\to$0)}
17: \def\bco {{\rm CO}($J$=2$\to$1)}
18: \def\cco {{\rm CO}($J$=3$\to$2)}
19: \def\dco {{\rm CO}($J$=4$\to$3)}
20: \def\eco {{\rm CO}($J$=5$\to$4)}
21: \def\fco {{\rm CO}($J$=6$\to$5)}
22: \def\gco {{\rm CO}($J$=7$\to$6)}
23: \def\hco {{\rm CO}($J$=8$\to$7)}
24: \def\ico {{\rm CO}($J$=9$\to$8)}
25: \def\jco {{\rm CO}($J$=10$\to$9)}
26: \def\kco {{\rm CO}($J$=11$\to$10)}
27: \def\cho {{\rm H$_2$O}(3$_{13}$$\to$2$_{20}$)}
28: \def\ahcn {{\rm HCN}($J$=1$\to$0)}
29: \def\bhcn {{\rm HCN}($J$=2$\to$1)}
30: \def\chcn {{\rm HCN}($J$=3$\to$2)}
31: \def\ahco {{\rm HCO$^+$}($J$=1$\to$0)}
32: \def\bri {BRI\,1335--0417}
33: %%-------------------------------------------
34:
35: \slugcomment{draft version \today, accepted for publication in the Astrophysical Journal Letters}
36:
37: \shorttitle{High--Resolution CO(2--1) Imaging of BRI\,1335--0417}
38: \shortauthors{Riechers et al.}
39:
40: \begin{document}
41:
42: \title{
43: Formation of a Quasar Host Galaxy through a Wet Merger \\ 1.4 Billion
44: Years after the Big Bang}
45: %Resolving the $z=4.4$ Quasar Host Galaxy of BRI\,1335-0417: \\
46: %Intimate Interaction or Massive Molecular Disk?}
47:
48: \author{Dominik A. Riechers\altaffilmark{1,2,5}, Fabian Walter\altaffilmark{1},
49: Christopher L. Carilli\altaffilmark{3}, Frank Bertoldi\altaffilmark{4},
50: and Emmanuel Momjian\altaffilmark{3}}
51:
52:
53: \altaffiltext{1}{Max-Planck-Institut f\"ur Astronomie, K\"onigstuhl 17,
54: Heidelberg, D-69117, Germany}
55:
56: \altaffiltext{2}{Astronomy Department, California Institute of
57: Technology, MC 105-24, 1200 East California Boulevard, Pasadena, CA
58: 91125; dr@caltech.edu}
59:
60: \altaffiltext{3}{National Radio Astronomy Observatory, PO Box O, Socorro, NM 87801}
61: % USA}
62:
63: \altaffiltext{4}{Argelander-Institut f\"ur Astronomie, Universit\"at Bonn, Auf dem
64: H\"ugel 71, Bonn, D-53121, Germany}
65:
66: \altaffiltext{5}{Hubble Fellow}
67:
68: %\email{dr@caltech.edu}
69:
70:
71: \begin{abstract}
72: We present high-resolution Very Large Array imaging of the molecular
73: gas in the host galaxy of the high redshift quasar \bri\
74: ($z=4.41$). Our \bco\ observations have a linear resolution of
75: 0.15$''$ (1.0\,kpc) and resolve the molecular gas emission both
76: spatially and in velocity. The molecular gas in \bri\ is extended on
77: scales of 5\,kpc, and shows a complex structure.
78: % From the CO
79: % luminosity, we find that the molecular gas reservoir has a total
80: % mass of 9.2 $\times$ 10$^{10}$\,\msol.
81: At least three distinct components encompassing about two thirds of
82: the total molecular mass of 9.2$\times$10$^{10}$\,\msol\ are
83: identified in velocity space, which are embedded in a structure that
84: harbors about one third of the total molecular mass in the
85: system. The brightest \bco\ line emission region has a peak
86: brightness temperature of 61$\pm$9\,K within 1\,kpc diameter, which
87: is comparable to the kinetic gas temperature as predicted from the
88: CO line excitation. This is also comparable to the gas temperatures
89: found in the central regions of nearby ultra-luminous infrared
90: galaxies, which are however much more compact than 1\,kpc. The
91: spatial and velocity structure of the molecular reservoir in \bri\
92: is inconsistent with a simple gravitationally bound disk, but
93: resembles a merging system. Our observations are consistent with a
94: major, gas-rich (`wet') merger that both feeds an accreting
95: supermassive black hole (causing the bright quasar activity), and
96: fuels a massive starburst that builds up the stellar bulge in this
97: galaxy. Our study of this $z$$>$4 quasar host galaxy may thus be
98: the most direct observational evidence that `wet' mergers at high
99: redshift are related to AGN activity.
100: \end{abstract}
101:
102: \keywords{galaxies: active --- galaxies: starburst --- galaxies: formation
103: --- galaxies: high-redshift --- cosmology: observations
104: --- radio lines: galaxies}
105:
106: %\citet{} xxx et al. (xxxx)
107: %\citep{} (xxx et al xxxx)
108: %\citep[bla][]{} (bla xxx et al xxxx)
109:
110: \section{Introduction}
111:
112: Great progress has been made in recent years both observationally and
113: theoretically to further our understanding of galaxy formation and
114: evolution from the early epochs of galaxy formation to the present-day
115: universe. One basic prediction of cosmological simulations is that
116: during the early epoch of hierarchical galaxy formation, some of the
117: most massive galaxies are already formed in major merger events (e.g.,
118: Springel \etal\ \citeyear{spr05}). These mergers are believed to
119: commonly trigger both AGN and starburst activity in such early
120: systems, which is regulated via AGN feedback (e.g., Hopkins \etal\
121: \citeyear{hop05}). Such feedback may be responsible for the
122: present-day ``$M_{\rm BH}$--$\sigma_v$'' relation between black hole
123: mass and bulge velocity dispersion (Ferrarese \& Merritt
124: \citeyear{fer00}; Gebhardt \etal\ \citeyear{geb00}).
125: % However, this has never been observed directly at high redshift.
126: % to date.
127:
128: Studies of molecular gas (the requisite material to fuel star
129: formation) in young galaxies are an essential ingredient to understand
130: their physical properties in more detail.
131: % (e.g., the nature of the observed relation
132: %%Schmidt-Kennicutt law
133: %between total gas mass and star formation rate;
134: %%Schmidt \citeyear{sch59}, \citeyear{sch63};
135: %e.g., Kennicutt \citeyear{ken98a},
136: %\citeyear{ken98b}; Kennicutt et al.\ \citeyear{ken07}; Krumholz \&
137: %Thompson \citeyear{kt07}).
138: Molecular gas (typically CO) has been detected in $\sim$50 galaxies at
139: $z$$>$1 to date, revealing large molecular reservoirs of
140: $>$10$^{10}$\,\msol\ in most cases (see Solomon \& Vanden Bout
141: \citeyear{sv05} for a review). However, these studies rely almost
142: exclusively on the integrated properties of the line emission, as the
143: molecular reservoirs in these distant galaxies are difficult to
144: resolve. To date, only the $z$=4.69 and $z$=6.42 quasars
145: BR\,1202--0725 and SDSS\,J1148+5251, two of the most distant gas-rich
146: galaxies could be resolved in molecular gas emission without the aid
147: of gravitational magnification (Omont \etal\ \citeyear{omo96}; Carilli
148: \etal\ \citeyear{car02}; Walter \etal\ \citeyear{wal04}).
149: %, and are due to their great
150: %relevance a major design goal for future (sub)mm/radio facilities like
151: %the Atacama Large Millimeter Array (ALMA). The constant improvement of
152: %the high-frequency capabilities of the Very Large Array (VLA) now
153: %enable us, for the first time, to sneak peek into the ALMA era, i.e.,
154: %to obtain high angular resolution (0.15$''$) observations of molecular
155: %gas in $z$$>$4 galaxies that also offer velocity resolution to study
156: %the dynamical structure of these galaxies.
157:
158: In this letter, we report on high angular resolution (0.15$''$;
159: 1.0\,kpc) Very Large Array (VLA)\footnote{The Very Large Array is a
160: facility of the National Radio Astronomy Observatory, operated by
161: Associated Universities, Inc., under a cooperative agreement with the
162: National Science Foundation.} observations of molecular gas
163: %(the fuel for star formation)
164: in the host galaxy of BRI\,1335--0417, a dust-rich, optically
165: identified quasar at a redshift of 4.41, corresponding to only 1.4 Gyr
166: after the Big Bang.
167: %($\sim$10\% of the current age of the Universe).
168: %Note that the emission is likely not enhanced or distorted by
169: %gravitational magnification, as high-resolution
170: Optical imaging with the Hubble Space Telescope reveals a single point
171: source without any evidence for
172: %of distortion or other signs
173: gravitational lensing (Storrie-Lombardi \etal\ \citeyear{sl96}). We
174: use a concordance, flat $\Lambda$CDM cosmology throughout, with
175: $H_0$=71\,\kms\,Mpc$^{-1}$, $\Omega_{\rm M}$=0.27, and
176: $\Omega_{\Lambda}$=0.73 (Spergel \etal\
177: %\citeyear{spe03},
178: \citeyear{spe07}).
179:
180:
181: \section{Observations}
182:
183: \begin{figure}
184: \epsscale{1.15}
185: \vspace{-5mm}
186:
187: \plotone{f1.eps}
188: %% BLC 464 482
189: %% TRC 537 555
190: %% PIXRANG -60e-6 350e-6
191: \vspace{-10mm}
192:
193: \caption{VLA map of the \bco\ emission toward BRI1335--0417 (integrated over the central 43.75\,MHz, or 308\,km\,s$^{-1}$), imaged using the combined B and C array dataset and natural weighting.
194: %The circle in the bottom left corner
195: %of each map
196: %represents the FWHM of the restoring CLEAN beam.
197: %The crosses indicate the two \bco\ positions
198: %identified by Carilli et al.\ (\citeyear{car02}).
199: %This map is integrated over the central 43.75\,MHz
200: %(308\,km\,s$^{-1}$) of the \bco\ emission.
201: %{\bf a}: Contours are shown at (-3, -2, 2, 3, 4, 5, 6, 7, 8, 9)$\times\sigma$ (1$\sigma = 55\,\mu$Jy beam$^{-1}$).
202: %This map is tapered to a resolution of 0.32\,$''$$\times$0.30\,$''$; natural weighting (robust 5) was used for imaging.
203: %{\bf b}:
204: Contours are shown at (-3, -2, 2, 3, 4, 5, 6, 7)$\times\sigma$
205: (1$\sigma = 50\,\mu$Jy beam$^{-1}$).
206: %Natural weighting was used.
207: The beam size (0.23\,$''$$\times$0.18\,$''$)
208: %; natural weighting)
209: is shown in the bottom left corner.
210: %This map is shown at the full resolution of 0.23\,$''$$\times$0.18\,$''$; natural weighting (robust 5) was used for imaging.
211: %{\bf c}: Contours are shown at (-2.8, -1.4, 1.4, 2.8, 4)$\times\sigma$ (1$\sigma = 64\,\mu$Jy beam$^{-1}$).
212: %This map is weighted towards a resolution of 0.17\,$''$$\times$0.15\,$''$; robust 0 was used for imaging.
213: %The source is clearly resolved in all maps and breaks up into two clumps at higher resolution.
214: \label{f1}}
215: %\vspace{-5mm}
216: %
217: \end{figure}
218:
219: We observed the \bco\ transition ($\nu_{\rm rest} = 230.53799\,$GHz)
220: towards \bri\ using the VLA in B configuration on 2005 April 11, 13,
221: and 22, and in C configuration on 2005 September 16. The total on-sky
222: integration time amounts to 29\,hr. At $z=4.4074$, the line is
223: redshifted to 42.6338\,GHz (7.03\,mm). Observations were performed in
224: fast-switching mode (e.g., Carilli \& Holdaway \citeyear{ch99}) using
225: the nearby source 13515-01513 for secondary amplitude and phase
226: calibration. Observations were carried out under very good weather
227: conditions with 25 antennas. The phase stability in all
228: %B configuration
229: runs was excellent (typically $<$15$^\circ$ for the longest
230: baselines).
231: %, and good in the C configuration run.
232: The phase coherence was checked by imaging the calibrator source
233: 13569+02144 with the same calibration cycle as that used for the
234: target source. For primary flux calibration, 3C\,286 was observed
235: during each run. Given the restrictions of the VLA correlator, two
236: 25\,MHz intermediate frequencies (IFs) with seven 3.125\,MHz channels
237: each were observed simultaneously centered at the \bco\ line
238: frequency, leading to an effective bandwidth of 43.75\,MHz
239: (corresponding to 308\kms\ at 42.6\,GHz). This encompasses a major
240: fraction of the CO line width as measured in the \eco\ transition
241: (420$\pm$60\,\kms\ FWHM, Guilloteau \etal\ \citeyear{gui97}), but does
242: not cover the line wings and the continuum. Earlier observations set a
243: 2\,$\sigma$ limit of 240$\mu$Jy on the continuum emission (Carilli
244: \etal\ \citeyear{car99}), in agreement with an expected flux of
245: $\sim$34\,$\mu$Jy as derived from the continuum spectral energy
246: distribution of this source.
247:
248: For data reduction and analysis, the AIPS
249: %\footnote{www.aoc.nrao.edu/aips/}
250: package was used.
251: %All data were mapped using the CLEAN algorithm.
252: Two data sets were created for the final analysis. The first dataset
253: includes both the B- and C-array data, and is imaged using natural
254: weighting. A velocity-integrated \bco\ map of this dataset is shown
255: in Fig.~\ref{f1}. The synthesized clean beam has a size of
256: 0.23\,$''$$\times$0.18\,$''$ (1.6$\times$1.2\,kpc). The final rms
257: over the full bandwidth of 43.75\,MHz (308\,\kms) is 50\,$\mu$Jy
258: beam$^{-1}$. To boost the resolution, the second dataset includes B
259: array data only, and are imaged using robust 0 (i.e., intermediate
260: between natural and uniform) weighting (Fig.~\ref{f2}), achieving a
261: resolution of 0.16\,$''$$\times$0.14\,$''$ (1.1$\times$0.95\,kpc).
262: %\footnote{Note that resolutions of order 1\,kpc
263: % are typically achieved with single-dish telescopes in studies of
264: % nearby galaxies in the \aco\ transition.}.
265: This results in an rms of 58\,$\mu$Jy\,beam$^{-1}$ per channel. In
266: Fig.~\ref{f3}, seven velocity channel maps (6.25\,MHz, or 44\,\kms\
267: each) of the \bco\ line based on the combined BC-array data are shown.
268: A Gaussian taper falling to 30\% level at a baseline length of
269: 800\,k$\lambda$ (longest baseline in the B array data:
270: $\sim$1.6M$\lambda$) was applied to the UV data, resulting in a
271: resolution of 0.32\,$''$$\times$0.30\,$''$ (2.2$\times$2.0\,kpc), and
272: %This results in
273: an rms of 150\,$\mu$Jy\,beam$^{-1}$ per channel. These channel maps
274: are combined into a three color overlay in Fig.~\ref{f4}.
275:
276: \begin{figure}
277: \epsscale{1.15}
278: \vspace{-5mm}
279:
280: \plotone{f2.eps}
281: %% BLC 441 482
282: %% TRC 529 571
283: %% PIXRANG 29e-6 385e-6
284: \vspace{-10mm}
285:
286: \caption{\bco\ emission line map of \bri\ at $\sim$1.0\,kpc resolution,
287: imaged using B array data only and robust 0 weighting.
288: % (zoomed-in version of Figure \ref{f1}).
289: Contours are shown at (--3, --2, 2, 3,
290: 4, 5, 6)$\times\sigma$ (1$\sigma = 58\,\mu$Jy beam$^{-1}$). The beam
291: size (0.16\,$''$$\times$0.14\,$''$) is shown in the bottom left
292: corner. The labels indicate the rest-frame peak brightness
293: temperatures $T_{\rm b}$ in the brightest
294: % individually resolved
295: molecular line emission regions.
296: \label{f2}}
297: %\vspace{-1mm}
298: %
299: \end{figure}
300:
301: \section{Results}
302:
303: \subsection{Morphology of the Molecular Gas Reservoir}
304:
305: \begin{figure*}
306: \epsscale{1.15}
307: \vspace{-10mm}
308:
309: \plotone{f3.ps}
310: %% PIXRANG -75e-6 850e-6
311: \vspace{-20mm}
312:
313: \caption{Channel maps of the \bco\ emission in \bri. The same region
314: is shown as in Figure \ref{f1}. One channel width is 6.25\,MHz, or
315: 44\,\kms\ (frequencies increase with channel number and are shown at
316: % 42621.300, 42630.675, 42640.050 and 42649.425
317: 42615.050, 42621.300, 42627.550, 42633.800, 42640.050, 42646.300,
318: and 42652.550\,MHz). Contours are shown at (--3, --2, 2, 3, 4, 5)$\times \sigma$ (1$\sigma = 150\,\mu$Jy\,beam$^{-1}$). The
319: beam size (tapered to 0.32\,$''$$\times$0.30\,$''$)
320: %; a Gaussian taper at 800\,k$\lambda$ was applied)
321: is shown in the bottom left corner; the crosses indicate
322: %the extent of the emission.
323: main peaks of the the northern (`N') and southern (`S') CO components in
324: Fig.~\ref{f2}.
325: %the same positions as in Fig.~1.
326: % The emission clearly moves from the northern (redwards) to the
327: % southern (bluewards) cross.
328: \label{f3}}
329: \end{figure*}
330:
331: In Figure \ref{f1}, the integrated \bco\ emission over the full
332: measured bandpass (308\,\kms ) is shown at a linear resolution of
333: $\sim$1.4\,kpc (0.2$''$). The emission is clearly spatially resolved
334: over many beams, and extended out to a scale of 5\,kpc ($\sim$0.7$''$,
335: deconvolved for the beam size). Two distinct peaks of the emission are
336: identified, both of which are extended, and embedded in a continuous,
337: morphologically complex molecular structure. The northern component
338: has a peak flux density of 216 $\pm$ 50\,$\mu$Jy beam$^{-1}$, and the
339: southern component has a peak flux density of 388 $\pm$ 50\,$\mu$Jy
340: beam$^{-1}$ (peaks labeled `N' and `S' in Fig.~\ref{f2}). In Figure
341: \ref{f2}, a zoomed-in version of this map at 1.0\,kpc ($\sim$0.15$''$)
342: linear resolution is shown.
343: %To boost the resolution, shorter C-array baselines were
344: %discarded from this dataset, and the image is weighted toward the
345: %longer baselines (robust 0, i.e., intermediate between natural and
346: %uniform weighting), which causes some of the more diffuse structure to
347: %be resolved out.
348: At this resolution, some of the more diffuse structure is resolved
349: out. Clearly, the southern component breaks up into multiple line
350: emission regions. Most notably, a resolved
351: %One of the
352: subcomponent extends from the main southern peak (`S') toward the
353: northern peak (`N').
354: %, consistent with the extension of the southern
355: %component in Figure \ref{f1}. A third, small peak of the southern
356: %component is found at 3$\sigma$ significance.
357: %In addition, it appears that the different spatial
358: %components are
359: % not aligned, but
360: %distributed in a
361: % more
362: %complex way.
363: % The molecular structure thus is not only extended over several
364: % kiloparsec, but also is rather disturbed.
365: In the higher resolution map, the northern component has a peak flux
366: density of 246 $\pm$ 58\,$\mu$Jy beam$^{-1}$, and the southern
367: component has a peak flux density of 375 $\pm$ 58\,$\mu$Jy
368: beam$^{-1}$. The connecting component has a peak flux density of 202
369: $\pm$ 58\,$\mu$Jy beam$^{-1}$. These three peaks alone encompass a
370: %minimum
371: molecular gas mass\footnote{As CO emission in currently known $z
372: \gtrsim 4$ quasar host galaxies appears to be thermalized typically
373: up to $J$=4$\to$3 (Riechers et al.\ \citeyear{rie06}; Wei\ss\ et al.\
374: \citeyear{wei07}), constant $T_{\rm b}$ is assumed between \bco\ and
375: \aco.
376: %Assuming constant $T_{\rm b}$ between \bco\ and \aco, and
377: The CO
378: %emission in $z \gtrsim 4$ quasar hosts such as \bri\
379: is more widespread than
380: %in the compact nuclear reservoirs
381: in nearby ultra-luminous infrared galaxies (ULIRGs), but shows similar
382: physical properties. We thus adopt a low ULIRG CO luminosity to H$_2$
383: mass conversion factor of
384: $\alpha$=0.8\,\msol\,(K\,\kms\,pc$^2$)$^{-1}$ (Downes \& Solomon
385: \citeyear{ds98}) rather than
386: $\alpha$=4--5\,\msol\,(K\,\kms\,pc$^2$)$^{-1}$ as in nearby spiral
387: galaxies (e.g., Scoville \& Sanders \citeyear{sco87a}; Solomon
388: \& Barrett \citeyear{sol91}). Such low $\alpha$ are also found for $z
389: \sim 2.5$ submm galaxies (Tacconi et al.\ \citeyear{tac08}).}
390: of $M({\rm H_2})=4.1 \times 10^{10}\,$\msol, corresponding to about
391: two thirds of the mass of $M({\rm H_2})=6.6 \times 10^{10}\,$\msol\
392: derived from the integrated emission seen in the naturally weighted
393: dataset ($I_{\rm CO(2-1)}$=0.43$\pm$0.02\,Jy\,\kms, in agreement with
394: previous results by Carilli \etal\ \citeyear{car02}).
395: %Correcting
396: Accounting for the flux
397: %missed
398: in the linewings that is not covered by our observations, the full
399: reservoir has an estimated mass of $M_{\rm tot}({\rm H_2})=9.2 \times
400: 10^{10}\,$\msol.
401:
402: At $z$=4.4074, the brightest peaks of the northern and southern
403: components in the highest resolution map corresponds to beam-averaged,
404: rest-frame brightness temperatures of $T_{\rm b}$=61 $\pm$ 9\,K and 40
405: $\pm$ 9\,K (see Figure \ref{f2}). Averaging over the area of the full
406: \bco\ reservoir gives $T_{\rm b}$=10.0 $\pm$ 0.5\,K, which implies
407: that part of the diffuse molecular structure may be rather cold.
408: %This sets a lower
409: %limit on the average temperature of the full spread-out molecular gas
410: %reservoir.
411: Note that the minimum excitation temperature of \bco\ is $T_{\rm
412: ex}$=16.6\,K, which however is only 1.8\,K above the cosmic microwave
413: background (CMB) temperature at $z$=4.41.
414:
415: \subsection{Dynamical Structure of the CO Distribution}
416:
417: In Figure \ref{f3}, the \bco\ emission is shown in seven 44\,\kms\
418: wide velocity channels, smoothed to a linear resolution of
419: $\sim$2\,kpc. The emission is clearly dynamically resolved and is
420: moving, to first order, from north to south between the red and blue
421: velocity channels (i.e., with increasing channel number). In addition,
422: the center channel (\#4) shows a bright, compact peak, which
423: corresponds to a narrow velocity component. This peak is weighted down
424: in the integrated line maps, and corresponds to the third small
425: 3$\sigma$ peak of the southern component in Figure \ref{f2} (labeled
426: with its $T_{\rm b}$ of 29 $\pm$ 9\,K). Figure
427: \ref{f4} shows a composite color map of the velocity channel maps and a position-velocity ($p-v$) diagram
428: %along the axis
429: %%with a position angle of 189$^\circ$, which
430: %connecting the brightest peaks of the northern and southern components
431: %in
432: %%in the highest resolution map
433: %Fig.~\ref{f2}
434: (left panel). The middle panel includes all velocity channels, while
435: the central channel (covering the range from -22\,\kms\ to +22\,\kms)
436: is excluded from the right panel. The emission is clearly moving from
437: redshifted to blueshifted emission in the right panel. However, the
438: emission in the (green) central channels does not
439: %mainly
440: follow the north-south extension, but also show an east-west
441: extension, indicating an additional, distinct velocity component in
442: the system. The overall peak of the emission where all components
443: overlap (seen as a `white spot' in the figure) corresponds to the
444: southern peak in the integrated line map. The velocity structure thus
445: is more complex than that of a simple, inclined, rotating disk (as
446: also seen in the $p-v$ slice).
447: %(which would be in agreement with the north--south extension),
448: %and also clearly disturbed. Moreover, the left panel clearly shows
449: %that the bright, narrow velocity component in the center channel is
450: %spatially offset from the southern peak of the integrated line map,
451: %and thus corresponds to an additional, distinct component in this
452: %system.
453: The central, compact, narrow velocity component
454: %has a peak flux
455: %density of 853 $\pm$ 150\,$\mu$Jy beam$^{-1}$, corresponding to a
456: %beam-averaged, rest-frame peak brightness temperature of $T_{\rm
457: %b}$=32 $\pm$ 6\,K. It
458: is likely
459: %that this components is
460: %may actually be
461: associated with the optical quasar. However, the relative astrometry
462: of the radio and optical observations (Storrie-Lombardi \etal\
463: \citeyear{sl96}) is unfortunately not accurate enough at present to
464: substantiate this conclusion.
465: %verify this assumption.
466:
467:
468: \section{Discussion}
469:
470: \begin{figure*}
471: %\epsscale{1.15}
472: %\vspace{3mm}
473: \plotone{f4.eps}
474: \vspace{1.5mm}
475:
476: \caption{RGB composite color map representation and position-velocity ($p-v$) diagram
477: of the \bco\ velocity channels shown in Fig.~\ref{f3}, with three
478: colors encoding the velocity range of the emission [red: redshifted
479: (channels 1-2 in Figure \ref{f3}), green: central (channels 3-5),
480: blue: blueshifted (channels 6-7)]. The velocity range covered
481: (relative to the line center) in \kms\ is indicated in the top left
482: corner, and the linear scale is indicated in the bottom left corner
483: of each panel. {\em Left}: $p-v$ diagram along a position axis of
484: 189$^\circ$, connecting the two brightest peaks of the integrated
485: line emission. {\em Middle}: All velocity channels (1-7). The
486: dashed line indicates the orientation of the $p-v$ slice. {\em
487: Right}: All except the central channel (channel 4).
488: %Image of the 4.885\,GHz radio continuum of BRI1335-0417 at 1.8$''$$\times$1.6$''$ resolution.
489: %Contours are (-3, -2, 2, 3, 4, 5, 6, 7, 8, 9) $\times \sigma$ (1$\sigma = 10\,\mu$Jy beam$^{-1}$).
490: %The source appears marginally resolved.
491: \label{f4}}
492: \end{figure*}
493:
494: We present maps of the molecular gas distribution in a quasar host
495: galaxy at $z$=4.4 at 1.0\,kpc spatial resolution. The VLA data show
496: that the molecular gas reservoir in this galaxy is not only
497: distributed over a scale of $\sim$5\,kpc, but also structured in
498: velocity space. This is the first time that the molecular gas in a
499: quasar host galaxy at such a high redshift (or, indeed, at any
500: redshift greater than $\sim$0) has ever been resolved both spatially
501: and dynamically over multiple beams (except for the lensed system
502: PSS\,J2322+1944 at $z$=4.12; Riechers et al.\ \citeyear{rie08}). The
503: emission in \bri\ is resolved into at least three distinct components
504: harboring at least 1--2 $\times$ 10$^{10}$\,\msol\ of molecular gas
505: each, and at least another 2 $\times$ 10$^{10}$\,\msol\ is found in
506: the more diffuse molecular medium in between these
507: concentrations. Each of the subcomponents hosts a few times the
508: molecular gas mass of nearby ULIRGs such as Arp\,220 (Downes \&
509: Solomon \citeyear{ds98}). The brightest peak corresponds to a 1\,kpc
510: diameter region with a gas surface density of $\Sigma_{\rm H_2}^{\rm
511: peak}$=2.3 $\times$ 10$^{10}$\,\msol\,kpc$^{-2}$, which is by a factor
512: of a few higher than in in the central regions of typical ULIRGs
513: (e.g., Wilson et al.\ \citeyear{wil08}).
514: %The high resolution and sensitivity of the present observations
515: %reveals more detailed structure than any comparable previous studies.
516: %Within those limitations,
517: The total molecular gas mass and distribution are reminiscent of those
518: in the $z$=4.69 and $z$=6.42 quasars BR\,1202--0725 and
519: SDSS\,J1148+5251 (Omont \etal\ \citeyear{omo96}; Carilli \etal\
520: \citeyear{car02}; Riechers \etal\ \citeyear{rie06};
521: Walter \etal\ \citeyear{wal04}), albeit concentrated
522: on almost an order of magnitude smaller scales than in BR\,1202--0725.
523: %and showing much more detailed structure than both.
524: %Also, the structure of the molecular reservoir
525: %Accounting for the temperature of the
526: %CMB, the
527: %brightest peak of the
528: The brightest peak of the \bco\ emission traces a region
529: %of 500\,pc radius
530: of 1\,kpc diameter where the gas has an average temperature of at
531: least 61\,K (depending on optical depth and beam dilution), which is
532: at the high end of but comparable to temperatures in the central
533: regions of nearby ULIRGs (e.g., Downes \& Solomon \citeyear{ds98}).
534: However, in ULIRGs, regions with molecular gas temperatures of
535: 30--60\,K are much more compact than 1\,kpc.
536: % in ULIRGs.
537: %Intriguingly, the 30--60\,K warm gas in \bri\
538: %is distributed on kpc scales, while gas of such high temperatures is
539: %much more concentrated in ULIRGs.
540: The CO excitation ladder in \bri\ is not constrained well at
541: present. However, the excitation so far is consistent with being
542: similar to the $z$=4.69 quasar host galaxy of BR\,1202--0725, for
543: which models of collisional excitation predict a kinetic gas
544: temperature of $T_{\rm kin}$=60\,K and a median gas density of $n({\rm
545: H_2}) = 10^{4.1}\,$cm$^{-3}$ (Riechers et al.\
546: \citeyear{rie06}). If indeed $T_{\rm kin} \simeq T_{\rm b}^{\rm
547: peak}$, this would indicate that the brightest CO peak is close to
548: being resolved. Also, it appears that the gas temperature and density
549: are on the high end of but comparable to ULIRGs, while the total gas
550: mass is by about an order of magnitude higher.
551: % and the warm gas is distributed on larger scales.
552: % However, since the \eco\ line is highly excited in this source
553: % (Guilloteau \etal\ \citeyear{gui97}), a fraction of the gas may be
554: % considerably warmer.
555:
556: Based on the FIR luminosity and SED shape of the source ($L_{\rm
557: FIR}$=3.1$\times$10$^{13}$\,\lsol; Benford et al.\ \citeyear{ben99}),
558: a star formation rate (SFR) of 4650\,\msol\,yr$^{-1}$ and a total dust
559: mass of 2$\times$10$^9$\,\msol\ can be derived.\footnote{Assuming
560: SFR=1.5$\times$10$^{-10}\,L_{\rm FIR}$(\msol\,yr$^{-1}$/\lsol)
561: (Kennicutt \citeyear{ken98a}).}
562: %(assuming standard conversions; Kennicutt
563: %\citeyear{ken98a},\citeyear{ken98b}).
564: This does not account for possible heating of the
565: %2$\times$10$^9$\,\msol\ of
566: dust by the AGN. However, if the dust in this source has an extension
567: and non-uniform distribution similar to that of the molecular gas,
568: local heating (rather than heating of a central source) is likely to
569: be responsible for most of the dust emission. This is consistent with
570: the finding that the (rest-frame) 4.0\,cm radio continuum emission in
571: this source dominantly originates from extended, kpc-scale regions and
572: shows a peak $T_{\rm b}$ of only few times 10$^5$\,K (Momjian et al.\
573: \citeyear{mom07}). Such $T_{\rm b}$ are typical for relativistic
574: electrons from supernova remnants and the interstellar medium in
575: nuclear starbursts, but by at least 2 orders of magnitude lower than
576: found in AGN-powered environments. This finding is also reflected in
577: the fact that \bri\ follows the radio-FIR correlation for star-forming
578: galaxies (Carilli et al.\ \citeyear{car99}). The FIR-derived SFR is
579: by about an order of magnitude higher than in local ULIRGs; however,
580: the ratio of SFR and total gas mass is consistent with the scaling
581: relation for ULIRGs (Solomon et al.\ \citeyear{sol97}) and other
582: high-$z$ galaxies (Solomon \& Vanden Bout \citeyear{sv05}; Riechers et
583: al.\ \citeyear{rie06}).
584: % Schmidt-Kennicutt law.
585: %The surface density of
586: %the molecular gas that is available on kpc scales in \bri\ is somewhat
587: %higher than in the centers of nearby ULIRGs, where
588: Moreover, in the centers of nearby ULIRGs, $\Sigma_{\rm H_2}^{\rm peak}$
589: %the peak H$_2$ surface density
590: correlates with the SFR (Wilson et al.\ \citeyear{wil08}), which
591: % This correlation
592: suggests that an increased availability of fuel for star formation at
593: a certain density leads to an increased SFR. Given its high
594: $\Sigma_{\rm H_2}^{\rm peak}$ on kpc scales, this also motivates the
595: high derived SFR for \bri. Assuming that the gas is converted into
596: stars at an efficiency of 5--10\% as in giant molecular cloud cores
597: (e.g., Myers et al.\
598: \citeyear{mye86}; Scoville et al.\ \citeyear{sco87}), the SFR
599: corresponds to a gas depletion timescale of 2--4 $\times$ 10$^8$\,yr.
600: %This means that the starburst
601: %can only be maintained for a relatively short time, unless gas is
602: %resupplied on timescales of $\sim$10$^7$\,yr.
603:
604: If the molecular gas in this system was gravitationally bound, the CO
605: linewidth (see Sect.~2) and distribution would predict a dynamical
606: mass of $M_{\rm dyn}$=1.0 $\times$ 10$^{11}\,\sin^{-2}\,i$\,\msol,
607: which could account for both the total molecular gas mass and the mass
608: of the black hole of $M_{\rm BH}$=6 $\times$ 10$^9$\,\msol\ (Shields
609: \etal\
610: \citeyear{shi06}), but not for a substantial fraction of a
611: $\sim$4$\times$10$^{12}$\,\msol\ stellar bulge as predicted if the
612: local $M_{\rm BH}$--$\sigma_v$ relation were to hold (Ferrarese \&
613: Merritt \citeyear{fer00}; Gebhardt \etal\
614: \citeyear{geb00}).\footnote{In principle, assuming an extreme inclination
615: toward the line-of-sight would significantly increase $M_{\rm dyn}$;
616: %toward the
617: %values required by the local $M_{\rm BH}$--$\sigma_v$ relation;
618: however, the implied large CO linewidths suggest that such a dynamical
619: structure would not be virialized (Riechers et al.\
620: \citeyear{rie08}).}
621: %; unless the inclination toward the line-of-sight is very high, which
622: %; would imply
623: %very
624: %excessively large CO linewidths; see also discussion by Riechers et
625: %al.\ \citeyear{rie08}).
626:
627: However, the overall structure of \bri\ looks rather disturbed, both
628: spatially and in its velocity structure. While the general north-south
629: extension of the source may be in agreement with
630: % following a systemic trend in velocity,
631: rotating structure, this is not the case for the central part of the
632: emission, in particular the compact peak seen in the central channel.
633: This, and the structure seen in the high-resolution \bco\ map (Figure
634: \ref{f2}), are more reminiscent of the disturbed gas reservoirs in
635: major mergers, such as seen in the nearby Antennae (NGC\,4038/39,
636: e.g., Wilson \etal\ \citeyear{wil00}). If \bri\ was an interacting
637: system, the northern component may be in the process of merging with
638: the southern component, which likely hosts the luminous quasar. The
639: connecting component then may be the part where the two galaxies
640: overlap and merge. Note that this region alone would be massive enough
641: to host more than 10 of the largest molecular complexes found in the
642: overlap region of the Antennae (Wilson \etal\ \citeyear{wil00}).
643:
644: Cosmological simulations predict that the merger rates at $z$=4.4 are
645: substantially higher than at $z$=0. Such scenarios imply that during
646: the early epoch of hierarchical galaxy formation, some of the most
647: massive galaxies form in major merger events (e.g., Springel \etal\
648: \citeyear{spr05}). Such major, `wet' mergers are believed to commonly
649: trigger both AGN and starburst activity, and lead to high excitation
650: of the molecular gas during both the hierarchical buildup of the host
651: galaxy and the quasar phase (e.g., Narayanan \etal\ \citeyear{nar07}).
652: In simulations, such objects often show multiple CO emisson peaks,
653: arising from molecular gas concentrations that have not yet fully
654: coalesced.
655:
656: % It cannot be definitely excluded based on our high resolution VLA
657: % data that the observed structure in \bri\ is not due to interaction.
658: % However, in a cosmological context, it is highly likely
659: We conclude that the observed properties of \bri\ (AGN and extreme
660: starburst activity, high CO excitation, disturbed morphology over
661: 5\,kpc scales) are connected to the ongoing buildup of the quasar host
662: galaxy.
663: % ``caught in the act'', rather than, e.g., a huge rotating molecular
664: % disk with strong outflows.
665: Such a signpost of early galaxy assembly then could be considered
666: direct observational proof of the scenarios proposed by cosmological
667: simulations, and enable us to directly investigate the connection
668: between quasar activity and high-mass merger events at early cosmic
669: times.
670:
671: % Such observations have only now become possible with the improved
672: % VLA, and are a vital foundation for molecular gas observations with
673: % the Atacama Large submillimeter/Millimeter Array (ALMA) in the
674: % coming decade.
675:
676: \acknowledgments
677: %The National Radio Astronomy Observatory is operated by Associated
678: %Universities Inc., under cooperative agreement with the National
679: %Science Foundation.
680: We thank the anonymous referee for valuable comments that helped to
681: improve the manuscript. DR acknowledges support from from NASA through
682: Hubble Fellowship grant HST-HF-01212.01-A awarded by the Space
683: Telescope Science Institute, which is operated by the Association of
684: Universities for Research in Astronomy, Inc., for NASA, under contract
685: NAS 5-26555, and from the Deutsche Forschungsgemeinschaft (DFG)
686: Priority Program 1177. CC acknowledges support from the
687: Max-Planck-Gesellschaft and the Alexander von Humboldt-Stiftung
688: through the Max-Planck-Forschungspreis 2005.
689:
690: %\clearpage
691:
692: \begin{thebibliography}{}
693:
694: % \bibitem[Bertoldi et al.(2003)]{ber03} Bertoldi, F., Cox, P.,
695: % Neri, R., \etal\ 2003, A\&A, 409, L47
696: \bibitem[Benford et al.(1999)]{ben99} Benford, D.~J., et al.\ 1999,
697: ApJ, 518, L65
698: \bibitem[Carilli et al.(1999)]{car99} Carilli, C.~L., Menten, K.~M.,
699: \& Yun, M.~S.\ 1999, ApJ, 521, L25
700: \bibitem[Carilli \& Holdaway(1999)]{ch99} Carilli, C.~L., \& Holdaway,
701: M.~A.\ 1999, Radio Science, 34, 817
702: \bibitem[Carilli et al.(2002)]{car02} Carilli, C.~L., Kohno, K.,
703: Kawabe, R., Ohta, K., \etal\ 2002, AJ, 123, 1838
704: \bibitem[Downes \& Solomon(1998)]{ds98} Downes, D., \& Solomon, P.~M.\
705: 1998, ApJ, 507, 615
706: \bibitem[Ferrarese \& Merritt(2000)]{fer00} Ferrarese, L., \&
707: Merritt, D.\ 2000, ApJ, 539, L9
708: \bibitem[Gebhardt et al.(2000)]{geb00} Gebhardt, K., Bender, R.,
709: Bower, G., \etal\ 2000, ApJ, 539, L13
710: \bibitem[Guilloteau et al.(1997)]{gui97} Guilloteau, S., Omont, A.,
711: McMahon, R.~G., Cox, P., \& Petitjean, P.\ 1997, A\&A, 328, L1
712: \bibitem[Hopkins et al.(2005)]{hop05} Hopkins, P.~F., et al.\ 2005,
713: ApJ, 625, L71
714: %Hernquist, L., Martini, P., Cox, T.~J., Robertson, B., Di Matteo, T.,
715: %\& Springel, V.\ 2005, ApJ, 625, L71
716: \bibitem[Kennicutt(1998)]{ken98a} Kennicutt, R.~C.\ 1998, ApJ, 498,
717: 541
718: %\bibitem[Kennicutt(1998b)]{ken98b} Kennicutt, R.~C.\ 1998b, ARA\&A,
719: % 36, 189
720: %\bibitem[Kennicutt et al.(2007)]{ken07} Kennicutt, R.~C., et al.\ 2007,
721: % ApJ, 671, 333
722: %\bibitem[Krumholz \& Thompson(2007)]{kt07} Krumholz, M.~R., \& Thompson,
723: % T.~A.\ 2007, ApJ, 669, 289
724: \bibitem[Momjian et al.(2007)]{mom07} Momjian, E., Carilli, C.~L.,
725: Riechers, D.~A., \& Walter, F.\ 2007, AJ, 134, 694
726: \bibitem[Myers et al.(1986)]{mye86} Myers, P.~C., et al.\ 1986, ApJ,
727: 301, 398
728: \bibitem[Narayanan et al.(2008)]{nar07} Narayanan, D., \etal\ 2008,
729: ApJS, 174, 13
730: %in press (arXiv:0707.3141)
731: \bibitem[Omont et al.(1996)]{omo96} Omont, A., et al.\
732: %Petitjean, P., Guilloteau, S., McMahon, R.~G., Solomon, P.~M.;
733: %\& P\'econtal, E.\
734: 1996, Nature, 382, 428
735: \bibitem[Riechers et al.(2006)]{rie06} Riechers, D.~A., Walter, F.,
736: Carilli, C.~L., \etal\ 2006, ApJ, 650, 604
737: \bibitem[Riechers et al.(2008)]{rie08} Riechers, D.~A., et al.\
738: %Walter, F., Brewer, B.~J, Carilli, C.~L., Lewis, G.~F.,
739: %Bertoldi, F., \& Cox, P.\
740: 2008, ApJ, in press (arXiv:0806.4616)
741: %\bibitem[Schmidt(1959)]{sch59} Schmidt, M.\ 1959, ApJ, 129, 243
742: %\bibitem[Schmidt(1963)]{sch63} Schmidt, M.\ 1959, ApJ, 137, 758
743: \bibitem[Scoville \& Sanders(1987)]{sco87a} Scoville, N. Z., \& Sanders,
744: D. B.\ 1987, in Interstellar Processes
745: %, ed. D.\ Hollenbach \& H.\ Thronson
746: (Dordrecht:\ Reidel), 21
747: \bibitem[Scoville et al.(1987)]{sco87} Scoville, N. Z., et al.\
748: %Yun, M.~S., Sanders, D.~B., Clemens, D.~P., \& Waller, W. H.\
749: 1987, ApJS, 63, 821
750: \bibitem[Shields et al.(2006)]{shi06} Shields, G.~A., Menezes, K.~L.,
751: Massart, C.~A., \& Vanden Bout, P.\ 2006, ApJ, 641, 683
752: \bibitem[Solomon \& Barrett(1991)]{sol91} Solomon, P.~M., \& Barrett,
753: J.~W.\ 1991, `Dynamics of Galaxies and Their Molecular Cloud
754: Distributions', IAUS, 146, 235
755: \bibitem[Solomon et al.(1997)]{sol97} Solomon, P.~M., Downes, D.,
756: Radford, S.~J.~E., \& Barrett, J.~W.\ 1997, ApJ, 478, 144
757: \bibitem[Solomon \& Vanden Bout(2005)]{sv05} Solomon, P.~M., \& Vanden
758: Bout, P.~A.\ 2005, ARA\&A, 43, 677
759: %\bibitem[Spergel et al.(2003)]{spe03} Spergel, D.~N., Verde, L.,
760: % Peiris, H.~V., \etal\ 2003, ApJS, 148, 175
761: \bibitem[Spergel et al.(2007)]{spe07} Spergel, D.~N., Bean, R.,
762: Dor\'e, O., \etal\ 2007, ApJS, 170, 377
763: \bibitem[Springel et al.(2005)]{spr05} Springel, V., Di Matteo, T.,
764: Hernquist, L.\ 2005, MNRAS, 361, 776
765: \bibitem[Storrie-Lombardi et al.(1996)]{sl96} Storrie-Lombardi, L.~J.,
766: McMahon, R.~G., Irwin, M.~J., \& Hazard, C.\ 1996, ApJ, 468, 121
767: \bibitem[Taccconi et al.(2008)]{tac08} Tacconi, L.~J., et al.\ 2008,
768: ApJ, in press (arXiv:0801.3650)
769: \bibitem[Walter et al.(2004)]{wal04} Walter, F., Carilli, C.,
770: Bertoldi, F., \etal\ 2004, ApJ, 615, L17
771: \bibitem[Wei\ss\ et al.(2007)]{wei07} Wei\ss, A., et al.\ 2007,
772: `From Z-Machines to ALMA: (Sub)Millimeter Spectroscopy of Galaxies',
773: ASPC, 375, 25
774: \bibitem[Wilson et al.(2000)]{wil00} Wilson, C.~D., Scoville, N.,
775: Madden, S.~C., \& Charmandaris, V.\ 2000, ApJ, 542, 120
776: \bibitem[Wilson et al.(2008)]{wil08} Wilson, C.~D., et al.\ 2008, ApJS,
777: accepted (arXiv:0806.3002)
778: \end{thebibliography}
779:
780: \end{document}
781: