1: \def\Nat{{\em Nature}}

2: \def\PASJ{{\em Publ.~astr.~Soc.~Japan}}

3: \def\PASP{{\em Publ.~astr.~Soc.~Pacif.}}

4: \def\PhD{{\em PhD thesis}}

5: \def\Pre{Preprint}

6: \def\PS{{\em Phys.~Scr.}}

7: \def\RMP{{\em Rev.~Mod.~Phys.}}

8: \def\SSR{{\em Space Sci.~Rev.}}

9: \def\etal{{et al.\thinspace}}

10: \def\cf{{\em cf.\ }}

11: \def\eg{{\em e.g.\ }}

12: \def\etc{{\em etc.\ }}

13: \def\ie{{\em i.e.\ }}

14: \def\spose#1{\hbox to 0pt{#1\hss}}

15: \def\approxlt{\mathrel{\spose{\lower 3pt\hbox{$\sim$}}

16:         \raise 2.0pt\hbox{$<$}}}

17: \def\approxgt{\mathrel{\spose{\lower 3pt\hbox{$\sim$}}

18:         \raise 2.0pt\hbox{$>$}}}

19: \def\multcent#1{\halign {\centerline{##}\cr #1}}

20: \def\multleft#1{\hbox to size{\vbox {\halign {\lft{##}\cr #1}}\hfill}\par}

21: \def\multright#1{\hbox to size{\vbox {\halign {\rt{##}\cr #1}}\hfill}\par}

22: \def\Mdot{\hbox{$\dot M$}}

23: \def\degmark{^\circ}

24: \def\boxit#1{\vbox{\hrule\hbox{\vrule\kern3pt\vbox{\kern3pt

25:           #1 \kern3pt}\kern3pt\vrule}\hrule}}

26: \font\big=cmr10 scaled\magstep2

27: \font\bigbf=cmbx10 scaled\magstep2

28: \font\bigit=cmti10 scaled\magstep2

29: %       Simple unitsn4151.tex

30: \def\cm{{\rm\thinspace cm}}

31: \def\dyn{{\rm\thinspace dyn}}

32: \def\erg{{\rm\thinspace erg}}

33: \def\eV{{\rm\thinspace eV}}

34: \def\g{{\rm\thinspace g}}

35: \def\ga{{\rm\thinspace gauss}}

36: \def\K{{\rm\thinspace K}}

37: \def\keV{{\rm\thinspace keV}}

38: \def\km{{\rm\thinspace km}}

39: \def\kpc{{\rm\thinspace kpc}}

40: \def\Lsun{\hbox{$\rm\thinspace L_{\odot}$}}

41: \def\m{{\rm\thinspace m}}

42: \def\Mpc{{\rm\thinspace Mpc}}

43: \def\Msun{\hbox{$\rm\thinspace M_{\odot}$}}

44: \def\pc{{\rm\thinspace pc}}

45: \def\ph{{\rm\thinspace ph}}

46: \def\s{{\rm\thinspace s}}

47: \def\yr{{\rm\thinspace yr}}

48: \def\sr{{\rm\thinspace sr}}

49: \def\Hz{{\rm\thinspace Hz}}

50: \def\chisq{\hbox{$\chi^2$}}

51: \def\delchi{\hbox{$\Delta\chi$}}

52: %       Compound units

53: \def\cntspixps{cts pixel$^{-1}$ s$^{-1}$}

54: \def\cmps{\hbox{$\cm\s^{-1}\,$}}

55: \def\cmsq{\hbox{$\cm^2\,$}}

56: \def\pcmsq{\hbox{$\cm^{-2},$}}

57: \def\cmcu{\hbox{$\cm^3\,$}}

58: \def\pcmcu{\hbox{$\cm^{-3}\,$}}

59: \def\pcmcuK{\hbox{$\cm^{-3}\K\,$}}

60: \def\dynpcmsq{\hbox{$\dyn\cm^{-2}\,$}}

61: \def\ergcmcups{\hbox{$\erg\cm^3\ps\,$}}

62: \def\ergpcmps{\hbox{$\erg\cm^{-3}\s^{-1}\,$}}

63: \def\ergpcmsqps{\hbox{$\erg\cm^{-2}\s^{-1}\,$}}

64: \def\ergpcmsq{\hbox{$\erg\cm^{-2}\,$}}

65: \def\ergpcmsqpspA{\hbox{$\erg\cm^{-2}\s^{-1}$\AA$^{-1}\,$}}

66: \def\ergps{\hbox{$\erg\s^{-1}\,$}}

67: \def\gpcm{\hbox{$\g\cm^{-3}\,$}}

68: \def\gpcmps{\hbox{$\g\cm^{-3}\s^{-1}\,$}}

69: \def\gps{\hbox{$\g\s^{-1}\,$}}

70: \def\kmps{\hbox{$\km\s^{-1}\,$}}

71: \def\Lsunppc{\hbox{$\Lsun\pc^{-3}\,$}}

72: \def\Msunpc{\hbox{$\Msun\pc^{-3}\,$}}

73: \def\Msunpkpc{\hbox{$\Msun\kpc^{-1}\,$}}

74: \def\Msunppc{\hbox{$\Msun\pc^{-3}\,$}}

75: \def\Msunppcpyr{\hbox{$\Msun\pc^{-3}\yr^{-1}\,$}}

76: \def\Msunpyr{\hbox{$\Msun\yr^{-1}\,$}}

77: \def\pcm{\hbox{$\cm^{-3}\,$}}

78: \def\pcmsq{\hbox{$\cm^{-2}\,$}}

79: \def\pcmK{\hbox{$\cm^{-3}\K$}}

80: \def\phpcmsqps{\hbox{$\ph\cm^{-2}\s^{-1}\,$}}

81: \def\pHz{\hbox{$\Hz^{-1}\,$}}

82: \def\pmpc{\hbox{$\Mpc^{-1}\,$}}

83: \def\pmpccu{\hbox{$\Mpc^{-3}\,$}}

84: \def\ps{\hbox{$\s^{-1}\,$}}

85: \def\psqcm{\hbox{$\cm^{-2}\,$}}

86: \def\psr{\hbox{$\sr^{-1}\,$}}

87: \def\pyr{\hbox{$\yr^{-1}\,$}}

88: \def\kmpspMpc{\hbox{$\kmps\Mpc^{-1}$}}

89: \def\Msunpyrpkpc{\hbox{$\Msunpyr\kpc^{-1}$}}

90: \def\arcs{\hbox{\arcsec}}

91: %\def\approx{\hbox{\sim}}

92: 

93: % The first item in a LaTeX file must be a \documentstyle command to

94: % declare the overall style of the paper.  The \documentstyle lines

95: % that are relevant for the AASTeX macros are shown; one is uncommented out

96: % so that the file can be processed.

97: 

98: %\documentclass[preprint]{aastex}

99: \documentclass{aastex}

100: \usepackage{emulateapj5}

101: \usepackage{psfig}

102: \newcommand{\ha}{\hbox{H$\alpha$}}

103: \newcommand{\nii}{\hbox{[{\ion{N}{2}}]}}

104: \newcommand{\oiii}{\hbox{[{\ion{O}{3}}]}}

105: \newcommand{\nix}{$\cdot\cdot\cdot$}

106: 

107: %\documentstyle[emulateapj,epsfig]{article}

108: \input epsf           % somewhere early on in your TeX file

109: \makeatletter

110: \newenvironment{inlinetable}{%

111: \def\@captype{table}%

112: \noindent\begin{minipage}{0.999\linewidth}\begin{center}\footnotesize}

113: {\end{center}\end{minipage}\smallskip}

114: 

115: \newenvironment{inlinefigure}{%

116: \def\@captype{figure}%

117: \noindent\begin{minipage}{0.999\linewidth}\begin{center}}

118: {\end{center}\end{minipage}\smallskip}

119: \makeatother

120: 

121: % The eqsecnum style changes the way equations are numbered.  Normally,

122: % equations are just numbered sequentially through the entire paper.

123: % If eqsecnum appears in the \documentstyle command, equation numbers will

124: % be sequential through each section, and will be formatted "(sec-eqn)",

125: % where sec is the current section number and eqn is the number of the

126: % equation within that section.  The eqsecnum option can be used with

127: % any substyle.

128: 

129: %\documentstyle[11pt,eqsecnum,aaspp4]{article}

130: 

131: % Authors are permitted to use the fonts provided by the American Mathematical

132: % Society, if they are available to them on their local system.  These fonts

133: % are not part of the AASTeX macro package or the regular TeX distribution.

134: 

135: %\documentstyle[12pt,amssym,aasms4]{article}

136: 

137: % Here's some slug-line data.  The receipt and acceptance dates will be 

138: % filled in by the editorial staff with the appropriate dates.  Rules will 

139: % appear on the title page of the manuscript until these are uncommented 

140: % out by the editorial staff.

141: 

142: %\received{}

143: %\accepted{}

144: 

145: \slugcomment{Submitted to the Astrophysical Journal Letters} 

146: 

147: 

148: % Authors may supply running head information, if they wish to do so, although

149: % this may be modified by the editorial offices.  The left head contains a

150: % list of authors, usually three allowed---otherwise use \etal  The right

151: % head is a modified title of up to roughly 44 characters.  Running heads

152: % are not printed.

153: 

154: %\shorttitle{Large Scale structure in the X-ray sky}

155: %\shortauthors{Yang, Mushotzky, Barger et al.}

156: 

157: % This is the end of the "preamble".  Now we wish to start with the

158: % real material for the paper, which we indicate with \begin{document}.

159: % Following the \begin{document} command is the front matter for the

160: % paper, viz., the title, author and address data, the abstract, and

161: % any keywords or subject headings that are relevant.

162: 

163: \begin{document}

164: 

165: \title{Imaging Large Scale Structure in the X-ray Sky }

166: 

167: \author{Y. Yang\altaffilmark{1,2}, R. F. Mushotzky\altaffilmark{2}, A. J. Barger\altaffilmark{3,4,5}, L. L. Cowie\altaffilmark{5}, D. B. Sanders\altaffilmark{5,6}, A. T. Steffen\altaffilmark{3}}

168: 

169: \altaffiltext{1}{Department of Astronomy, University of Maryland, College Park 20742-2421}

170: \altaffiltext{2}{Laboratory for High Energy Astrophysics, Goddard Space Flight Center, Code 660, NASA, Greenbelt, MD, 20770}

171: \altaffiltext{3}{Department of Astronomy, University of Wisconsin at Madison, 5534 Sterling Hall, 475, Madison, WI 53760} 

172: \altaffiltext{4}{Department of Physics and Astronomy, University of Hawaii, 2505 Correa Road, Honolulu, HI 96822}

173: \altaffiltext{5}{Institute for Astronomy, University of Hawaii, 2680 Woodlawn drive, Honolulu, HI 96822} 

174: \altaffiltext{6}{Max-Planck-Institut fur Extraterrestrische Physik, D-85740, Garching, Germany}

175: 

176: 

177: 

178: \begin{abstract}

179: We present the first results from a wide solid angle, moderately 

180: deep {\it Chandra} survey of the Lockman Hole North-West region. 

181: Our 9 ACIS-I fields cover an effective solid angle of 0.4 deg$^{2}$ and 

182: reach a depth of  $3 \times 10^{-16}$~\ergpcmsqps in the 0.4--2 keV band and 

183: $3 \times 10^{-15}$~\ergpcmsqps in the 2--8 keV band. The best fit logN-logS

184: for the entire field, the largest contiguous {\it Chandra} field yet observed, 

185: matches well onto that of the  {\it Chandra} Deep Field North. We show that 

186: the full range of the `cosmic variance' previously seen in different 

187: {\it Chandra} fields is reproduced in this small region of the sky. 

188: Counts-in-cells analysis shows that the hard band sources are

189: more strongly correlated than the soft band sources.        

190: \end{abstract}

191: 

192: 

193: \keywords{cosmology: observations --- large-scale structure of the universe --- x-rays: diffuse background--- galaxies: nuclei}

194: 

195: 

196: \section{Introduction}

197: Recent {\it Chandra} and {\it XMM} observations have resolved over 85\% of the 

198: 2--8~keV X-ray background (XRB) into discrete sources, presumably active 

199: galaxtic nuclei (AGNs; Mushotzky et al. 2000; Brandt et al. 2001; 

200: Tozzi et al. 2001; Campana et al. 2001; Cowie et al. 2002; Giacconi et al. 

201: 2002; Hasinger et al. 2001).  However, the nature of the sources is unclear, 

202: with many of  the faint sources showing little or no optical activity (Barger 

203: et al. 2001, 2002; Hornschemeier et al. 2001; Rosati et al. 2002).  There is  

204: a large field-to-field variance in the source counts in the {\it Chandra} 

205: deep surveys in the  2--8~keV band (Cowie et al. 2002).

206: This cosmic variance is generally believed to arise from the 

207: underlying large-scale structure (LSS) that 

208: is traced, in some fashion, by the {\it Chandra} sources. 

209: Clustering of the XRB sources in redshift space has been seen in both 

210: {\it Chandra}  Deep Fields (Barger et al. 2002; Hasinger 2002),

211:  indicating that LSS does exist in the XRB source distribution.  

212: From a cosmological point of view, AGNs are expected to be highly 

213: biased tracers of cosmic structure formation at medium to large redshifts.

214: The XRB sources, which have $\sim10$ times the areal density of 

215: optically selected  AGNs, provide for the first time a sufficiently 

216: high density to be used as tracers of the LSS. 

217: 

218: 

219: \begin{figure*}[t]

220: \centerline{\psfig{figure=fig1a.ps,width=0.4\textwidth,angle=90}\psfig{figure=fig1b.ps,width=0.4\textwidth,angle=90} }

221: \centerline{\psfig{figure=fig1c.ps,width=0.4\textwidth,angle=90}\psfig{figure=fig1d.ps,width=0.4\textwidth,angle=90} }

222: \caption{The cumulative number counts in the 9 fields (a), 

223: (b) and the whole field (c), (d).

224: For each (a) and (b), the insert shows the number counts of each 

225: ISONW field at flux level of $2\times 10^{-15}$~\ergpcmsqps (soft) 

226: and  $9 \times 10^{-15}$~\ergpcmsqps (hard). The solid line is the 

227: logN-logS of CDF-N. Field numbers 10 and 11 represents 

228: the CDF-N and CDF-S value respectively. 

229: The highest counts in the 9 fields are about 2.3 (soft) and 3.3 (hard) 

230: times higher than that from CDF-S, which has the lowest normalization among 

231: all the deep surveys (Cowie et al. 2002). For the hard band, number counts of 8 of the 9 fields are $\sim 1\sigma$ about the mean, but ISONW5 is much lower. For each (c) and (d), 

232: cumulative number counts for the whole field (diamonds with solid 

233: lines representing the  Poisson error) is compared with that from CDF-N 

234: (solid histogram), CDF-S (dash-dotted histogram) and 

235: SSA13 (dashed line). The inserts shows the sensitive sky area vs flux. }

236: \end{figure*}

237: 

238: 

239: The clustering at various scales of X-ray selected AGNs has been pursued 

240: previously using observations from {\it HEAO-1} ( e.g. Barcons \& Fabian 1988;

241:  Mushotzky \& Jahoda 1992) and {\it ROSAT} (e.g. Soltan et al. 1997; 

242: Vihklinin \& Forman 1995). 

243: The {\it ROSAT} results based on deep pointings (Carrera et al. 1998) show a 

244: much smaller amplitude of correlation than predicted from optical samples, 

245: while an analysis of the {\it ROSAT} North Ecliptic Pole sample, which is 

246: shallower but covers a wider solid angle (Mullis 2002), comes to the 

247: opposite conclusion.  

248: It has been claimed that at a flux level below $1.5 \times 10^{-15} $ 

249: \ergpcmsqps (0.5--2 keV) the XRB exhibits no clustering (Soltan \& 

250: Hasinger 1994). This contradicts the observed cosmic variance in the

251: ultradeep surveys by {\it Chandra} and {\it XMM}. How to combine these 

252: seemingly disparate results is one of the new mysteries of the XRB. 

253:   

254: 

255: To allow a direct measurement of the auto-correlation functions (ACFs) of 

256: the XRB sources, and to optically identify these 

257: objects to obtain their redhshifts, one needs arcsecond spatial resolution, 

258: a wide contiguous field of view, and sufficient sensitivity. 

259: To achieve these goals we have performed a large solid angle, 

260: moderately deep {\it Chandra} survey of the Lockman Hole North-West region.

261: This survey currently has a contiguous sky coverage of 

262: $\sim 0.4$ deg$^{2}$ and is sensitive to X-ray flux levels of  

263: $3 \times 10^{-16}$ \ergpcmsqps (0.4--2 keV) 

264: and $3 \times 10^{-15}$ \ergpcmsqps (2--8 keV). 

265: Most of the XRB is resolved at these flux levels (Cowie et al. 2002).   

266:  

267: \section{Observations and Data Reduction}   

268: The survey covers the Lockman Hole North-West region centered 

269: at $\alpha=10^{h}34^{m}$, $\delta=57\arcdeg40\arcmin$ (J2000). 

270: The region has very low Galactic column density ($N_{H} \equiv 

271: 5.72\times 10^{19}$\pcmsq, Dickey \& Lockman 1990) and was covered 

272: by the deepest 170$\mu$m ISOPHOT field observed from {\it ISO} 

273: (hereafter ISONW).

274: The field  has also been observed at 850$\mu$m with the SCUBA camera on the 

275: James Clerk Maxwell Telescope and at cm wavelengths with the VLA, as well as 

276: at optical and near-IR wavelengths with the Subaru (using the unique 

277: Suprime-Cam instrument) and Keck telescopes. The ISONW region will be intensively observed 

278: with {\it SIRTF} as part of the LEGACY program. The multiwavelength analysis

279: of this field will be presented in subsequent papers. 

280: 

281: 

282: The observation consists of 9 ACIS-I pointings (labeled as ISONW1--ISONW9) 

283: separated from each other by $\sim 10\arcmin$ to allow close to uniform 

284: sky coverage.

285: %The 3.2~s standard frame times were used. 

286: ISONW1 has an exposure time of $\sim 70$~ks while the other 8 pointings have 

287: exposure times of $\sim 40$~ks. 

288: 

289: The data were reduced with CIAO 2.2.1 and CALDB 2.15. The CXC 

290: {\it Science Threads} were followed in data preparation. 

291: %All observations which were done before May 2, 2002 were

292: %corrected for known aspect offset problems. 

293: The resulting event lists were binned into 

294: 2 energy bands, the soft (0.4--2 keV) and the hard (2--8 keV). 

295: 

296: Point sources were detected for each pointing in both energy bands 

297: with {\it wavdetect} within the CIAO package. The wavelet scales 

298: of square root series $1,\sqrt 2, 2, 2 \sqrt 2, 4, 4 \sqrt 2,8$ and 

299: a false  detection threshold of $1 \times 10^{-7}$  were used.   

300: Spectrally weighted  monochromatic exposure maps were created, 

301: assuming a power law with photon index of 1.2 for

302: the hard band and 1.4 for the soft band (Mushotzky et al. 2000; Barger et al. 

303: 2001). Count rates were converted to flux assuming the above   

304: power law spectra with only Galactic absorption. The conversion factors  

305: are $4.74 \times 10^{-12}$ \ergpcmsq (soft) and 

306: $2.34 \times 10^{-11}$ \ergpcmsq (hard).

307: The degradation of the ACIS low energy quantum efficency during 

308: the flight was corrected using the measurements of the ACIS team. 

309: 

310: The catalogs for all observations were merged. Source properties for 

311: objects detected in more than one observation (due to the overlapping

312: of fields) were taken from the field in which the source had the smallest

313: off-axis angle. In the soft band, 431 sources were detected, and in the hard 

314: band, 278. The combined catalog contains 554 sources. 

315: 

316: 

317: \section{Analysis and Results}  

318: Since the effective area decreases and the point spread function (PSF)  

319: increases with off-axis angle, the sensitivity  is not 

320: uniform across the field. We quantified this with Monte-Carlo simulations. 

321: First we constructed background maps by removing the wavelet detected 

322: sources from the observed images and filled the holes with Poisson noise 

323: sampled from regions surrounding the sources. Sources with fluxes drawn from 

324: the LogN-LogS derived from the {\it Chandra} deep fields (Cowie et al. 2002;

325:  Garmire 2002) were generated and distributed uniformly within $8\arcmin$ from

326:  the aim point. PSF images of each source were made using {\it mkpsf}.  

327: The exposure maps described in the previous section were then applied to 

328: simulate the effect of vignetting. The simulated sources were added to 

329: the background maps to create simulated images. About 100 simulations were 

330: performed for each band and exposure (for details see  Yang et al. 2003). 

331: Above flux thresholds of 10 cts/exposure (soft) and 12 cts/exposure (hard), 

332: the detection is complete and the derived fluxes are consistent with the 

333: input values. We use the flux unit in cts/exposure 

334: because we found the {\it wavdetect} exposure-corrected count 

335: rates threshold in such unit only depends weakly on the exposure time in 

336: our observations due to the low backgrounds. We obtain the same results 

337: by assuming a uniform background and  the {\it wavdetect} detection thresholds 

338: described in Freeman et al.(2002).  Within $4.5\arcmin$ of the aim point, 

339: the detection is complete above 5 cts/exposure in the soft band and 

340: 7 cts/exposure in the hard band. 

341: 

342: Above flux thresholds of $1.2 \times 10^{-15}$ \ergpcmsqps (soft) and

343: $8 \times 10^{-15}$ \ergpcmsqps (hard), the source catalog

344: in the combined field within $8\arcmin$ from each aim point is complete. 

345: This defines a flux limited sample that contains 115 hard band and 298 

346: soft band sources. The solid angle covered by the complete sample field is 

347: 0.33 deg$^{2}$.  

348: 

349: %\subsection{Number Counts}

350: We constructed the cumulative number counts (LogN-LogS)

351: for each of 9 fields, as well as the whole merged catalog 

352: (Fig. 1). The Eddington bias found at these flux levels from 

353: the simulations is small and can be ignored.

354: 

355: %

356: % Fig. 2

357: %

358: \begin{inlinefigure}

359: \centerline{\psfig{figure=fig2.ps,width=0.9\textwidth,angle=0}}

360: %\vspace{12pt}

361: \caption{

362: The density map of the flux limited samples (region outlined with a dashed line). The maps have been adaptively smoothed using {\it fadapt} in {\it FTOOLS} with a 6 counts kernel counts threshold. The gray scale map represents the source density in the soft band. The contours shows the source density in the hard band. The 7 contour levels are proportional to the logarithm of the source density. }

363: \label{fig2}

364: \addtolength{\baselineskip}{10pt}

365: \end{inlinefigure}

366: 

367: Fig. 1a and 1b show comparisons of the LogN-LogS in the 9 pointings. 

368: Although the overlapping of fields

369: could minimize the differences between them, this is the simplest  way to 

370: demonstrate the cosmic variance because exactly the same procedures 

371: were used in each field and 8 of the 9 fields have virtually identical 

372: exposures. Variance is seen in both the soft and hard bands.

373: The full range of variance seen in the {\it Chandra}

374: deep surveys published to date is reproduced in this set of 

375: contiguous fields.  

376: 

377: 

378: The LogN-LogS of the whole field is compared with that from {\it Chandra}

379: deep field North (CDF-N; Brandt et al 2001), the {\it Chandra} Deep Field

380: south (CDF-S, Rosati et al. 2001) and SSA13 (Mushotzky et al. 2000) in 

381: Fig.~1c and 1d. 

382: Our LogN-LogS connects smoothly with that of CDF-N, which has the highest 

383: normalization of all the published {\it Chandra} deep fields. 

384: 

385: The LogN-LogS distribution of the combined fields is modeled with power 

386: laws in the form of $N(>S)=A(S/S_{0})^{-\alpha}$, using the area 

387: weighted maximum likelihood method (Murdoch, Crawford \& Jauncey 1973).

388: For the soft band,  between $2 \times 10^{-15}$ and $10^{-14}$ 

389: \ergpcmsqps, $A=630,S_{0}=2\times10^{-15}$, and $\alpha=0.72\pm0.18$; 

390: above $1 \times 10^{-14}$ \ergpcmsqps, $A=152,S_{0}=1\times10^{-14},

391: and \alpha=1.58\pm0.23$. For the hard band above 

392: $1 \times 10^{-14}$ \ergpcmsqps, the best fit parameters are 

393: $A=359,S_{0}=1\times10^{-14}, and \alpha=1.56\pm0.16$.

394: 

395: To visualize the source distributions, we used adaptive smoothing

396: to create density maps of the flux limited sample (Fig.2). 

397: Structure is visible in both the soft and hard band maps, but

398: the hard band sources are more clustered.  

399: %The lump on the west side 

400: %of the field has the highest density. The number of sources in this 

401: %lump is about $2.7-4 \sigma$ higher than the mean of the whole field, 

402: %depending on the definition of the lump. 

403: To test whether the observed over-density could arise from 

404: Poisson fluctuations, we employed the likelihood test described in 

405: Carrera et al.(1998). Comparing with 10000 simulated samples which are

406: Poisson  distributed, we found 97.67\% of the soft and 99.99\% of the  

407: hard band simulations had better likelihood than the observations.  

408: The significance of clustering is therefore

409: 2$\sigma$ (soft) and 4$\sigma$ (hard).  

410: 

411: 

412: Using the source distribution in cells tiling the field (counts-in-cells),

413: we can estimate the correlation scales of the sources.

414: The variance of counts-in-cells defined as

415: $\mu_{2} \equiv \langle(N-\bar{N})^{2}\rangle$, 

416: where $\bar{N}$ is the mean counts in the cell, is directly related to the 

417: angular correlation function (Peebles 1980) by

418: \begin{equation}

419: \mu_{2}=\bar{N}+\frac{\bar{N}^{2}}{\Omega^{2}}\int{w(\theta)d\Omega_{1}d\Omega_{2}}

420: \end{equation} 

421: where $\Omega$ is the cell size. The first term is the Poisson fluctuation. 

422: For correlation functions with a power-law form $ w(\theta)=({\theta}/{\theta_{0}})^{1-\gamma}$ (where $\gamma$ is the power law index of the spatial 

423: correlation function) and square cells with size 

424: $\Omega=\Theta \times \Theta$~deg$^{2}$, the integration can be obtained as

425: (Totsuji \& Kihara 1969; Lahav \& Saslaw, 1992), 

426: \begin{equation}

427: \sigma^{2} \equiv \frac{\mu_{2}-\bar{N}}{(\bar{N}/\Omega)^2}=C_{\gamma}\theta_{0}^{1-\gamma}\Theta^{5-\gamma}

428: \end{equation}

429: where $\sigma^{2}$ is the normalized variance. $C_{\gamma}$ is a function of $\gamma$ and is calculated numerically. We calculate $\sigma^{2}$ for square

430: cells of various sizes that tile the whole field. 

431: By fitting the $\sigma^{2}-\Theta$ relation we should be able to estimate 

432: $\theta_{0}$ and $\gamma$. We found the present data cannot  

433: constrain both parameters accurately. By fixing $\gamma=1.8$, the 

434: ``universal'' slope measured in galaxies and in groups and clusters of 

435: galaxies (Bahcall 1988), and 

436: minimizing $\chi^2$, we found $\theta_{0}=40 \pm 11\arcsec $ and $\theta_{0}=4 \pm 2 \arcsec$ for the hard and soft band sources respectively. 

437: While the soft band sources agree very well with the angular correlation scale

438: previously seen in {\it ROSAT} surveys (Vikhlinin \& Forman, 1995), the hard band sources are much more strongly correlated. 

439: 

440: The striking difference in clustering between the soft and hard band sources 

441: indicates the hard sources which are not detected in the soft band are highly 

442: clustered. About 60\% of the hard-band-only sources lie in overdense regions

443: which form a `band' connecting the `lumps' on the western and eastern side of 

444: the field (Fig. 2). This band includes only  about 1/3 of solid angle of the 

445: whole field. The counts-in-cells analysis (Fig. 3) also indicates that these 

446: hard-band-only sources have larger correlation scales than the soft 

447: band or hard band sources.      

448:     

449: 

450: 

451: \section{Discussion} 

452: {\it ASCA} observations have shown that the rms variance of the 2--10 keV 

453: XRB on a scale of 0.5~deg$^{2}$ is $\sim 6\%$(Kushino et al. 2002). With a sky 

454: coverage of 0.4~deg$^{2}$ and a depth that resolves  

455: $> 50\%$ of the hard band XRB, the normalization of LogN-LogS derived from 

456: our observations should be very close to the ``true'' value. 

457: The fact that the observed LogN-logS connects onto the CDF-N field at low 

458: fluxes then indicates that $>90\%$ of the XRB 

459: is resolved using the {\it ASCA/ROSAT} XRB normalization (Chen et al. 1997). 

460: The main uncertainty is the normalization of XRB itself.  

461:  

462: The LSS seen in our field has reproduced the cosmic

463: variance observed previously in deep field surveys.  

464: It is noticeable (Fig. 1b) that on scales of a 

465: {\it Chandra} field the variance is demonstrated as holes rather than lumps,

466: i.e., the number counts in 9 fields are close to the mean, while only 2

467: field have very low value (ISONW5 \& CDF-S).   

468: This indicates the existence of voids in the X-ray LSS, which should, 

469: within a factor of a few, be of the same angular size as an ACIS-I field. 

470: Most of the hard-band-only sources cluster in relatively small regions, 

471: which may be topologically connected. If this is not a result of a 

472: projection effect, then it is the 

473: first time a wall-like structure has been seen in the X-ray sky.     

474: 

475: We have found more variance in the hard band than in the soft band, 

476: consistent with the lower mean redshift of the hard X-ray sources.

477: However, the difference in the mean redshift cannot account for most of 

478: the large difference in the soft and hard band angular correlation length

479: found by the counts-in-cells statistic. The mean redshift found previously

480: for the {\it Chandra/XMM} deep fields sources is $\sim 0.8$.  

481: Unless most of the hard band sources are at very low redshift, which does 

482: not seem likely because we see no low redshift spikes in the redshift 

483: distribution of deep field surveys, the relatively low redshifts for the 

484: hard X-ray selected sources could not account for the factor of 10

485: larger differences in correlation scales seen in these sources. 

486: The similarity of the redshift distributions for the sources found in 

487: CDF-N and in an observation with a similar exposure to ours 

488: (Castander et al. 2003), indicates that the luminosity functions have 

489: likely been sampled below the ``knee'' for all redshifts

490: in our observations. Thus we feel that the stronger correlation function 

491: for the hard sources seen in our data is likely due to their stronger 

492: spatial correlation rather than a redshift effect.

493: 

494: The large variance in the {\it Chandra} source counts 

495: indicates that they must be highly biased tracers of matter. 

496: Direct calculation of the bias will require redshifts and an understanding 

497: of how the luminosity function changes with redshift, but it is already clear 

498: that the {\it Chandra} sources show much more variance than galaxy counts 

499: (Cohen et al 2000) at similar optical magnitudes. 

500: 

501: \begin{inlinefigure}

502: \centerline{\psfig{figure=fig3.ps,width=0.9\textwidth,angle=90}}

503: %\vspace{12pt}

504: \caption{

505: Normalized variance $\sigma^{2}$ measured with different sized square cells which tiles the whole field. Squares: the hard band sources; triangles: the soft band sources. Crosses: the hard band only sources.  The errors are estimated via a  boots-trap technique. The lines shows the best fit variance with $\gamma=1.8$(fixed) and  $\theta_{0}=40\arcsec \pm 11$ for the hard band, $\theta_{0}= 4\arcsec \pm 2$ for the soft band sources. The hard-band-only points are not fitted due to the small number of sources. }

506: \label{fig3}

507: \addtolength{\baselineskip}{10pt}

508: \end{inlinefigure}

509: 

510: 

511: 

512: \acknowledgements

513: 

514: We thank CXC for their excellent work, 

515: G. Garmire and N. Brandt for providing the CDF-N logN-logS, 

516: C. Reynolds, K. Arnold, K. Jahoda, D. Davis, K. Kuntz, 

517: D. Smith and A. Young for very helpful discussions. 

518: The project is funded under IDS program of 

519: R. Mushotzky, NSF grants AST-0084847 (A.J.B.) and 

520: AST-0084816 (L. L. C. ), the University of Wisconsin

521: Research Committee with funds granted by the Wisconsin Alumni

522: Research Foundation (A.J.B.) and the Alfred P. Sloan Foundation (A.J.B.). 

523: D. B. S. gratefully acknowledges the hospitality of the 

524: MPE and is grateful for support from a senior award from the 

525: Alexander von Humboldt-Stiftung and from NASA through Chandra Award 

526: number GO2-3191C.

527: 

528: \begin{references} 

529: 

530: \reference{} {Bahcall, N. A., 1988 ARA\&A, 26, 631}

531: 

532: \reference{} {Barcons, X. \&  Fabian, A. C. 1988, \mnras, 230, 189} 

533: 

534: \reference{} {Barger, A. J., Cowie, L. L., Mushotzky, R. F., Richards, E. A.

535:   2001 \aj,121,662}

536: 

537: \reference{} {Barger, A. J., Cowie, L. L., Brandt, W. N., Capak, P., Garmire, G. P., Hornschemeier, A. E., Steffen, A. T., Wehner, E. H.  2002, \aj,

538:  124, 1839}

539: 

540: \reference{} {Brandt, W. N., Alexander, D. M., Hornschemeier, A. E., 

541: Garmire, G. P., Schneider, D. P., Barger, A. J., 

542: Bauer, F. E., Broos, P. S. et al. 2001 \aj, 122, 2810}

543: 

544: \reference{} {Campana, S., Moretti, A., Lazzati, D., Tagliaferri, G.\ 2001, \apjl, 560, L19}

545: 

546: \reference{} {Carrera, F. J., Barcons, X., Fabian, A. C., Hasinger, G., Mason, K. O., McMahon, R. G., Mittaz, J. P. D., Page, M. J. 1998, \mnras 299, 229}

547: 

548: \reference{} {Castander, F. J., Treister, E., Maza,Z., Coppi,P.,  Maccarone, T., Zepf, Z., Guzman, R., Ruiz, M. T.  2003, Proceedings of the "X-ray surveys, in the light of the new observatories" workshop, Astron. Nachr., in press}

549: 

550: \reference{} {Cohen, J. G., Hogg, D. W., Blandford, R., Cowie, L. L.,

551:  Hu, E., Songaila, A., Shopbell, P., Richberg, K. et al. 2000, \apj, 538, 2} 

552: 

553: 

554: 

555: \reference{} {Cowie, L. L., Garmire, G. P., Bautz, M. W., Barger, A. J., 

556: Brandt, W. N., Hornschemeier, A. E. 2002, \apjl, 566, L5}

557: 

558: %\reference{} {Croft,R.A.C., Weinberg,D.H., Katz,N., Hernquist,L. 1997, \mnras,

559: %291,305}

560: 

561: \reference{} {Chen, L.-W.; Fabian, A. C.; Gendreau, K. C. 1997 \mnras, 

562: 285, 449}  

563: 

564: %\reference{} {Collins,C.A., Guzzo,L., B\"{o}hringer,H., Schuecker,P., 

565: %Chincarini,G., Cruddace,R., De Grandi,S., MacGillivray,H.T., Neumann,D.M., 

566: %Schindler,S. et al. 2000 \mnras 319, 939}

567:  

568: \reference{} {Dickey \& Lockman, 1990, ARA\&A, 28, 215}

569: 

570: \reference{} {Freeman, P. E., Kashyap, V., Rosner, R., Lamb, D. Q. 2002, \apjs,

571: 138, 185}

572: 

573: \reference{} {Garmire, 2002, private communication}

574: 

575: \reference{} {Giacconi, R., Zirm, A.,  Wang, J., Rosati, P., Nonino, M., Tozzi, P., Gilli, R., Mainieri, V. et al. 2002, \apjs, 139, 369}

576: 

577: \reference{} {Hasinger, G., Altieri, B., Arnaud, M., Barcons, X., Bergeron, J., Brunner, H., Dadina, M., Dennerl, K. et al. 2001, A\&A, 365, L51} 

578: 

579: \reference{} {Hasinger,G. 2002, AAS Conference 199 No.138.19}

580:  

581: \reference{} {Hornschemeier, A. E., Brandt, W. N., Garmire, G. P., Schneider, D. P., Barger, A. J., Broos, P. S., Cowie, L. L., Townsley, L. K. et al. 2001 

582: \apj 554, 742}

583: 

584: \reference{} {Kushino, A., Ishisaki, Y., Morita, U., Yamasaki, N. Y., Ishida, M., Ohashi, T., Ueda, Y. 2002,  PASJ, 54, 327}

585:  

586: \reference{} {Lahav, O.,  Saslaw, W. C. 1992, \apj, 396, 430L}

587: 

588: \reference{} {Mullis, C. R., 2002, PASP 114,688}

589: 

590: \reference{} {Murdoch, H. S., Crawford, D. F., Jauncey, D. L. 1973, \apj, 183, 1}

591: \reference{} {Mushotzky, R. F., Cowie, L. L., Barger, A. J., Arnaud, K. A., 2000, \Nat, 404, 459}

592: 

593: \reference{} {Mushotzky, R. \& Jahoda, K., 1992, in {\it The X-ray Background}, eds. Barcons,X., Fabian, A. C., Cambridge University Press, Cambridge}

594: 

595: \reference{} {Peebles, P. J. E., 1980, {\it The Large Scale Structure of the Universe},  Princeton University Press}

596: 

597: \reference{} {Rosati, P., Tozzi, P., Giacconi, R., Gilli, R., Hasinger, G., Kewley, L., Mainieri, V., Nonino, M. et al.  2002, \apj, 566, 667}

598: 

599: \reference{} {Soltan, A. \& Hasinger, G. 1994, A\&A, 288, 77}

600: 

601: \reference{} {Soltan, A. M., Hasinger, G., Egger, R., Snowden, S., 

602: Truemper, J. 1997, A\&A, 320, 705}

603: 

604: \reference{} {Totsuji, H.,  Kihara, T. 1969, PASP, 21 , 221}

605: 

606: \reference{} {Tozzi, P., Rosati, P., Nonino, M., Bergeron, J., Borgani, S., Gilli, R., Gilmozzi, R., Hasinger, G. et al. 2001, \apj, 562, 42}

607: 

608: \reference{} {Vikhlinin, A., Forman, W. 1995, \apj, 455, L109}

609: 

610: \end{references}

611: 

612: %\plotone{fig1a.ps}

613: %\plotone{fig1b.ps}

614: %\plotone{fig2.ps}

615: %\plotone{fig3.ps}

616: %\plotone{fig4.ps}

617: 

618: \end{document}

619: