1: %

2: % Lensing SMMs (or cold?)

3: %

4: \documentstyle[draft,psfig]{mn}

5: %

6: 

7: % For correct printing on US Letter, while still working on A4

8: \topmargin-1cm

9: 

10: 

11: \def\gs{\mathrel{\raise0.35ex\hbox{$\scriptstyle >$}\kern-0.6em

12: \lower0.40ex\hbox{{$\scriptstyle \sim$}}}}

13: \def\ls{\mathrel{\raise0.35ex\hbox{$\scriptstyle <$}\kern-0.6em

14: \lower0.40ex\hbox{{$\scriptstyle \sim$}}}}

15: %\addtolength{\textheight}{0.5in}

16: 

17: \date{\fbox{\sc Draft: \today\ --- Do Not Distribute}}

18: 

19: \title[Lensing in blank-field SCUBA surveys]{

20: The effect of lensing on the identification

21: of bright SCUBA galaxies.}

22: 

23: \author[Chapman et al.]

24: {S.\,C.\ Chapman,$^{\! 1}$

25: Ian Smail,$^{\! 2}$

26: R.\,J.\ Ivison$^{3}$ \&

27: A.\,W.\ Blain$^1$

28: \vspace*{1mm}\\

29: 	$^1$ California Institute of Technology, Pasadena, CA 91125\\

30:         $^2$ Department of Physics, University of Durham, South Road,

31:         Durham DH1 3LE\\

32:         $^3$ Astronomy Technology Centre, Royal Observatory, Blackford Hill,

33:         Edinburgh EH9 3HJ}

34: 

35: 

36: 

37: \date{Accepted ... ; Received ... ; in original form ...}

38: 

39: \pagerange{000--000}

40: 

41: \begin{document}

42: 

43: \maketitle

44: 

45: 

46: \begin{abstract}

47: Spectroscopic surveys of luminous submillimetre-selected sources have

48: uncovered optically-bright galaxies at $z\ls 1$ close to the positions

49: of several submillimetre (submm) sources.  Naive statistical analyses

50: suggest that these galaxies are associated with the submm emission.

51: However, in some cases, it is difficult to understand this association

52: given the relatively modest redshifts and unpreposessing spectral

53: characteristics of the galaxies. These are in stark constrast to those

54: expected from the massive dust-enshrouded starbursts and AGN thought

55: to power the bulk of the bright submm population.  We present new

56: observations of optically-bright counterparts to two luminous submm

57: sources, along with a compilation of previously proposed

58: optically-bright counterparts with $z\ls 1$.  We suggest that the

59: majority of these associations between bright galaxies and submm

60: sources may be due to the action of the foreground galaxies as

61: gravitational lenses on the much fainter and more distant submm

62: sources.  We discuss the implications of this conclusion for our

63: understanding of the SCUBA population.

64: \end{abstract}

65: 

66: \begin{keywords}

67:    galaxies: starburst

68: -- galaxies: formation 

69: -- infrared: galaxies

70: \end{keywords}

71: 

72: %

73: %

74: %

75: \section{Introduction}

76: 

77: 

78: %

79: % Figure 1: OBRS

80: %

81: \begin{figure*}

82: \centerline{\psfig{figure=scott.eps,angle=0,width=6.0in}}

83: \caption{\footnotesize

84: Images (18\arcsec$\times$18\arcsec) of the two $z<1$ candidate submm

85: sources from our ESI survey: SMM\,J16370+4106 (left panel, $V$-band)

86: and SMM\,J22173+0014 (centre panel, $K$-band).  The right-hand panel

87: shows a zoomed $7''\times 7''$ view of SMM\,J22173+0014 as seen in an

88: {\it HST} STIS $R'_{573}$ image.  Radio contours are overlaid starting

89: at $3\sigma$, and increasing in $1\sigma$ intervals.

90: }

91: \label{fig1}

92: \end{figure*}

93: 

94: The intentional use of gravitational lenses has been a particularly

95: profitable route to probe the nature of the submm galaxy population

96: (Smail, Ivison \& Blain 1997; Chapman et al.\ 2002a; Smail et al.\

97: 2002; Cowie, Barger \& Kneib 2002).  The lens amplification

98: facilitates the identification and characterisation of counterparts at

99: other wavelengths (Smail et al.\ 2002), and has produced some of the

100: best-studied examples of the submm population (e.g.\ Ivison et al.\

101: 1998, 2000; Soucail et al.\ 1999).  Blank field submm surveys have

102: also been undertaken (e.g.\ Hughes et al.\ 1998; Barger et al.\ 1999;

103: Eales et al.\ 1999, 2000; Borys et al.\ 2002; Scott et al.\ 2002; Webb

104: et al.\ 2002), these cover large contiguous areas and are in principle

105: simpler to analyse than the lens surveys.  However, without the boost

106: from a cluster lens, identifying counterparts in the blank field

107: surveys has proved to be an arduous task.  For example, significant

108: effort has been expended in attempting to track down the brightest

109: submm source in the Hubble Deep Field North, HDF\,850.1, with no clear

110: resolution yet.

111: 

112: The published blank field SCUBA surveys have reliably identified a

113: small fraction of their submm sources in the optical and near-infrared

114: (Gear et al.\ 2000; Lutz et al.\ 2001), with a mere handful having

115: spectroscopic redshifts, e.g.\ Westphal-MMD11 (Chapman et al.\ 2002b);

116: CUDSS\,3.8, 3.10, 14.13 and 14.18 (Eales et al.\ 2000; Webb et al.\

117: 2002).  Although the spectral properties of some of these proposed

118: counterparts appear reasonable given their redshifts, several of them

119: must have very low dust temperatures compared to similar luminosity

120: galaxies in the local Universe, if they are to produce the observed

121: submm emission at the proposed redshifts, $z\ls 1$ (Eales et al.\

122: 1999; Dunne \& Eales 2001).

123: 

124: There are only three SCUBA galaxies with spectroscopic redshifts which

125: have been confirmed in CO line emission, all these lie at $z\geq 1$

126: (e.g.\ Frayer et al.\ 1998, 1999), consistent with the estimates of

127: the median redshift for the whole population of $z\sim 2$--3 from

128: their submm and radio spectral properties (Hughes et al.\ 1998;

129: Carilli \& Yun 2000; Smail et al.\ 2000; Yun \& Carilli 2002), with

130: few galaxies at $z\ll 1$.  The existence of a moderate fraction of

131: SCUBA galaxies with redshifts of $z\ll 1$ would therefore indicate a

132: bimodal redshift distribution for the population, and suggest that two

133: physically distinct classes of SCUBA galaxies contribute to the faint

134: submm counts.  Alternatively, these apparent low redshift counterparts

135: could be misidentifications -- but why are they so frequent?

136: 

137: Given the expected high median redshift of the submm population and

138: the steep number counts seen for the brighter submm sources (Blain et

139: al.\ 1999; Fox et al.\ 2002), gravitational lensing by foreground

140: galaxies is one possible explanation for the anomolously high rate of

141: association of submm sources with low redshift, optically bright

142: galaxies.  Surveys of distant QSOs have provided estimates of the rate

143: of lensing for optical samples of $z>1$ QSOs, 0.7\% (e.g.\ Surdej et

144: al.\ 1993; Jaunsen et al.\ 1995; Kochanek et al.\ 1995, see also

145: Kochanek 1993).  A more reliable constraint comes from lens searches

146: in the radio waveband where obscuration in the lens is not a concern,

147: these suggest that a modest fraction of flux limited radio-loud AGN

148: samples, 0.25--0.57\%, are strongly lensed by field galaxies at $z\ls

149: 1$ (e.g.\ King et al.\ 1999; Myers et al.\ 1999).  This same

150: population of lenses will also act on the similarly distant SCUBA

151: population, although differences in the intrinsic count slope of the

152: two populations will lead to a different rate of occurence of strong

153: (and weak) lensing in flux limited submm samples.  The issue of the

154: effects of gravitational lensing on field surveys in the submm

155: waveband have been discussed by Blain (1996; 1998), who suggested that

156: $\sim 2$\% of SCUBA galaxies with S$_{\rm 850 \mu m}\simeq 10$\,mJy

157: will be amplified by $\geq 2\times$ by foreground lenses.  The number

158: of submm sources in published samples is now sufficient that we should

159: be able to observationally test this prediction and in doing so

160: investigate the nature of those SCUBA galaxies which apparently lie at

161: $z\ll 1$.

162: 

163: %

164: %

165: %

166: \section{Observations and Reduction}

167: 

168: We have embarked on a program to obtain spectroscopic redshifts for a

169: large sample of radio-selected, optically bright submm sources.  The

170: goal of this project is to identify the optically-brightest SCUBA

171: galaxies, such as SMM\,J02399$-$0136 and SMM\,J14011+0252 (Ivison et

172: al.\ 1998; 2000, 2001) which are suitable for detailed follow-up on

173: 10-m class telescopes.  To achieve this we extend the previous work on

174: submm observations of optically-faint radio-selected sources (e.g.\

175: Chapman et al.\ 2001b) to brighter counterparts.  Chapman et al.\

176: (2002e) have demonstrated the success of this approach, recovering

177: $\sim$60\% of the SCUBA population.  The expectation is that the

178: remaining 40\% include both optically luminous SCUBA galaxies (Ivison

179: et al.\ 1998, 2000) and very high redshift sources which are too faint

180: in the radio waveband to be detected (e.g.\ Frayer et al.\ 2000).  The

181: bias against very distant SCUBA galaxies in radio surveys depends on

182: their dust temperature: hotter sources can be seen out to higher

183: redshifts, $z\gs 3$, while colder sources are more difficult to detect

184: (Chapman et al.\ 2002d).

185: 

186: %

187: % Figure 2

188: %

189: \centerline{\psfig{figure=N2_850_1spec.ps,angle=0,width=2.5in}}

190: \centerline{\psfig{figure=SA22_west.ps,angle=0,width=2.5in}}

191: \noindent{\footnotesize \addtolength{\baselineskip}{-3pt}

192: {\bf Fig.~2:}  Keck/ESI spectra of SMM\,J16370+4106 (upper

193: panel) and SMM\,J22173+0014 (lower panel). We identify redshifted

194: emission lines in the spectra of both galaxies which place them at

195: $z<1$.

196: 

197: }

198: 

199: \subsection{ESI Observations}

200: 

201: The initial observations for our survey of optically bright SCUBA

202: sources involved spectroscopy of 7 sources with the Echellette

203: Spectrograph and Imager (ESI) on the Keck\,II telescope in 2001 July.

204: The echellette mode provides complete coverage from 0.32--1\,$\mu$m at

205: high spectral resolution $\sim$1\AA, allowing very good sky

206: subtraction into the atmospheric OH forest at the red end of the

207: spectra.  The spectral integrations were 1800\,s for each source.

208: High signal-to-noise flats and wavelength calibrations were taken

209: shortly before the observations of each target, and fluxes were

210: calibrated using spectra of red standard stars.  The data were reduced

211: with the {\sc makee} software using the reduction recipe described in

212: Barlow \& Sargent (1997).

213: 

214: %

215: % Table 1 (Catalogue)

216: %

217: \begin{table*}

218: {\footnotesize

219: \centerline{\sc Properties of Sample}

220: \begin{center}

221: \begin{tabular}{lcccccccl}

222: \hline\hline

223: \noalign{\smallskip}

224:  {Source}  & $\delta r^a$ &  $z_{\rm lens}$ & $z_{\rm CY}$ & T$_{\rm d}^b$ &  

225:    $I$-mag$^c$ &  $S_{850 \rm \mu m}$ & $S_{1.4 \rm GHz}$ & \cr

226:      {}    &  $('')$    &                 &              & (K)         &     &   (mJy)              & ($\mu$Jy)         & \cr

227: \noalign{\smallskip}

228: \hline

229: \noalign{\smallskip}

230:  SMM\,J16370+4106 & 2.0 & ~0.845 &2.2 

231: & 23 & 21.4 &  11.2$\pm$2.0 & 99$\pm$20 & N2\,850.1, Ivison et al.\ (2002) \cr

232:  SMM\,J22173+0014 & 1.7 & ~0.510 &2.4

233: & 19 & 19.4 &  15.0$\pm$3.0 & 145$\pm$18 & SSA22\,850.2, Chapman et al.\ (2001a) \cr

234: \noalign{\medskip}

235:  SMM\,J00266+1708$^d$ & 1.5 & ~0.44 & 2.9 & 16 & 22.0 &  18.6$\pm$1.5 & 100$\pm$15 & ERO: M12, Frayer et al.\ (2000) \cr

236:  SMM\,J04431+0210$^d$ & 2.3 & ~0.18 & $>2.6$~ & 14 & 18.4 &  7.2$\pm$1.5 & $<57$, 3$\sigma$ & ERO: N4, Smail et al.\ (1999) \cr

237:  SMM\,J12369+6212 & 0.7 & $\sim$1.1 & 4.1 & 18 & 22.6 & 7.4$\pm$0.5 & $<24$, 3$\sigma$ & HDF\,850.1, Downes et al.\ (1999) \cr

238: \hline

239: \end{tabular}

240: 

241: \begin{tabular}{l} 

242: $^a$ Angular separation between the bright galaxy and the

243: nominal position of the submm, mm or radio centroid.\cr

244: $^b$  T$_{\rm d}$ is the dust temperature required for the source to

245: have a radio/submm redshift consistent with the spectroscopic \cr

246: ~~ measurement of the galaxy.  \cr

247: $^c$  The $I$-band magnitude of the optically bright counterpart. \cr

248: $^d$  Fluxes have not been corrected for cluster lensing. \cr

249: \end{tabular}

250: 

251: \end{center}

252: }

253: \label{tab1}

254: \end{table*}

255: 

256: These observations generated redshifts for five new candidate

257: counterparts to submm galaxies, to be discussed in a future paper.

258: However, the proposed optically-bright counterparts to two of these

259: sources lie at much lower redshifts than expected from their radio and

260: submm properties, and we discuss these further here.  These two

261: sources are: SMM\,J16370+4106 from the SCUBA survey of Scott et al.\

262: (2002) and SMM\,J22173+0014 from Chapman et al.\ (2001a).  We show

263: radio/optical overlays of these two sources in Fig.~1 and summarise

264: their properties in Table~1.  The calibrated spectra are shown in

265: Fig.~2 with their spectral features identified.  These spectra show

266: that the proposed optically-bright counterpart to SMM\,J16370+4106

267: lies at $z=0.845$, while the galaxy identified with SMM\,J22173+0014

268: has $z=0.510$.  We note that the spectrum of SMM\,J22173+0014 shows

269: broad lines indicative of a low luminosity type~I AGN, although such

270: activity is not infrequent in the high-redshift field population

271: (Cowie et al.\ 1996).

272: 

273: As we discuss below the properties of both of these systems strongly

274: suggest that the optically bright, spectroscopically-identified

275: galaxies are not in fact the source of the submm emission, but instead

276: they may represent foreground gravitational lenses which are

277: amplifying the more distant, optically-faint SCUBA galaxy.

278: 

279: %

280: %

281: %

282: \section{Analysis and Discussion}

283: 

284: We have cataloged two new, $>10$\,mJy submm sources which are within

285: 1--2$''$ of $I<21.5$ galaxies at $z=0.510$ and $0.845$.  In the

286: absence of detections in deep, high resolution radio maps, the coarse

287: spatial resolution of the SCUBA detections would mean that these

288: offsets would not be significant, and combined with the low surface

289: density of $I<21.5$ galaxies this would lead to a high probability

290: that the submm source is associated with the optically bright

291: galaxy. There are two possible explanations for these associations:

292: 1) either the optically-bright galaxies are the source of the submm

293: emission, or 2) these galaxies are lensing the true submm source,

294: which is then both fainter and lies at much higher redshifts.

295: 

296: If the spectroscopically-identified galaxies are the source of the

297: submm and radio emission then we can estimate the characteristic dust

298: temperatures, T$_{\rm d}$, which these galaxies must have to be

299: consistent with the measured radio/submm spectral index. We list these

300: values in Table~1 and compare them to the distribution of dust

301: temperatures (Dunne et al.\ 2000) as a function of galaxy luminosity

302: seen locally in Fig.~3. The distribution is shown as log normal,

303: consistent with the best fit distribution for the local 1.2\,Jy IRAS

304: sample (Chapman et al.~2002f).

305: 

306: In the case that these systems represent gravitational lenses we list

307: for comparison in Table~1 the properties of three sources from

308: published SCUBA surveys which are all believed to be lensed by

309: foreground galaxies: SMM\,J00266+1708 (Frayer et al.\ 2000) and

310: SMM\,J04431+0210 (Smail et al.\ 1999), both of these lie within $\sim

311: 2''$, but are not coincident with, bright foreground galaxies (images

312: of these systems can be found in the relevant references).  We also

313: include the candidate lensed source, HDF\,850.1 from Hughes et al.\

314: (1998).  The millimetre interferometry map of HDF\,850.1 by Downes et

315: al.\ (1999) indicates that the submm source is offset by 0.7\arcsec\

316: from a moderately bright Elliptical galaxy at modest redshift,

317: suggesting it is likely to be lensed (Downes et al.\ 1999).  

318: Very recently an extremely faint $K$-band source has been

319: found at the millimetre position, providing further support for the

320: identification of this system as a high redshift lensed SCUBA galaxy

321: (Dunlop et al.\ 2002).

322: 

323: We now discuss the properties of the two new submm sources to attempt

324: to distinguish between the two possible explanations of their

325: properties.

326: 

327: For SMM\,J16370+4106 we find a close pair of radio sources -- the

328: brighter radio component is consistent with being coincident with the

329: optically-bright galaxy, within the optical/radio astrometric

330: uncertainty ($\sim0.2$\arcsec).  The second, fainter radio component

331: lies $\sim 2''$ east of the bright optical galaxy.  The two radio

332: components could both be associated with the bright galaxy as the

333: chance of a radio/submm source being lensed by another foreground

334: radio source is low, due to the low radio source density at these flux

335: levels.  At the same time, the astrometric uncertainty does not allow

336: us to reject the possibility that this system is a highly amplified

337: double image lensing configuration, straddling the optical lens, which

338: would circumvent the radio source number density/probability argument.

339: The required T$_{\rm d}$ for SMM\,J16370+4106 at $z=0.845$ is $23\pm

340: 5$\,K, placing it $\sim 4\sigma$ below the local relation in Fig.~3

341: and strengthening the lensing hypothesis.

342:          

343: SMM\,J22173+0014 lies only 30\arcsec\ from a known submm source at the

344: centre of a $z=3.1$ galaxy overdensity (Steidel et al.\ 1998), yet

345: appears to be associated with a luminous early-type spiral at

346: $z=0.510$. However, the radio centroid is significantly offset from

347: the optical galaxy center, five times the relative astrometric error

348: in the VLA/optical frames, which is dominated by the source

349: centroiding, ${\rm FWHM}/(2\times S/N) = {5\arcsec}/{17} =

350: 0.3\arcsec$, with the radio emission falling close to a faint $K$-band

351: extension from the galaxy. The {\it HST} image obtained subsequent to

352: the spectroscopic observations reveals a morphologically complex

353: systems, reminiscent of a merger, coincident with this extension.  In

354: addition, the calculated T$_{\rm d}$ required to reproduce the

355: radio/submm spectral index at this redshift is only $18\pm 2$\,K,

356: putting this source $\sim 5\sigma$ below the local mean T$_d$ for its

357: luminosity (Fig.~3).  Together, these arguments suggest that

358: SMM\,J22173+0014 is a distant submm source lensed by the foreground

359: galaxy at $z=0.510$.  The configuration of this system is very similar

360: to that seen in SMM\,J00266+1708 and SMM\,J04431+0210, which have

361: bright edge-on spiral galaxies within 2--3$''$ of the submm source

362: (Frayer et al.\ 2000; Smail et al.\ 1999).  These configurations are

363: also consistent with the results from Blain, M\"oller \& Maller (1999)

364: and Bartelmann \& Loeb (1998), who have shown that nearly edge-on

365: spirals provide a high lensing probability, due to the large projected

366: surface mass density.

367: 

368: 

369: %

370: % Figure 3

371: %

372: 

373: \centerline{\psfig{figure=lensTd.ps,angle=0,width=3.5in}}

374: \noindent{\footnotesize\addtolength{\baselineskip}{-3pt}

375: {\bf Fig.~3:} The distribution of T$_{\rm d}$ versus far-infrared

376: luminosity for local luminous, dusty galaxies (circles), the best fit

377: relationship is shown as a solid line and the error bounds by dashed

378: lines.  We show our two new systems by filled squares, assuming that

379: the submm and radio emission is associated with the bright optical

380: galaxy, as well as the lensed submm sources from the literature (open

381: squares).  We also show two distant, luminous {\it ISO}-FIRBACK

382: galaxies from Chapman et al.\ (2002d) which have low dust temperatures

383: (stars).

384: 

385: }

386: 

387: While there is strong evidence that the two new submm sources are both

388: lensed systems, we first need to consider selection effects in the

389: radio/submm sample.  The local luminous {\it IRAS} galaxies in Fig.~3

390: show a broad distribution in T$_{\rm d}$, with no examples as cold as

391: the SCUBA galaxies would need to be to be $z<1$.  However, the {\it

392: IRAS} galaxies represent a restframe 60-$\mu$m selected sample, while

393: the SCUBA galaxies are initially identified in terms of their 1.4-GHz

394: and 850-$\mu$m emission.  This will tend to bias their selection

395: towards galaxies with lower T$_{\rm d}$ (Eales et al.\ 1999).  This

396: effect has recently been demonstrated for two SCUBA sources from the

397: {\it ISO}-FIRBACK survey (Chapman et al.\ 2002d).  These 170-$\mu$m

398: selected sources have accurate radio positions which identify them

399: with relatively low redshift galaxies ($z=$0.45 and 0.91) which both

400: exhibit clear merger morphologies. The best fit dust temperatures for

401: these luminous galaxies (calculated for consistancy with the sample in

402: Fig.~3) are 28\,K and 33\,K, indicating that some cold, but luminous

403: galaxies do exist at high redshifts (Fig.~3).  There is also some

404: evidence that the dust temperature distribution may be broader at

405: higher redshifts (Chapman et al.\ 2002f), making these apparently cold

406: sources more common.  The 1-$\sigma$ errors on SMM\,J16370+4106

407: overlap with those of the cold FIRBACK sources, suggesting that the

408: cold dust explanation is conceivable for this source.  However, none

409: of these galaxies have as low T$_{\rm d}$ as is required for

410: SMM\,J22173+0014 and we therefore conclude that SMM\,J22173+0014 most

411: probably represents a lensed SCUBA galaxy.

412:  

413: Assuming that at least one, and perhaps both, of these two submm

414: sources are likely to be lensed we now estimate the proportion of

415: similarly lensed sources in a typical SCUBA survey.  Our survey

416: started with a subsample of the total submm population which were

417: radio identified, with the requirement that the radio source aligned

418: with an $I<23.5$ optical galaxy to within 2\arcsec.  The fraction of

419: submm sources in total which have a bright ($I<23.5$) galaxy nearby

420: (Chapman et al.\ 2002c) represents 16\% of radio identified submm

421: population.  Radio selection detects $\sim60$\% of all bright

422: ($S>5$\,mJy) submm galaxies.  We found above that 1--2 from 5 sources

423: in our Keck/ESI sample were likely to be lensed.  This implies 3--5\%

424: of the radio detected submm population could be lensing candidates,

425: with a lower limit of 2--3\% for the whole SCUBA population.

426: 

427: We have also discussed three lensed SCUBA galaxies from the literature

428: which are very similar to the two candidate systems presented here.

429: The submm source HDF\,850.1 appears to be lensed by a foreground

430: elliptical galaxy, this gives a rate of lensed sources of 1/8, or

431: $\sim 13$\% with a large uncertainty, for the HDF sources discussed by

432: Hughes et al.\ (1998) and Serjeant et al.\ (2002).  The other two

433: lensed sources come from the Smail et al.\ (2002) cluster sample,

434: these are of course more likely to suffer galaxy-galaxy lensing due to

435: the high foreground galaxy concentration in the clusters used in this

436: survey.  For that reason we take the rate of galaxy-lensing from this

437: survey as an upper limit, 2/15 sources, or $\leq 13$\%.  Overall this

438: suggests that around 3--5\% of submm sources from SCUBA surveys are

439: likely to be gravitationally amplified by foreground galaxies.

440: 

441: The rate of lensing we find is nearly an order of magnitude higher

442: than the incidence of multiple-imaging in similarly distant QSO

443: samples, which combined with the radio morphologies of the galaxies

444: discussed here indicates that these are not highly-amplified,

445: multiply-imaged SCUBA sources.  With our current crude knowledge of

446: the lensing configurations in these systems, it is impossible to

447: produce detailed lens models for these sources.  However, the relative

448: separation of the source and lens, along with the apparent lack of

449: multiple images suggests amplifications of $\ls 5$.

450: 

451: The probability of lensing in SCUBA surveys depends on the form of the

452: submm counts, with the steep count slope seen for submm sources

453: brighter than 5\,mJy (Scott et al.\ 2002), $\sim2$\% of the

454: $\sim10$\,mJy sources could be amplified by $\geq 2\times$ (Blain

455: 1998).  This figure is roughly in agreement with our findings,

456: especially if SMM\,J16370+4106 actually represents a cold, luminous

457: galaxy.

458: 

459: We note that weak amplification bias similar to, but weaker than, that

460: invoked here has been invoked as a possible explanation for the strong

461: cross correlation between foreground bright galaxies and submm sources

462: seen by Almaini et al.\ (2002).  We also note that lensing is also

463: expected to manifest at the $\sim 1$\% level given the surface density

464: of sources now reached in the deepest radio maps, $\sim 15\mu$Jy

465: (e.g.\ Formalont et al.\ 2002; Owen et al.\ 2002). However, the effect

466: cannot be much greater as $\mu$Jy count slopes are close to Euclidean,

467: and the $N(z)$ is less extended than for the submm-selected samples

468: ($<\! z\! > \sim 0.6$, Richards et al.\ 1999; Barger, Cowie \&

469: Richards 2000; Chapman et al.\ 2002c).

470: 

471: %

472: %

473: %

474: \vspace*{-2mm}

475: \section{Conclusions}

476: 

477: We have identified two submm sources which have apparent counterparts

478: which are optically bright galaxies, $I<21.5$, lying at modest

479: redshifts, $z<1$. We suggest two possible explanation for these

480: sources, and similar systems in the literature.  The first explanation

481: states that the suggested identifications are correct, in which case

482: these galaxies represent a class of very cold, luminous submm source

483: (Chapman et al.\ 2002d).  For one source, SMM\,J16370+4106 at

484: $z=0.845$, this interpretation appears possible, however, it seems

485: less likely for the second and even colder source, SMM\,J22173+0014.

486: In both cases, the most powerful test to reject this hypothesis would

487: be to confirm the absence of luminous molecular CO emission at the

488: redshift of the optically bright galaxies.

489: 

490: The second explanation for these systems is that the optically bright

491: galaxy represents a foreground gravitational lens, which is amplifying

492: the more distant SCUBA source.  This explanation is consistent with

493: both the observed source configurations and their spectral properties.

494: In this case we estimate that up to 3--5\% of the $\gs 10$\,mJy submm

495: sources in blank field SCUBA surveys could be gravitationally

496: amplified by foreground galaxies.  The most important result of this

497: amplification is that it leads to the misidentification of the submm

498: source with the nearby bright, and typically low redshift, galaxies.

499: These misidentifications would produce a false tail of low-redshift,

500: $z\ll 1$, submm sources in all far-infrared/submillimetre/millimetre

501: surveys.  This is especially a concern for those surveys where deep,

502: high resolution data is not available to confirm the correspondance

503: between the radio (pin-pointing the longer wavelength emission source)

504: and optical sources with a precision of $\ll 1''$.

505: 

506: \vspace*{-2mm}

507: \section*{Acknowledgements}

508: We acknowledge our collaborators on the HST-SCUBA morphology project,

509: R.\ Windhorst \& E.\ Richards.

510: SCC acknowledges support from NASA through HST grant 9174.1.

511: IRS acknowledges support from a Royal Society URF and a

512: Philip Leverhulme Prize Fellowship.  

513: 

514: \begin{thebibliography}{}

515: \bibitem{} Almaini, O., et al., 2002, MNRAS, in press, (astro-ph/0108400)

516: \bibitem{} Barger, A., Cowie, L., Richards, E., 2000, AJ, 119, 2092

517: \bibitem{} Barger, A., et al., 1999, AJ, 117, 2656

518: \bibitem{} Barlow, T., Sargent, W.L.W., 1997, AJ, 113, 136

519: \bibitem{} Bartelmann, M., Loeb, A., 1998, ApJ, 503, 48 

520: \bibitem{} Blain A.W., Smail I., Ivison R.J., Kneib J.-P., 1999, 

521:         MNRAS, 302, 632

522: \bibitem{} Blain, A., M\"oller, O., Maller, A., 1999, MNRAS, 303, 423

523: \bibitem{} Blain, A., 1996, MNRAS, 283, 1340

524: \bibitem{} Blain, A., 1998, MNRAS, 297, 502

525: \bibitem{} Borys, C., Chapman, S., Halpern, M., Scott, D., 2002, MNRAS, 

526: 	in press, (astro-ph/0107515) 

527: \bibitem{} Carilli, C., Yun, M., 2000, ApJ, 530, 618

528: \bibitem{} Chapman, S.C., Lewis, G., Scott, D., Richards, E., et al.,

529: 	2001a, ApJL, 548, 17

530: \bibitem{} Chapman, S.C., Richards, E., Lewis, G., Wilson, G., Barger, A.,  

531: 	2001b, ApJL, 548, 147

532: \bibitem{} Chapman, S.C., Scott D., Borys C., Fahlman G.G., 2002a, 

533:         MNRAS, 330, 92

534: \bibitem{} Chapman, S.C., et al. 2002b, ApJL, submitted 

535: \bibitem{} Chapman, S.C., et al. 2002c, AJ, submitted 

536: \bibitem{} Chapman, S.C., et al. 2002d, ApJ, in press, May 1 issue, 

537: 	(astro-ph/0111157)

538: \bibitem{} Chapman, S.C., et al. 2002e, ApJ, in press, (astro-ph/0203068) 

539: \bibitem{} Chapman, S.C., et al. 2002f, ApJ, submitted 

540: \bibitem{} Cowie, L.L., Songaila, A., Hu, E.M., Cohen, J.C., 1996, AJ, 112, 839

541: \bibitem{} Cowie, L.L., Barger, A.J., Kneib, J.-P., 2002, AJ, in press

542: \bibitem{} Downes, D., et al., 1999, A\&A, 347, 809

543: \bibitem{} Dunlop, J.S., et al., 2002, in prep 

544: \bibitem{} Dunne, L., Eales, S., 2001, MNRAS, 327, 697

545: \bibitem{} Dunne, L., et al., 2000, MNRAS, 301, 230

546: \bibitem{} Eales, S., et al., 1999, ApJ, 515, 518 

547: \bibitem{} Eales, S., et al., 2000, AJ, 120, 2244

548: \bibitem{} Fomalont, E., et al., 2002, in prep

549: \bibitem{} Fox, M.J., et al., 2002, MNRAS, submitted (astro-ph/0107585)

550: \bibitem{} Frayer, D.T., Ivison, R.J., Scoville, N.Z., Yun, M., Evans, A.S., Smail, I., Blain, A.W., Kneib, J.-P., 1998, ApJL, 506, L7

551: \bibitem{} Frayer, D.T., Ivison, R.J., Scoville, N.Z., Evans, A.S., Yun, M., Smail, I., Barger, A.J., Blain, A.W., Kneib, J.-P., 1999, ApJL, 514, L13

552: \bibitem{} Frayer, D.T., Smail, I., Ivison, R.J., Scoville, N.Z., 2000, AJ, 120, 1668

553: \bibitem{} Gear, W., et al., 2000, MNRAS, 316, 51

554: \bibitem{} Hughes, D., et al., 1998, Nature, 394, 241

555: \bibitem{} Ivison, R.J., et al., 1998, MNRAS, 298, 583

556: \bibitem{} Ivison, R.J., et al., 2000, MNRAS, 315, 209

557: \bibitem{} Ivison, R.J., et al., 2001, ApJ, 561, 45L

558: \bibitem{} Ivison, R.J., et al., 2002, MNRAS, submitted

559: \bibitem{} Jaunsen, A., Jablonski, M., Pettersen, B., Stabell, R., 1995, 

560: 	A\&A, 300, 323

561: \bibitem{} King, L., et al., 1999, MNRAS, 307, 225

562: %\bibitem{} Kochanek, C., 1993, ApJ, 419, 12

563: \bibitem{} Kochanek, C., Falco, E., Schild, R., 1995, APJ, 452, 109

564: \bibitem{} Lutz, D., et al., 2001, A\&A, 378, 70

565: \bibitem{} Myers, S.T., et al., 1999, AJ, 117, 2565

566: \bibitem{} Owen, F., et al., 2002, in prep.

567: \bibitem{} Richards, E., et al., 1999, ApJ, 526, 73L

568: \bibitem{} Scott, S., et al., 2002, MNRAS, in press, (astro-ph/0107446)

569: \bibitem{} Serjeant, S., et al. 2002, MNRAS, (astro-ph/0201502)

570: \bibitem{} Smail, I., Ivison, R.J., Blain, A.W., 1997, ApJ, 490, 5

571: \bibitem{} Smail, I., Ivison, R.J., Kneib, J.-P., Cowie, L.L.,

572: 	Blain, A.W., Barger, A.J., Owen, F.N., Morrison, G.,

573: 	1999, MNRAS, 308, 1061

574: \bibitem{} Smail, I., et al., 2000, ApJ, 528, 612

575: \bibitem{} Smail, I., Ivison, R.J., Blain, A.W., Kneib, J.-P.,

576: 	2002, MNRAS, in press

577: \bibitem{} Soucail, G., Kneib, J.-P., Bezecourt, J., Metcalfe, L., 

578: 	Altieri, B., LeBorgne, J.F., 1999, A\&A, 343, 70 

579: \bibitem{} Steidel, C.C., Adelberger, K.L., Dickinson, M., Giavalisco, M., Pettini, M., Kellogg, M., 1998, ApJ, 492, 428

580: \bibitem{} Surdej, J., et al., 1993, AJ, 105, 2064

581: \bibitem{} Webb, T., et al., 2002, ApJ, submitted, (astro-ph/0201180)

582: \bibitem{} Yun, M., Carilli, C., 2002, in press, (astro-ph/0112074)

583: \end{thebibliography}

584: 

585: \end{document}

586: