Quasars in the MAMBO blank field survey

a r X i v :a s t r o -p h /0511117v 1 4 N o v 2005

Astronomy &Astrophysics manuscript no.3712February 5,2008

(DOI:will be inserted by hand later)

Quasars in the MAMBO blank ?eld survey ?

Hauke V oss 1,Frank Bertoldi 1,2,Chris Carilli 3,Frazer N.Owen 3,Dieter Lutz 4,Mark Holdaway 3,Michael

Ledlow 5??,and Karl M.Menten 1

1Max-Planck Institut f¨u r Radioastronomie,Auf dem H¨u gel 69,D-53121Bonn,Germany 2Radioastronomisches Institut der Universit¨a t Bonn,Auf dem H¨u gel 71,D-53121Bonn,Germany 3National Radio Astronomy Observatory,P.O.Box,Socorro,NM 87801,USA 4Max-Planck Institut f¨u r extraterrestrische Physik,Giessenbachstra?e,D-85748Garching,Germany 5

Gemini Observatory,Southern Operations Center,AURA,Casilla 603,La Serena Chile

Received date /Accepted date

Abstract.Our MAMBO 1.2mm blank ?eld imaging survey of ～0.75sqd has uncovered four unusually bright sources,with

?ux densities between 10and 90mJy,all located in the Abell 2125?eld.The three brightest are ?at spectrum radio sources with bright optical and X-ray counterparts.Their mm and radio ?ux densities are variable on timescales of months.Their X-ray luminosities classify them as quasars.The faintest of the four mm bright sources appears to be a bright,radio-quiet starburst at z ～3,similar to the sources seen at lower ?ux densities in the MAMBO and SCUBA surveys.It may also host a mildly obscured AGN of quasar-like X-ray luminosity.The three non-thermal mm sources imply an areal density of ?at spectrum radio sources higher by at least 7compared with that expected from an extrapolation of the lower frequency radio number counts.

Key words.Galaxies:active –Galaxies:high redshift –Galaxies:starburst –Galaxies:individual:MMJ1540+6605,

MMJ1541+6630,MMJ1541+6622,MMJ1543+6621–quasars:general –Submillimeter

1.Introduction

Blank ?eld surveys with the SCUBA and MAMBO bolome-ter arrays were able to resolve a good fraction of the extra-galactic far-infrared (FIR)background radiation that was dis-covered by COBE (Puget et al.1996)into point sources,re-

2Hauke V oss et al.:Quasars in the MAMBO blank?eld

survey

J1541+6630 J1541+6622

J1540+6605

J1543+6621

Fig.1.The MAMBO signal to noise250GHz map centered on Abell2125.The rms noise level is below1mJy per11arcsec beam in the inner contour(ca.400arcmin2),and below3mJy within the outer contour(ca.1200arcmin2).

the four brightest sources ever found in mm or submm blank ?eld surveys.

2.Observations

Using the Max-Planck Millimeter Bolometer(MAMBO, Kreysa et al.(1998,2002))1.2mm(250GHz)37-and117-element arrays at the IRAM30m telescope we have im-aged four?elds:the Abell2125?eld covers an area of about1600arcmin2(Fig.1),the COSMOS?eld covers ～500arcmin2,the Lockman Hole covers～400arcmin2and the NTT deep?eld covers～200arcmin2,adding up to 2700arcmin2imaged to an rms noise level ranging from0.5 to3mJy per11arcsec beam.More than60signi?cant(>4σ) sources have been detected,most of which show a faint(< 100μJy)1.4GHz radio counterpart(Bertoldi et al.,in prepa-ration).

We discovered four previously unknown sources with S1.2mm 10mJy in this MAMBO survey(Tables1and2). They are all located in the Abell2125?eld,which is cen-tered on the di?use,low density galaxy cluster Abell2125 at z=0.25(Owen et al.1999).Their mm?ux densities were monitored with MAMBO pointed observations,show-ing considerable variability on timescales of months for the three brighter sources MMJ1540+6605,MMJ1541+6622and MMJ1543+6621(Fig.2).Those were also observed simulta-neously at1mm and3mm with the Plateau de Bure interfer-ometer(PdBI),showing them to have spectra declining with frequency(Fig.4).

Deep(6.5μJy rms)VLA20cm images(Owen et al. 2005a)show likely radio counterparts to all four sources, providing positions to about0.5arcsec accuracy.Short ob-Fig.2.250GHz?ux density measurement of the four brightest sources in the MAMBO?elds,plotted as a function of time. The measurements are slightly o?set horizontally for clarity. servations were also made a4.86and8.46GHz in August 2000with the VLA D con?guration,which are reported by Owen et al.(2005b),and324MHz observations were made in April2002(Owen&Clarke,private communication). We performed near-simultaneous cross-scan observations of MMJ1540+6605,MMJ1541+6622and MMJ1543+6621with the E?elsberg100m telescope at 4.85,8.35,10.45and 14.6GHz on March24th,2005,and cross-scans at1.41GHz were taken at E?elsberg two month earlier.These near-simultaneous observations show?at radio spectra.Further cross-scan measurements were taken at4.85and10.45GHz nine and eight month earlier,showing radio?uxes that are vari-able on timescales of months for all three sources.

We found1.4GHz?ux measurements for the three brighter sources also in the NVSS catalog(Condon et al.2002),for two of them at325MHz in the WENSS catalog(Rengelink et al. 1997),and at5GHz in the87GB catalog(Gregory&Condon 1992).

All four sources are covered by our mapping of the Abell2125MAMBO?eld with the Spitzer Space Telescope, using the Multiband Imaging Photometer for Spitzer(MIPS)in photometry mode at24μm(rms～11μJy)and the Infrared Array Camera(IRAC)at3.6μm(rms～0.5μJy)and5.8μm (rms～3μJy).Three of the sources are also covered by the IRAC4.5and8.0μm(rms～0.6and2.5μJy)paral-lel observations.The basic calibrated data have been further reduced using the mopex software package.A median?lter was applied to substract the background from the images.The source?ux densities were measured by?tting a point response function that was determined from bright sources in the?eld. MMJ1540+6605,MMJ1541+6622and MMJ1543+6621were detected in all bands,while MMJ1541+6630was not detected in any band.

Spectroscopic redshifts for MMJ1540+6605and MMJ1541+6630are not https://www.wendangku.net/doc/c03935752.html,ler et al.(2004) report that MMJ1541+6622has a redshift of1.382based on the C III]and Mg II lines.These broad lines indicate that the object is a quasar.

Hauke V oss et al.:Quasars in the MAMBO blank ?eld survey

3

Fig.3.The radio-to-IR SED of MMJ1541+6630.The dotted line shows a greybody with T dust =50K and β=1.5and a synchrotron spectrum with α=0.7at z =3,whose radio-to mm ?ux

ratio is set according to the correlation found in local galaxies (Carilli &Yun 1999).The solid line is the SED of Arp 220shifted to z =3.5and scaled to match the radio and mm measurements.

A long slit spectrum of MMJ1543+6605was obtained us-ing Gemini-N and GMOS on June 30th ,2003for a total expo-sure of 30minutes.The B600grating was used which produced a spectrum with about 3.3?resolution.The spectrum shows one strong wide emission line (Fig.5)with a number of associ-ated pairs of absorption lines.This combination and the wave-length separation of the absorption doublets clearly indicates that the line is Mg II.The redshift we derive is 1.2362.Also the low level emission near 6600?is very consistent with the Fe II emission line complex at the same redshift.

We estimate a photometric redshift of 3+1.9

?1.2for MMJ1541+6630(Fig.3),using the correlation between the radio and FIR luminosities found in the local star-forming galaxies (Condon 1992;Carilli &Yun 1999,2000).

We found X-ray counterparts for MMJ1540+6605,MMJ1541+6622and MMJ1543+6621in the ROSAT 1WGA catalog (White et al.1994).MMJ1541+6630is detected in a 20ksec ROSAT PSPC observation.MMJ1540+6605is de-tected in a 83ksec archival Chandra observation (Fig.6;Wang et al.(2004)).The 0.5-10keV ?ux measured by Chandra is (1.1±0.1)·10?13erg cm ?2s ?1.

The o ?sets of all identi?cations to the VLA positions are less than the 1σpositional error quoted in the catalogs,ex-cept for the NVSS counterpart of MMJ1543+6621and the ROSAT counterpart of MMJ1541+6630,for which the o ?sets are within the 2σpositional error.

Fig.4.The radio-to-IR SEDs of the three brightest MAMBO sources,arbitrarily scaled for clarity.The dashed lines connect the simultaneous PdBI measurements (90and 230GHz).The solid lines connect near-simultaneous E ?elsberg measurements (below 20GHz).The E ?elsberg measurements were taken dur-ing the same day,except for the 1.4GHz measurements,which were taken two month earlier.For J1540+6605measurements below 8GHz were not taken because of confusion with a neigh-bouring source.

10 e r g s c m H z

?17 ?1 ?2 ?15800 6000 6200 6400 6600 6800

8

64

57Fig.5.The GMOS /Gemini-N spectrum of MMJ1543+6605,showing the Mg II line at z =1.2362.

3.Discussion

To quantify the signi?cance of our source identi?cations,we calculated the corrected Poissonian probability,P ,of a chance association (Browne &Cohen 1978;Downes et al.1986).For the identi?cations from the NVSS,WENSS and 87GB cata-logs,P <0.1%,so that the found counterparts correspond to the MAMBO sources with >99.9%con?dence.For the

4Hauke V oss et al.:Quasars in the MAMBO blank?eld survey

Table1.The radio to mm source properties of the four brightest MAMBO sources

Source(J2000)S0.3S1.4S4.9S8S10S15S95S250GHz zα230

95

[mJy][mJy][mJy][mJy][mJy][mJy][mJy][mJy]

Origin of the measurements:0.3GHz:VLA&Westerbork;1.4&8GHz:VLA,E?elsberg;4.9GHz:E?elsberg,VLA&Green Bank300ft; 10&15GHz:E?elsberg;95GHz:PdBI;250GHz:PdBI&MAMBO

The spectral indexαis de?ned as Sν∝ν?αin this paper.

Table2.The mid-infrared to X-ray properties of the four brightest MAMBO sources

Source(J2000)S24S8S5.8S4.5S3.6μm R0.1–2.4keV Hardness L0.1?2.4keV [μJy][μJy][μJy][μJy][μJy][mag][erg cm?2s?1][erg s?1]

Origin of the measurements:24μm-3.6μm:Spitzer MIPS&IRAC;R:KPNO;0.1–2.4keV:ROSAT PSPC

Hardness:ROSAT HR1is de?ned as(M+J?C)/(M+J+C)where M,J,C are the countrates in the0.5–1.3,1.5–2.1and0.1–0.3keV bands. For the luminosities we assumed?M=0.27,?Λ=0.73and H0=71km s?1Mpc?1.For MMJ1540+6605we assumed a redshift of0.5,which is suggested by optical photometry.

ROSAT identi?cations the con?dence is93%,83%,98%and 99%for MMJ1540+6605,MMJ1541+6630,MMJ1541+6622, and MMJ1543+6621,respectively.

As the absorption of X-rays by hydrogen decreases with increasing photon energy,the hardness ratios can be used for rough estimates of the absorbing column densities.Three of the sources show absorption by N H～1?4·1020cm?2,com-parable to the galactic value of2.4·1020cm?2toward the Abell2125?eld estimated from the Leiden/Dwingeloo survey (Hartmann&Burton1997).For MMJ1540+6605we estimate N H≈(2±0.4)·1021cm?2,which hints at intrinsic absorption.

3.1.MMJ1541+6630:A thermal source

MMJ1541+6630has a faint1.4GHz radio counterpart with a ?ux density typical of the MAMBO or SCUBA galaxies.The 1.4GHz?ux density measurement and the upper limits at4.9 and8GHz indicate a falling radio spectrum,consistent with a typical synchrotron spectrum withα～0.7(Sν∝ν?α).The ratio between the1.4and250GHz?ux densities is consis-tent with a typical starburst galaxy SED at z～ 3.No opti-cal counterpart could be identi?ed down to an R magnitude of 27.2,which is not unusual for even the brightest mm galax-ies,e.g.MMJ120456-0741.5(Dannerbauer et al.2002,2004). The standard interpretation of these data is that this object is a luminous high redshift starburst.

In starforming galaxies in which no AGN contributes sig-ni?cantly to the X-ray emission the radio and the FIR luminos-ity are also found to be well correlated with the X-ray luminos-ity(Ranalli2004).The X-ray source close to the radio position of MMJ1541+6630is detected at a low signal to noise ratio,making the position determination uncertain.The con?dence of an association of the X-ray source and MMJ1541+6630is 83%.In the following we assume that the ROSAT source is associated with MMJ1541+6630.The inferred X-ray luminos-ity is～80times higher than that expected from the correla-tions between X-ray,radio and mm luminosities for starform-ing galaxies.This X-ray excess reveals the presence of a radio-quiet AGN,which is also not unusual for(sub)mm galaxies: Alexander et al.(2004)suggest that～40%of the(sub)mm galaxies host AGN,which however contribute less then20% to the IR bolometric luminosity.Following the classi?cation scheme suggested by Hasinger(2003),the implied X-ray lumi-nosity classi?es MMJ1541+6630as a quasar.

Taking the SED of the local ULIRG Arp220,redshifted to z=3.5,and?tting it to the mm and radio measurements of MMJ1541+6630would yield?ux densities in the Spitzer IRAC bands that are3.5to20times higher than the measured IRAC upper limits.A typical ULIRG SED with emission from hotter dust(T>50K)than that in Arp220,located at higher redshifts than3would still match the radio and mm measure-ments(Blain1999)but would yield lower?ux densities in the IRAC bands.It could be that MMJ1541+6630is therefore at z>3.

Scaling the average SED for a radio-quiet quasar from Elvis et al.(1994)to match the observed X-ray?ux,we would expect MMJ1541+6630to have an R magnitude of20.5,much brighter than what is observed.The expected?ux densities in the Spitzer IRAC and MIPS24μm bands would also exceed the measurements by an order of magnitude.Dust obscuration by 6.7magnitudes atλ=0.16μm(rest frame)could ex-plain the optical faintness.Assuming galactic dust properties

Hauke V oss et al.:Quasars in the MAMBO blank ?eld survey 5

this would translate to an A V 2.6.To match the ?ux densities observed in the Spitzer bands,with galactic dust properties A V needs to be of order 50?100,which is not atypical for star-burst galaxies;e.g.in Arp 200A V ～50.Assuming a galactic gas to dust ratio of 150,A V 50would imply a hydrogen col-umn density N H 9·1022cm ?2,which is about two orders of magnitude larger than the value inferred from the X-ray hard-ness.Due to the low signal to noise ratio of the ROSAT mea-surement,the X-ray inferred column density is uncertain by at least a factor of ?ve.Another uncertainty arises from the high redshift,whereby the soft part of the X-ray spectrum,which is most a ?ected by absorption,is shifted out of the ROSAT bandpass,so that much higher column densities would still be consistent with the observed hardness ratio (see e.g.Alexander et al.2004).The di ?erence in inferred column densities might also hint on di ?erent line of sights toward the starburst and the quasar.

Studies of quasars at mm wavelengths show almost no cor-relation

between the optical emission of quasars and their mm emission (Omont et al.2003,and references therein).Radio quiet quasars studied at cm and (sub)mm wavelengths show the same ratio of cm and (sub)mm inferred luminosities as known from starforming galaxies (Petric et al.in preparation;Beelen et al.in preparation).Besides this indirect evidence for a sig-ni?cant,if not dominant dust heating by a starburst in quasars,for the high redshift quasars PSS J2322+1944(Carilli et al.2003),SDSS J114816.64+525150.3(Bertoldi et al.2003)and the Cloverleaf (H1413+117,Wei?et al.2003)a starburst could directly be identi?ed as the dominant dust heating source.The Spitzer measurements of MMJ1541+6630rule out a signi?-cant contribution to the dust heating by an AGN,except if this source were a heavily obscured z ～5quasar with moderate star forming activity.

With these uncertainties in mind,we ?nd the properties of MMJ1541+6630consistent with typical X-ray luminous (sub-)mm galaxies such as those studied by Alexander et al.(2004).In the following we shall assume that MMJ1541+6630is a quasar with its FIR emission dominated by star formation.For a dust temperature of 50K and an emissivity index β=1.5,typical for local starburst galaxies (Dunne &Eales 2001),at z =3MMJ1541+6630would have an infrared lu-minosity of (4±1)·1013L ⊙,making it one of the most lu-minous starbursts known.It would be more luminous in the (sub)mm and X-rays than the type-1quasar SMMJ04135that was recently discovered in SCUBA blank ?elds (Knudsen et al.2003),or the type-2quasar SMMJ02399(Ivison et al.1998;Genzel et al.2003).Note that the relation between 250GHz ?ux density and infrared luminosity is nearly independent of redshift from z ≈0.5to 10.

With a Salpeter initial mass function and a starburst age of ～50Myr (Omont et al.2001)the inferred star formation rate would be 4500M ⊙yr ?1.Such a high star formation rate would imply a very massive object.Adopting a dust emissivity κ125μm =18.75cm 2g ?1(Hildebrand 1983)the inferred dust mass of MMJ1541+6630is ～1·109M ⊙.

The likely detection of a massive starburst associated with a quasar adds to the growing evidence that the growth of central supermassive black holes and the formation of the ?rst stel-

Fig.6.Smoothed ROSAT PSPC map of the Abell 2125?eld with inlayed smoothed Chandra ACIS-I map.The logarithmic grey scale ranges from 1·10?3ct /s to 6·10?3ct /s in the ROSAT image and from 1·10?4ct /s to 8·10?3ct /s in the Chandra inlay.The positions of the four bright MAMBO sources are indicated by squares.

lar populations are closely linked (e.g.Haehnelt &Kau ?mann 2000,and references therein).Today we see the fossil records of this connection in the correlation between black hole and stellar bulge masses in local galaxies (Magorrian et al.1998;Ferrarese &Merritt 2000;Gebhardt et al.2000).Following Omont et al.(2001),for MMJ1541+6630we estimated the ra-tio between star formation rate,˙M

SF ,and black hole growth rate,˙M

BH ,as well as the bulge to black hole mass ratio,M bul /M BH .The former is estimated by relating the FIR lumi-nosity to the bolometric luminosity arising from the quasar,which we calculate from the X-ray luminosity using the av-erage radio-quiet quasar SED from Elvis et al.(1994).For the latter estimate we assume the dust properties mentioned above.

We ?nd ˙M

SF /˙M BH ≈3400.This is ～10times higher than expected if black holes and stars formed simultaneously with the same ratio of bulge mass to black hole mass as observed in

local galaxies.For SMMJ02399we ?nd ˙M

SF /˙M BH ≈2200,still a factor of ～6higher.For M bul /M BH we ?nd ～1200and ～850for MMJ1541+6630and SMMJ02399respectively,a factor of ～2?3above the value found in local galaxies.Given the uncertainties in our M bul /M BH estimate and the scat-ter in M bul /M BH for local galaxies,the estimated ratios are sim-ilar to those in local https://www.wendangku.net/doc/c03935752.html,pared with the optically selected quasars of Omont et al.(2001),the bolometric lumi-nosities of the (sub)mm selected quasars MMJ1541+6630and SMMJ02399appear to be dominated by the starburst activity.

6Hauke V oss et al.:Quasars in the MAMBO blank?eld survey 3.2.Flat spectrum sources

The?at and time-variable radio to mm SEDs of the three

brightest sources,their relatively bright optical counterparts

and their high X-ray luminosities indicate that these sources are

?at spectrum,radio loud quasars.The broad emission lines in

the optical spectrum of MMJ1541+6622also indicate that this

source is a quasar.The Spitzer SEDs of MMJ1541+6622and

MMJ1543+6621

can well be?t with the average SEDs of ra-

dio loud quasars(Elvis et al.1994)shifted to the spectroscop-

ically measured redshifts.The95to230GHz spectral index

measured with the PdBI also suggests a non-thermal spectrum,

which is however steeper by～0.5–1than the radio spectrum

between1and95GHz.The fact that the radio spectrum is?at

up to95GHz indicates that the steepening occurs very close

to this frequency.A similar mm-break is also seen e.g.in3

quasars selected from the3C catalog(van Bemmel&Bertoldi

2001).

Flat spectrum radio sources(FSS)are well studied at ra-

dio wavelengths,but little is known about their SEDs at fre-

quencies above90GHz and below the MIR.To see how the

mm bright FSS relate to the radio selected FSS,we used the

90GHz observations of Holdaway et al.(1994),as interpreted

by the model of Holdaway&Owen(2005),to extrapolate the

source counts up to250GHz.This model uses90GHz obser-

vations of?at spectrum quasars selected at cm wavelengths to

sample the spectral index distribution of the sources’compact

cores from8.4GHz to90GHz.To push the model to frequen-

cies above90GHz,Holdaway and Owen determined plausible

distributions of turnover frequencies in these FSS,assuming

the spectral index turns over to0.8above these frequencies.

The sampling of the spectral index distribution of the FSS

population up to90GHz builds on the work of Patnaik et al.

(1992),who selected sources with S5GHz>200mJy and

1.4to5GHz spectral indexα<0.5.The 1.4GHz

data(Condon&Broderick1985,1986)and the5GHz data

(Condon et al.1989)that de?ned their survey were taken with

the NRAO91m telescope.Patnaik et al.(1992)observed their

FSS sample with the VLA A array at8.4GHz.Holdaway et al.

(1994)de?ned a subsample of Patnaik’s sample,selecting367

sources with8.4GHz core?uxes ranging from200mJy to

35Jy.Holdaway et al.(1994)observed these sources with the

NRAO12m telescope.As the8.4GHz observations with0.2

arcsecond resolution provided excellent core?uxes,and from

8.4GHz to90GHz,the steep spectrum emission is decreased

by a factor between10and100,a good spectral index distri-

bution for the FSS cores between8.4GHz and90GHz was

obtained.No evidence was found for a spectral index distribu-

tion depending on core?ux.

Condon(1984)presents source counts of FSS at5GHz,

complete down to10mJy,from a variety of single dish ob-

servations.These counts,which included contributions from

both a?at spectrum core and steep spectrum extended emis-

sion,were reduced by a mean core fraction of0.8and then

scaled up to90GHz with the8.4GHz to90GHz spectral in-

dex distribution.To scale the FSS counts to other frequencies,

we need a model for how the spectral index varies at higher fre-

quency.Such a model is hinted at by the distribution of spectral

Fig.7.The cumulative source counts based on the MAMBO

and SCUBA surveys.The rightmost three data points corre-

spond to the FSS and were not included in the counts below

10mJy.The lines show the model?t(solid)of the thermal

population’s counts and its1σ-error(dashed).The dotted lines

represent the extrapolation of the5GHz FSS counts using the

model of Holdaway&Owen(2005).

index between8.4GHz and90GHz,which can be decom-

posed into two Gaussian components,indicating that81%of

the observed sources have“turned over”to a steep spectral in-

dex by90GHz,and19%of the observed sources remain?at at

90GHz.By combining the theoretical knowledge that?at spec-

trum quasars have a spectral index of about0.0,and that steep

spectrum,optically thin synchrotron emission has a spectral in-

dex of about0.8,and by making reasonable assumptions on the

spread about those spectral index values,Holdaway&Owen

(2005)were able to solve for a distribution of turn over fre-

quencies.This distribution of turn over frequencies can then be

combined with the well-studied5GHz FSS counts and the core

fraction correction to estimate the FSS counts at any frequency

above8.4GHz.

One thing remains undetermined in this model:what hap-

pens to those19%of still-?at spectrum sources above90GHz?

An upper limit to source counts at250GHz can be obtained

by assuming that these sources remain?at spectrum through

250GHz.A lower limit can be obtained by assuming they all

turn over right at90GHz.A“best guess”can be obtained by

continuing the best?t power law distribution of turn over fre-

quencies up to181GHz,at which point all of the remaining

?at spectrum sources have turned over.

The extrapolated number counts are shown in Fig.7,with

no spectral steepening up to250GHz and with the“best guess”

model for spectral steepening.The FSS source counts from the

MAMBO?elds are a factor of9+33

?3

or7+30

?3

higher than pre-

dicted from the extrapolation with and without steepening,re-

Hauke V oss et al.:Quasars in the MAMBO blank?eld survey7

spectively.Lensing ampli?cation is unlikely to be a cause for the observed source overdensity since the four sources are far from the inner2×2arcmin region of the galaxy cluster where mild lensing ampli?cation could be expected.It could be that the number of FSS in the MAMBO?eld is high by chance. With Monte Carlo methods we simulated a?eld of1000square degrees with source counts that are consistent with the model and sources at random positions.We took1600independent small?elds of the MAMBO survey size from the large sim-ulation and analysed the distribution of source counts for the small?elds.We found that less than0.4%of these arti?cial ?elds contain three or more sources brighter than10mJy.This discrepancy might also be due to a deviation from the found spectral index distribution for sources below the90GHz detec-tion limit of75mJy.Although the spectral index distribution was found not to depend on radio?ux density,this is based on the brighter sources(S5GHz>200mJy)that were also detected at90GHz.It is possible that some fainter sources have?atter or more inverted spectra and thus contribute noticeably to the mm number counts.

The four sources are among the?rst quasars to be discov-ered in mm or submm blank?eld surveys.Although based on small number statistics,they allow a?rst study of the contribu-tion of non-thermal sources in(sub)mm observations.Figure7 shows the cumulative source counts for the MAMBO and SCUBA deep?eld surveys.We have?tted the counts with a model of a population of dust enshrouded starburst galaxies. The model is based on the Saunders60μm luminosity function of local galaxies(Saunders et al.1990)which consists of a?at power law for low luminosities and a steep power law for lu-minosities above～109L⊙.An additional exponential cuto?at ～1012L⊙,and a luminosity and density evolution with redshift were applied to match the(sub)mm number counts and their redshift distribution.The model was found to be consistent with the SED of the extragalactic FIR background.Including MMJ1541+6630in the number counts,we?nd that they are in good agreement with the extrapolation of the fainter source counts.The three brightest sources however indicate a?atten-ing in the source counts at250GHz?ux densities>10mJy. The mm background appears to change from a declining pop-ulation of high redshift starbursts to a shallower distribution of less distant,radio loud quasars.

Acknowledgements.We thank J¨u rgen Kerp from RAIUB and Thomas Boller from MPE Garching for their help with the ROSAT data and Matthias Kadler from the MPIfR for his help with the Chandra data. We thank Endrik Kr¨u gel from MPIfR for his useful comments on optical dust properties,Emmanouil Angelakis and Alex Kraus from the MPIfR for their help on E?elsberg observations,and the IRAM sta?for their great support of this https://www.wendangku.net/doc/c03935752.html, acknowledges support from the Max-Planck-Forschungspreis,granted by the Alexander von Humboldt Foundation and the Max-Planck Society.HV is member of the International Max Planck Research School(IMPRS)for Radio and Infrared Astronomy.

This research has made use of the NASA/IPAC Extragalactic Database(NED)which is operated by the Jet Propulsion Laboratory, California Institute of Technology,under contract with the NASA, and of data obtained through the High Energy Astrophysics Science Archive Research Center Online Service,provided by the NASA/Goddard Space Flight Center.References

Alexander,D.M.,Bauer,F.E.,Chapman,S.C.,et al.2004, arXiv:astro-ph/0401129

Barger,A.J.,Cowie,L.L.,Sanders,D.B.,et al.1998,Nature, 394,248

Bertoldi,F.,Carilli,C.L.,Menten,K.M.,et al.2000,A&A, 360,92

Bertoldi,F.,Cox,P.,Neri,R.,et al.2003,A&A,409,L47 Bertoldi,F.,Menten,K.M.,Kreysa,E.,Carilli,C.L.,&Owen, F.2002,Highlights in Astronomy,12,473

Blain,A.W.1999,MNRAS,309,955

Blain,A.W.,Smail,I.,Ivison,R.J.,Kneib,J.-P.,&Frayer, D.T.2002,Phys.Rep.,369,111

Browne,I.W.A.&Cohen,A.M.1978,MNRAS,182,181 Carilli,C.L.,Lewis,G.F.,Djorgovski,S.G.,et al.2003, Science,300,773

Carilli,C.L.&Yun,M.S.1999,ApJ,513,L13

Carilli,C.L.&Yun,M.S.2000,ApJ,530,618 Chapman,S.C.,Blain,A.W.,Smail,I.,&Ivison,R.J.2005, ApJ,622,772

Condon,J.J.1984,ApJ,287,461

Condon,J.J.1992,ARA&A,30,575

Condon,J.J.&Broderick,J.J.1985,AJ,90,2540 Condon,J.J.&Broderick,J.J.1986,AJ,91,1051 Condon,J.J.,Broderick,J.J.,&Seielstad,G.A.1989,AJ,97, 1064

Condon,J.J.,Cotton,W.D.,Greisen,E.W.,et al.2002,VizieR Online Data Catalog,8065,0

Dannerbauer,H.,Lehnert,M.D.,Lutz,D.,et al.2002,ApJ, 573,473

Dannerbauer,H.,Lehnert,M.D.,Lutz,D.,et al.2004,ApJ, 606,664

Downes,A.J.B.,Peacock,J.A.,Savage,A.,&Carrie,D.R. 1986,MNRAS,218,31

Dunne,L.&Eales,S.A.2001,MNRAS,327,697

Eales,S.,Bertoldi,F.,Ivison,R.,et al.2003,MNRAS,344, 169

Elvis,M.,Wilkes,B.J.,McDowell,J.C.,et al.1994,ApJS,95, 1

Ferrarese,L.&Merritt,D.2000,ApJ,539,L9

Gebhardt,K.,Bender,R.,Bower,G.,et al.2000,ApJ,539,L13 Genzel,R.,Baker,A.J.,Tacconi,L.J.,et al.2003,ApJ,584, 633

Gregory,P.C.&Condon,J.J.1992,VizieR Online Data Catalog,8014,0

Greve,T.R.,Bertoldi,F.,Smail,I.,et al.2005,MNRAS,359, 1165

Greve,T.R.,Ivison,R.J.,Bertoldi,F.,et al.2004,MNRAS, 354,779

Haehnelt,M.G.&Kau?mann,G.2000,MNRAS,318,L35 Hartmann,D.&Burton,W.B.1997,Atlas of galactic neu-tral hydrogen(Cambridge;New York:Cambridge University Press,ISBN0521471117)

Hasinger,G.2003,arXiv:astro-ph/0310804

Hildebrand,R.H.1983,QJRAS,24,267+

Holdaway,M.A.&Owen,F.N.2005,ALMA memo,520 Holdaway,M.A.,Owen,F.N.,&Rupen,M.P.1994,MMA

8Hauke V oss et al.:Quasars in the MAMBO blank?eld survey memo,123

Hughes,D.H.,Serjeant,S.,Dunlop,J.,et al.1998,Nature,394,

241

Ivison,R.J.,Smail,I.,Le Borgne,J.-F.,et al.1998,MNRAS,

298,583

Knudsen,K.K.,van der Werf,P.P.,&Ja?e,W.2003,A&A,

411,343

Kreysa,E.,Gemuend,H.,Gromke,J.,et al.1998,in Proc.SPIE

V ol.3357,p.319-325,Advanced Technology MMW,Radio,

and Terahertz Telescopes,Thomas G.Phillips;Ed.,319–325

Kreysa,E.,Gem¨u nd,H.-P.,Raccanelli,A.,Reichertz,L.A.,

&Siringo,G.2002,in AIP Conf.Proc.616:Experimental

Cosmology at Millimetre Wavelengths,262–269

Magorrian,J.,Tremaine,S.,Richstone,D.,et al.1998,AJ,115,

2285

Miller,N.A.,Owen,F.N.,Hill,J.M.,et al.2004,ApJ,613,

841

Neri,R.,Genzel,R.,Ivison,R.J.,et al.2003,ApJ,597,L113

Omont,A.,Beelen,A.,Bertoldi,F.,et al.2003,A&A,398,857

Omont,A.,Cox,P.,Bertoldi,F.,et al.2001,A&A,374,371

Owen,F.N.,Keel,W.C.,Ledlow,M.J.,Morrison,G.E.,&

Windhorst,R.A.2005a,AJ,129,26

Owen,F.N.,Keel,W.C.,Wang,Q.D.,Ledlow,M.J.,&

Morrison,G.E.2005b,in preperation

Owen,F.N.,Ledlow,M.J.,Keel,W.C.,&Morrison,G.E.

1999,AJ,118,633

Patnaik,A.R.,Browne,I.W.A.,Wilkinson,P.N.,&Wrobel,

J.M.1992,MNRAS,254,655

Puget,J.-L.,Abergel,A.,Bernard,J.-P.,et al.1996,A&A,308,

L5+

Ranalli,P.2004,arXiv:astro-ph/0407140

Rengelink,R.B.,Tang,Y.,de Bruyn,A.G.,et al.1997,A&AS,

124,259

Saunders,W.,Rowan-Robinson,M.,Lawrence,A.,et al.1990,

MNRAS,242,318

Smail,I.,Ivison,R.J.,&Blain,A.W.1997,ApJ,490,L5+

van Bemmel,I.M.&Bertoldi,F.2001,A&A,368,414

Wang,Q.D.,Owen,F.,&Ledlow,M.2004,ApJ,611,821

Wei?,A.,Henkel,C.,Downes,D.,&Walter,F.2003,A&A,

409,L41

White,N.E.,Giommi,P.,&Angelini,L.1994,Bulletin of the

American Astronomical Society,26,1372