Planck intermediate results. VII. Statistical properties of infrared and radio extragalactic sources from the Planck Early Release Compact Source Catalogue at frequencies between 100 and 857 GHz

2.6 MB·English

by  Planck Collaboration

Astronomy&Astrophysicsmanuscriptno.counts˙planck c ESO2012
(cid:13)
July20,2012
Planck intermediate results. VII. Statistical properties of infrared
and radio extragalactic sources from the Planck Early Release
Compact Source Catalogue at frequencies between 100 and
857GHz
PlanckCollaboration:P.A.R.Ade84,N.Aghanim59,F.Argu¨eso19,M.Arnaud73,M.Ashdown70,5,F.Atrio-Barandela17,J.Aumont59,
C.Baccigalupi83,A.Balbi36,A.J.Banday88,8,R.B.Barreiro67,E.Battaner90,K.Benabed60,87,A.Benoˆıt57,J.-P.Bernard8,M.Bersanelli33,50,
M.Bethermin73,R.Bhatia6,A.Bonaldi69,J.R.Bond7,J.Borrill12,85,F.R.Bouchet60,87,C.Burigana49,35,P.Cabella37,J.-F.Cardoso74,1,60,
2 A.Catalano75,72,L.Cayo´n30,A.Chamballu55,R.-R.Chary56,X.Chen56,L.-YChiang63,P.R.Christensen80,38,D.L.Clements55,
1 S.Colafrancesco46,S.Colombi60,87,L.P.L.Colombo22,68,A.Coulais72,B.P.Crill68,81,F.Cuttaia49,L.Danese83,R.J.Davis69,P.deBernardis32,
0 G.deGasperis36,G.deZotti45,83,J.Delabrouille1,C.Dickinson69,J.M.Diego67,H.Dole59,58,S.Donzelli50,O.Dore´68,9,U.Do¨rl78,
2 M.Douspis59,X.Dupac41,G.Efstathiou64,T.A.Enßlin78,H.K.Eriksen65,F.Finelli49,O.Forni88,8,P.Fosalba61,M.Frailis47,E.Franceschi49,
S.Galeotta47,K.Ganga1,M.Giard88,8,G.Giardino42,Y.Giraud-He´raud1,J.Gonza´lez-Nuevo67,83,K.M.Go´rski68,91,A.Gregorio34,47,
l
u A.Gruppuso49,F.K.Hansen65,D.Harrison64,70,S.Henrot-Versille´71,C.Herna´ndez-Monteagudo11,78,D.Herranz67,S.R.Hildebrandt9,
J E.Hivon60,87,M.Hobson5,W.A.Holmes68,T.R.Jaffe88,8,A.H.Jaffe55,T.Jagemann41,W.C.Jones25,M.Juvela24,E.Keiha¨nen24,

9 T.S.Kisner77,R.Kneissl40,6,J.Knoche78,L.Knox27,M.Kunz16,59,N.Kurinsky23,H.Kurki-Suonio24,44,G.Lagache59,A.La¨hteenma¨ki2,44,
1 J.-M.Lamarre72,A.Lasenby5,70,C.R.Lawrence68,R.Leonardi41,P.B.Lilje65,10,M.Lo´pez-Caniego67,J.F.Mac´ıas-Pe´rez75,D.Maino33,50,
N.Mandolesi49,4,M.Maris47,D.J.Marshall88,8,E.Mart´ınez-Gonza´lez67,S.Masi32,M.Massardi48,S.Matarrese31,P.Mazzotta36,
] A.Melchiorri32,51,L.Mendes41,A.Mennella33,50,S.Mitra54,68,M.-A.Miville-Descheˆnes59,7,A.Moneti60,L.Montier88,8,G.Morgante49,
O
D.Mortlock55,D.Munshi84,J.A.Murphy79,P.Naselsky80,38,F.Nati32,P.Natoli35,3,49,H.U.Nørgaard-Nielsen15,F.Noviello69,S.Osborne86,
C
F.Pajot59,R.Paladini56,D.Paoletti49,B.Partridge43,F.Pasian47,G.Patanchon1,O.Perdereau71,L.Perotto75,F.Perrotta83,F.Piacentini32,
h. M.Piat1,E.Pierpaoli22,S.Plaszczynski71,E.Pointecouteau88,8,G.Polenta3,46,N.Ponthieu59,52,L.Popa62,T.Poutanen44,24,2,G.W.Pratt73,
p S.Prunet60,87,J.-L.Puget59,J.P.Rachen20,78,W.T.Reach89,R.Rebolo66,13,39,M.Reinecke78,C.Renault75,S.Ricciardi49,T.Riller78,
- I.Ristorcelli88,8,G.Rocha68,9,C.Rosset1,M.Rowan-Robinson55,J.A.Rubin˜o-Mart´ın66,39,B.Rusholme56,A.Sajina23,M.Sandri49,G.Savini82,
o D.Scott21,G.F.Smoot26,77,1,J.-L.Starck73,R.Sudiwala84,A.-S.Suur-Uski24,44,J.-F.Sygnet60,J.A.Tauber42,L.Terenzi49,L.Toffolatti18,67,
r
t M.Tomasi50,M.Tristram71,M.Tucci71,M.Tu¨rler53,L.Valenziano49,B.VanTent76,P.Vielva67,F.Villa49,N.Vittorio36,L.A.Wade68,
s B.D.Wandelt60,87,29,M.White26,D.Yvon14,A.Zacchei47,andA.Zonca28
a
[ (Affiliationscanbefoundafterthereferences)

1
Submitted19-Jul-2012/Accepted??
v
6
0
Abstract
7
4
We make use of the Planck all-sky survey to derive number counts and spectral indices of extragalactic sources – infrared and radio sources
.
7 – from the Planck Early Release Compact Source Catalogue at 100 to 857GHz (3mm to 350µm). Three zones (deep, medium and shallow)
0 of approximately homogeneous coverage areused toensure acleancompleteness correction. Our sample, prior tothe80%completeness cut,
2 contains between 217 sources at 100GHz and 1058 sources at 857GHz over about 12,800 to 16,550deg2 (31 to 40% of the sky). After the
1 80%completenesscut,between122and452andsourcesremain,withfluxdensitiesabove0.3and1.9Jy.Usingthemulti-frequencycoverage
: ofthePlanckHighFrequencyInstrument,allthesourceshavebeenclassifiedintodust-dominated(infraredgalaxies)orsynchrotron-dominated
v (radiogalaxies)spectralenergydistributions(SED).Wefindanapproximatelyequalnumberofsynchrotronanddustysourcesbetween217and
i 353GHz;at353GHzorhigher(respectively217GHzorlower)frequencies,thenumberisdominatedbydusty(synchrotron)sources,asexpected.
X
Formostofthesources,thespectralindicesarealsoderived.Weprovidebrightcountsfrom353to857GHzandthecontributionsfromdusty
r andsynchrotronsourcesatallHFIfrequencies, inthekeyspectral rangewherethesespectraarecrossing,forthefirsttimeinacoherent way.
a
The observed counts are inthe Euclidean regime. The number counts are compared topreviously published data (from earlier Planck results,
Herschel,BLAST,SCUBA,LABOCA,SPT,andACT)andmodels.Wederivethemulti-frequencyEuclideanlevel–theplateauinthenormalised
differential counts at high flux-density – and compare it to WMAP, Spitzer and IRAS results. The submillimetre number counts are not well
reproducedbycurrentevolutionmodelsofdustygalaxies,whereasthemillimetrepartappearsreasonablywellfittedbythemostrecentmodelfor
synchrotron-dominatedsources.Finallyweprovideestimatesofthelocalluminositydensityofdustygalaxies,providingthefirstmeasurements
at545and857GHz.
Keywords.Cosmology:observations–Surveys–Galaxies:statistics-Galaxies:evolution–Galaxies:starformation–Galaxies:active–Radio
continuum:galaxies–Submillimetre:galaxies
1. Introduction reasonto probebrightobjectsisto studythe numbercountsof
extragalacticsourcesandtheirspectralshapes.Inthefar-infrared
Among other advantages, all-sky multifrequency surveys have (FIR) the sources detected by these surveys are usually domi-
thebenefitofprobingrareand/orbrightobjectsinthesky.One
2 PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz
Figure1. Comparison between our Planck 857GHz mask (red Figure2. Comparison between our Planck 353GHz mask (red
means removed from the analysis) and WMAP 7-year KQ75 means removed)and WMAP 7-year KQ85 mask (green means
mask (green means removed); unlike the case for the mask removed). The background (blue) is the sky area used for the
employed in this paper, the WMAP mask excludes some point analysis.
sources.Thebackground(blue)istheskyareausedforouranal-
ysis.ThismapisaMollweideprojectionoftheskyinGalactic
coordinates. The Planck1 all-sky survey covers nine bands between 30
and 857GHz. It gives us for the first time robust extragalactic
countsoverawideareaofskyatthesewavelengths,andthefirst
all-skycoveragebetween3mm(WMAP)and160µm (Akari)–
e.g.seeTable1ofPlanckCollaborationVII(2011).Thecounts
giveuspowerfulconstraintsonthelong-wavelengthspectralen-
ergy distribution (SED) of the dusty galaxies investigated e.g.
natedbylow-redshiftgalaxieswithz < 0.1,e.g.IRASat60µm byIRAS,andontheshort-wavelengthSEDoftheactivegalax-
(Ashbyetal.,1996)butalsorevealthepresenceofafewextreme iesstudiedatradiowavelengths,e.g.byWMAPorground-based
objects like the lensed F10214source (Rowan-Robinsonet al., facilities.
1991).However,theradiorangepopulationisdominatedbysyn-
chrotronsources(inparticular,blazars)athigherredshift(seede
Zottietal.,2010,forarecentreview).Previousmultifrequency
all-skysurveyswerecarriedoutintheinfrared(IR)rangebythe
IRASsatellite(between12and100µm;Neugebaueretal.1984),
andmorerecentlybyAkari(between2and180µm;Murakami
et al. 2007) and WISE (between 3.4 and 22µm; Wright et al.
2010). They were carried out in the IR and microwave range
byCOBE(between1.2µmand1cm;Boggessetal.1992),and
more recently in the microwave range by WMAP (between 23
and 94GHz; Bennett et al. 2003;Wright et al. 2009;Massardi
etal.2009;deZottietal.2010).
Atlargefluxdensities(typically1Jyandabove),FIRextra-
galactic numbercounts from all-sky surveysshow a Euclidean
component,i.e. a distribution of the number of source per flux
densitybinS atobservedfrequencyν(dN/dS inJysr 1)pro-
ν ν −
portional to S 2.5 (see Eq. 1) – in line with expectations from
ν−
a local Universe uniformly filled with non-evolving galaxies
(Lonsdale&Hacking,1989;Hacking&Soifer,1991;Bertin&
Dennefeld,1997;Massardietal.,2009;Wrightetal.,2009).In
Figure3.CumulativedistributionofPlanckhitcountsonthesky
the radio range,the Euclideanpartis modifiedby the presence
(hereat857GHzwithNside=2048),withthecorrespondingin-
of higher-redshift sources. At flux densities smaller than typi-
tegrationtimeperskypixel(giveninthetopaxis).Thethreesky
cally 0.1 to 1Jy, an excess in the number counts compared to
zonesused in the analysis are definedat 857GHz: shallow (50
the Euclidean level is an indication of evolution in luminosity
to 75% of the hit count distribution, short-spaced lines, blue);
and/or density of the galaxy populations. This effect is clearly
medium (75 to 95% of the hit counts, medium-spaced lines,
seen in deeper surveys in the FIR – e.g. (Genzel & Cesarsky,
green);anddeep(95%andabovehits,widely-spacedlines,red).
2000;Dole etal., 2001,2004;Frayeretal., 2006;Soifer etal.,
2008;Betherminetal.,2010a)–inthesubmillimetre(submm)
1 Planck (http://www.esa.int/Planck) is a project of the
range–e.g.(Bargeretal.,1999;Blainetal.,1999;Ivisonetal.,
EuropeanSpaceAgency(ESA)withinstrumentsprovidedbytwosci-
2000;Greveetal.,2004;Coppinetal.,2006;Weissetal.,2009;
entificconsortiafunded byESAmember states(inparticularthelead
Patanchonetal.,2009;Clementsetal.,2010;Lapietal.,2011) countriesFranceandItaly),withcontributionsfromNASA(USA)and
– and millimetre and radio ranges – e.g. (de Zotti et al., 2010; telescopereflectorsprovidedbyacollaborationbetweenESAandasci-
Vieiraetal.,2010;Vernstrometal.,2011). entificconsortiumledandfundedbyDenmark.
PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz 3
Figure4. Thethreesky zonesusedin theanalysisat857GHz: Figure5.Thethreesky zonesusedin theanalysisat100GHz:
deep(red);medium(green);andshallow(blue);Thesearebased deep(red);medium(green);andshallow(blue).Thesearebased
onthe857GHzhitcounts. onthe100GHzhitcounts.
Wepresentheretheextragalacticnumbercountsandspectral The Planck data used in this paper (unlike other interme-
indices of galaxies selected at high Galactic latitude and using diate Planck papers) correspond to the first 1.6 sky surveys
identifications, detected at the six highest Planck frequencies, (PlanckCollaborationI,2011;PlanckHFICoreTeam,2011a,b),
between 100and 857GHz. Planck numbercountsand spectral from which the Early Release Compact Source Catalogue, or
indicesofextragalacticradio-selectedsourceswerealreadypub- ERCSC(PlanckCollaborationVII,2011;PlanckCollaboration,
lished for the frequency range 30 to 217GHz using LFI and 2011) has been extracted and documented. First results from
HFIinstruments(PlanckCollaborationXIII,2011).Thetransi- the ERCSC are published as Planck early papers (Planck
tion betweensynchrotron-dominatedsourcesand thermaldust- Collaboration XIII, 2011; Planck Collaboration XIV, 2011;
dominated occurs in the crucial spectral range 200-800GHz . Planck Collaboration XV, 2011; Planck Collaboration XVI,
Thusourbroadfrequencydataallowabetterspectralcharacteri- 2011).WeuseHFIdata,coveringthe100–857GHzrangeinsix
sation,usefulforestimatinge.g.thepointsourcecontamination. bands.
Planckhastheuniquecapabilitytostudybothpopulations,tak-
ingadvantageofitswidespectralandall-skycoverage.
2.1.Galacticmasks
WeusetheWMAP7yearbest-fitΛCDMcosmology(Larson
etal., 2011),with H = 71 kms 1Mpc 1, Ω = 0.734andΩ Toobtainreliableextragalacticnumbercounts,uncontaminated
0 − − Λ M
=0.266. by Galactic sources, we mask out areas of the sky strongly af-
fected by Galactic sources, defining a set of “Galactic masks”.
Thesearebasedonremovingafractionoftheskyaboveaspec-
ified levelin sky surface brightness.We use two masks, one at
2. Planckdata,masksandsources highfrequencies(857and545GHz),andoneatlowerfrequen-
Planck(Tauberetal.,2010;PlanckCollaborationI,2011)isthe
thirdgenerationspacemissiontomeasuretheanisotropyofthe
cosmic microwave background (CMB). It observes the sky in
ninefrequencybandscovering30–857GHzwith highsensitiv-
ity and angular resolution from 31 to 5. The Low Frequency
′ ′
InstrumentLFI;(Mandolesietal.,2010;Bersanellietal.,2010;
Mennellaetal.,2011)coversthe30,44,and70GHzbandswith
amplifierscooledto20K.TheHighFrequencyInstrument(HFI;
Lamarreet al. 2010;Planck HFI Core Team 2011a)coversthe
100, 143, 217, 353, 545, and 857GHz bands with bolometers
cooled to 0.1K. Polarization is measured in all but the highest
twobands(Leahyetal.,2010;Rossetetal.,2010).Acombina-
tionofradiativecoolingandthreemechanicalcoolersproduces
the temperatures needed for the detectors and optics (Planck
Collaboration II, 2011). Two Data Processing Centers (DPCs)
checkandcalibratethe dataandmakemapsofthesky(Planck
HFICoreTeam,2011b;Zaccheietal.,2011).Planck’ssensitiv-
ity, angularresolution, and frequencycoveragemake it a pow-
erful instrument for galactic and extragalactic astrophysics as
wellascosmology.EarlyastrophysicsresultsaregiveninPlanck
Collaboration VIII–XXVI 2011, based on data taken between Figure6. Completeness (vs. flux density of sources) coming
13August2009and7June2010.Intermediateastrophysicsre- fromtheMonte-CarlorunsfortheERCSCandderivedforeach
sultsarenowbeingpresentedinaseriesofpapersbasedondata zone. The horizontal dashed line represents our threshold for
takenbetween13August2009and27November2010. numbercountanalysis.
4 PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz
cies (353GHz and below). The use of two different masks is to the focal plane geometry; each zone will thus have slight
motivated by the different astrophysical components dominat- differences in geometry from one frequency to another, lead-
ingthehigherHFIfrequenciesandthelowerfrequencies,which ing to slightly different surface areas. Table 2 summarizes the
arenotnecessarilyspatiallycorrelated.WhiletheGalacticdust surface area of each zone (deep, medium and shallow); typi-
dominatesat857 GHz,its spectrumdecreaseswith decreasing cally, the deep zone covers 1000deg2, the medium zone about
frequency.On the contrary,the synchrotronand free-freecom- 3000deg2,andtheshallowabout12000deg2.Fig.4(5,respec-
ponentsdominateat100GHz. tively)representsthethreedifferentzonesusedinthisanalysis:
Beforeapplyingabrightnesscuttothemaps,wedegradethe deep,mediumandshallowat857GHz(100GHz,respectively).
angularresolutionofthemapsfrom1.5(N =2048inHealpix;
′ side
Go´rskietal.2005)downto55(N =64).Themapsatlowres-
′ side 2.3.Sampleselectionandvalidation
olutionaretheninterpolatedattheoriginalhighangularresolu-
tion.CreatingaGalacticmaskusingthisprocedurehasthedou- The sample is drawn from the first public catalogue of Planck
ble benefit of:1) not masking the brightsources (because they sources,theERCSC(PlanckCollaborationVII,2011),contain-
aresmoothed);and2)smoothingtheGalacticstructure. inghighlyreliablesources.Theselectionisperformedwiththe
The857GHzGalacticmaskkeeps48%oftheskyforanal- followingstepsateachHFIfrequencyindependently:
ysis(thusremoving52%ofthesky),andthiscorrespondstoa
cutof2.2MJysr−1 at857GHz.Thismaskisappliedat857and – selectsourceswithineachzone:deep,mediumandshallow;
545GHz.The353GHzGalacticmaskkeeps64%oftheskyfor – select point sources, using the keyword “EXTENDED” set
analysis(thusremoving36%ofthe sky),andcorrespondstoa tozero;
cut of 0.28 MJysr 1 at 353GHz.This mask is applied at 353,
−
217,143and100GHz. Thesecriteria shouldfavourthe presenceofgalaxiesrather
Figs.1and2showthePlanckmasks,comparingthemwith thanGalacticsources.Tovalidatethis,wemakethreechecksin
the WMAP 7 year KQ75 and KQ85 masks (Gold et al., 2011; additiontousingconservativemasks.
Jarosiketal.,2011).NoticethatwedonotusetheWMAPmasks 1.We measurethemid-IRvs. submmfluxdensityratiosof
inthiswork. known Galactic cold cores (from the Planck Early Cold Core
catalogue, ECC, Planck Collaboration XXIII 2011) and con-
verselyofknowngalaxiesoftheERCSC.UsingWISE(Wright
2.2.Threezonesinthesky:deep,mediumandshallow
et al., 2010) W3 and W4 bands (when available with the first
Three main zones are identified to ensure homogeneous cov- publicrelease),andthe857GHzHFIband,wemeasureafactor
erage of the sky by the Planck detectors at each frequency, of100to200betweenthesubmm-to-mid-infaredratiosofgalax-
therebyallowingasubsequentestimateofthecompleteness.The iesandECCsources.Whenmeasuringthisratioinoursample,
Planckdatausedherecorrespondtoapproximately1.6complete weseethatthesubmm-to-mid-infaredcoloursofsourcesinour
surveys of the sky and have a non-uniform coverage (Planck sample are compatible with galaxy colours, and not with ECC
CollaborationI,2011;PlanckHFICoreTeam,2011a,b).While colours.
performing statistics on sources drawn from a non-uniformly 2.TheCIRRUSflagintheERCSCgivesanestimateofthe
coveredsurveyisfeasible,thenatureofthePlanckdata(includ- normalisedneighboursurfacedensityofsourcesat857GHz,as
ingscanningstrategy,andmaskingofplanets–seeERCSCar- aproxyforacirrusestimate.ThemedianvalueoftheCIRRUS
ticlePlanckCollaborationVII2011)anditsheterogeneouscov- flaginoursampleis0.093at857GHz,alowvaluecompatible
erage(byafactormorethan103, see Fig.3)–makeitdifficult withnocirruscontaminationwhenusedinconjunctionwiththe
toimplement.Wethereforeselectthreezonesinthesky,ineach EXTENDED=0flag(e.g.Herranzetal.2012).
ofwhichtheobservationsareapproximativelyhomogeneousin 3. We query the NED and SIMBAD databases at the po-
integrationtime. sitions of all our ERCSC sources using a 2.5 radius search.
′
We define the hit count by counting the number of times a Each Planck source has many matches (many of them com-
singlePlanckdetectorobservesoneskypositioninthesky.The pletely unrelated, e.g. foreground stars), and the identification
hitcountcanalsobedefinedforafrequencyband:itisthenum- is more complex at higher frequencies. However, as seen later
beroftimeseachskypixelhasbeenhitbyanyPlanckdetector with the counts estimates, our N(> S) cumulative distribution
atagivenfrequency.Wewillbeusingthislatterdefinition.This ofsourcesisalwayslessthan200sourcespersteradian,i.e.less
quantityissimilartoNobs inWMAPdatafiles. than3.3×10−4ERCSCsourcesper2.′5searchradiusonaverage.
The three zones have hit counts varying by not more than We thus search for the most probable match by identify-
a factor of two, exceptin the smaller deepzone (atthe ecliptic ing the source type in the order Galactic, then extragalactic.
poles)wherethereishighredundancy.Theyaredefinedas(and The Galactic typesincludesupernovaremnants,planetaryneb-
illustratedinFig.3): ulae,nebulae,Hiiregions,stars,molecularclouds,globular/star
clusters.We calla source“Galactic unsecure”whenoneofthe
– deep:95%ormoreofthe cumulativehitcountdistribution two databasesreturnsnoidentificationandtheothera Galactic
atagivenfrequency; identification.Wedonotuse“Galacticsecure”or“Galacticun-
– medium:75to95%ofthecumulativehitcountdistribution secure”sourcesin this paper.The statistics of identificationsis
atagivenfrequency; giveninTable1.
– shallow:50to75%ofthecumulativehitcountdistribution Ourfinalsampleiscomposedofconfirmedgalaxies,thevast
atagivenfrequency. majoritybeingNGC,IRAS,radiosourcesandblazarobjects,as
well as some unidentified sources (a small fraction of the to-
Thus,pixelsinthedeepzone(atagivenfrequency)allhave tal number). These completely unidentified sources, where no
ahitcountvaluegreaterthanorequaltothehitcountvaluecor- SIMBAD or NED ID was found, are interpreted as potential
responding to 95% of the total distribution at this frequency. galaxies,andhenceareincludedinourcounts,becausetheyhave
Notice that each frequency map has different hit counts, due asmallcirrusflagvalue(seepoint2above),
PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz 5
Table1.PercentageofPlancksourceidentificationsusingtheSIMBADandNEDdatabases(withthenumberofsourcesinparen-
theses).
ν Extragalactic Galactic Galactic Unidentified Total
[GHz] Secure%(Nb.) Unsecure%(Nb.) Secure%(Nb.) % (Nb.) % (Nb.)
857 deep 91.2 (73) 2.5 (2) 1.2 (1) 5.0 (4) 100 (80)
857medium 95.5 (255) 1.9 (5) 0.0 (0) 2.6 (7) 100(267)
857shallow 94.6 (697) 1.6 (12) 0.8 (6) 3.0 (22) 100(737)
545 deep 76.5 (39) 7.8 (4) 2.0 (1) 13.7 (7) 100 (51)
545medium 91.1 (143) 2.5 (4) 0.0 (0) 6.4 (10) 100(157)
545shallow 91.8 (301) 2.1 (7) 0.9 (3) 5.2 (17) 100(328)
353 deep 81.6 (31) 2.6 (1) 5.3 (2) 10.5 (4) 100 (38)
353medium 87.0 (94) 0.9 (1) 2.8 (3) 9.3 (10) 100(108)
353shallow 78.0 (170) 4.6 (10) 4.6 (10) 12.8 (28) 100(218)
217 deep 77.3 (17) 0.0 (0) 22.7 (5) 0.0 (0) 100 (22)
217medium 92.5 (62) 1.5 (1) 1.5 (1) 4.5 (3) 100 (67)
217shallow 88.5 (170) 0.5 (1) 4.2 (8) 6.8 (13) 100(192)
143 deep 86.7 (13) 0.0 (0) 13.3 (2) 0.0 (0) 100 (15)
143medium 100.0 (48) 0.0 (0) 0.0 (0) 0.0 (0) 100 (48)
143shallow 96.8 (182) 0.5 (1) 1.6 (3) 1.1 (2) 100(188)
100 deep 77.8 (14) 0.0 (0) 22.2 (4) 0.0 (0) 100 (18)
100medium 100.0 (45) 0.0 (0) 0.0 (0) 0.0 (0) 100 (45)
100shallow 93.9 (154) 0.0 (0) 3.7 (6) 2.4 (4) 100(164)
Table 2. Numberof extragalacticERCSC sourcesin oursample priorto the completenesscut(andafter the completenesscutin
parentheses),togetherwiththesurfaceareaofthezones.
ν number number number totalsource surfacearea surfacearea surfacearea totalsurface
[GHz] deep medium shallow number deep medium shallow [deg2]
857 77(24) 262(115) 719(313) 1058 [phone] 9800 12969
545 46 (8) 153 (69) 318(143) 517 [phone] 9551 12749
353 35(14) 104 (59) 198(151) 337 (224) 1086 2971 12373 16431
217 17(15) 65 (57) 183 (71) 265 (143) 1104 3169 11900 16174
143 13 (8) 48 (44) 184 (90) 245 (142) 1111 2972 11977 16061
100 14 (6) 45 (39) 158 (77) 217 (122) 1072 2870 12611 16554
Table3.NumberofextragalacticERCSCsourcesactuallyusedinthenumbercounts:deep–medium–shallow,withthenumbers
ofunidentifiedsourcesaftertheslash.
S S range 857GHz 545GHz 353GHz 217GHz 143GHz 100GHz
h νi ν
[Jy] [Jy]
0.398 0.303 – 0.480 0 0 0 7/0-28/1-92/9 3/0-12/0-81/1 0
0.631 0.480 – 0.762 0 0 8/0-31/4-83/14 3/0-13/1-34/2 2/0-13/0-43/0 4/0-15/0-68/3
1.000 0.762 – 1.207 0 0 4/0 - 18/2 - 38/7 4/0 - 8/0 - 21/1 4/0-10/0-25/0 4/0-11/0-47/1
1.585 1.207 – 1.913 0 5/0-26/6-78/9 1/1 - 6/0 - 14/0 1/0 - 4/0 - 11/0 2/0 - 4/0 - 12/0 2/0 - 7/0 - 13/0
2.512 1.913 – 3.032 11/0-31/1-139/4 0/0-20/1-28/1 1/0 - 2/0 - 8/0 0/0 - 1/0 - 3/0 0/0 - 3/0 - 5/0 0/0 - 3/0 - 11/0
3.981 3.032 – 4.805 7/0 - 33/2 - 95/4 1/0-16/1-20/1 0/0 - 1/0 - 6/0 0/0 - 2/0 - 1/0 0/0 - 1/0 - 3/0 0/0 - 2/0 - 3/0
6.310 4.805 – 7.615 3/0 - 22/2 - 41/7 1/0 - 6/0 - 8/1 0 0/0 - 1/0 - 1/0 0/0 - 1/0 - 1/0 0/0 - 0/0 - 2/0
10.000 7.615 – 12.069 2/0 - 20/1 - 23/1 1/0 - 1/0 - 5/0 0/0 - 1/0 - 0/0 0 0/0 - 0/0 - 1/0 0/0 - 1/0 - 1/0
15.849 12.069–22.801 1/0 - 9/0 - 15/0 0/0 - 0/0 - 4/0 0/0 - 0/0 - 2/0 0 0 0
Table 2 summarises the source number and surface area of three zones presented in Fig. 6. The uncertaintiesin complete-
eachzone(deep,mediumandshallow).Wefindatotalnumber ness are at the 5% level, as discussed in Planck Collaboration
ofsourcesrangingfrom217at100GHzto1058at857GHz. VII(2011)andPlanckCollaboration(2011).Thecorrectionfor
incompleteness is then applied to the number counts of each
zoneseparately.
2.4.Completeness
We will use a completeness level threshold of 80% for all
TheERCSCPipeline(PlanckCollaborationVII,2011)usedex- frequencies. This ensures: 1) a minimal source contamination;
tensive Monte-Carlo simulations to assess various parameters 2) no photometric biases (Planck Collaboration, 2011); and 3)
such as positional or flux density accuracies. Here, we use the good photometric accuracy(Planck Collaboration, 2011) – see
resultsofthoserunstoestimatethecompletenessineachofthe Sect. 2.5. The number of sources actually used to estimate the
6 PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz
Figure7. Distributionofspectralindices(forthesourcespresentintheERCSCatagivenfrequencyinoursamplewithcomplete-
nessof80%orabove).Hereα217 isshownasadottedline,α353 asadashedline,andα857 asasolidline.Theregion2 α 4
143 217 545 ≤ ≤
is typical of thermal dust emission. In red we show the dusty sources, whereas in blue we show the synchrotron sources. The
sourcesin all three samples(deep, medium,and shallow) are combinedhere. Note that, as expected,the 100GHz, and 143GHz
samplesaredominatedbyradiogalaxies,whereasthe545GHzand857GHzsamplesaredominatedbydustygalaxies.At217GHz
and 353GHz we observethe transition betweenthe two populations,with significantnumbersof both typesbeingpresentin the
samples.
numbercountsisgiveninTab.3,whichalsoincludesthenum- ment Planck Collaboration (2011). Here we use the “FLUX”
berofunidentifiedsources.We finallyuseanumberofsources field for flux densities. Notice that 100 and 217GHz flux den-
rangingfrom122at100GHzto452at857GHz(Tab.2). sities can be affected by Galactic CO lines (Planck HFI Core
Team,2011b).
We would like to emphasise that the extensive simulations
performed in the process of generating/validating the ERCSC
allowustoderivereliablecompletenessestimatesforeachzone
(seeSect.2.4)andalsotoestimatethequalityofthephotometry.
In the faintest flux density bins that we are using (correspond-
ing to 80% completeness), there is no photometric offset, and
the photometric accuracy from the Monte-Carlo simulations is
about(PlanckCollaboration,2011,;Fig.7forreference):35%
at 480mJy for 100GHz; 30% at 300mJy for 143GHz; 20%
at 300mJy for 217GHz; 20% at 480mJy for 353GHz; 20%
at 1207mJy for 545GHz; and 20% at 1913mJy for 857GHz.
Thisscatter(inthefaintestfluxdensitybins)stronglydecreases
at larger flux densities. Notice that photometric uncertainties
can bias the determination of the counts slope (e.g. Murdoch,
Crawford&Jauncey,1973).
From our sample, we also create “Band-filled catalogues”.
For each frequency/zonesample, we take each source position
from the ERCSC and perform aperture photometry from the
corresponding images in the other frequencies. We adopt 4σ
Figure8. Fraction of galaxy types as a function of frequency,
asthedetectionthreshold.Theseaperturephotometrymeasure-
on the sample having completeness of 80% and above: dusty
ments(andupperlimits) are usedfor thespectralclassification
(red squares) and synchrotron (blue triangles). Error bars are
of sources and in the spectral index determinations, but not in
Poissonian.
the number counts measurements(which rely only on ERCSC
fluxdensities).WedefinethespectralindexαbyS να.
ν
∝
Thederivedspectralindicesareusedtodeterminethecolour
correctionoftheERCSCfluxdensities(PlanckHFICoreTeam,
2.5.Photometry
2011b). This correction changes the flux densities by at most
ThephotometryoftheERCSCisextensivelydetailedinPlanck 5% at 857GHz, 15% at 545GHz, 14% at 353GHz, 12% at
Collaboration VII (2011) as well as in the explanatorysupple- 217GHz,and1%at143GHzand100GHz.
PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz 7
Table4.Plancknumbercountsat353,545and857GHz.
S S range dN/dS S2.5 N> S dN/dS S2.5 N> S dN/dS S2.5 N> S
h νi ν 857 857 857 545 545 545 353 353 353
[Jy] [Jy] [Jy1.5sr 1] [sr 1] [Jy1.5sr 1] [sr 1] [Jy1.5sr 1] [sr 1]
− − − − − −
0.631 0.480 – 0.762 0 0 0 0 30.6 3.7 50.1 3.3
± ±
1.000 0.762 – 1.207 0 0 0 0 26.4 4.0 21.0 2.1
± ±
1.585 1.207 – 1.913 0 0 131.4 21.4 60.3 4.1 17.6 4.1 8.4 1.3
± ± ± ±
2.512 1.913 – 3.032 466.2 70.5 127.7 6.0 104.6 20.4 28.9 2.7 18.3 5.7 4.2 0.9
± ± ± ± ± ±
3.981 3.032 – 4.805 613.6 96.6 71.8 4.4 160.2 33.8 16.3 2.1 23.3 9.0 2.0 0.6
± ± ± ± ± ±
6.310 4.805 – 7.615 573.4 103.4 35.0 3.0 129.4 37.5 6.7 1.3 0 0
± ± ± ±
10.000 7.615 – 12.069 755.2 150.3 17.7 2.1 119.5 47.8 2.8 0.9 13.2 13.3 0.6 0.3
± ± ± ± ± ±
15.849 12.069–22.801 837.1 200.5 6.3 1.3 136.2 70.4 1.0 0.5 52.9 37.6 0.4 0.3
± ± ± ± ± ±
Table5.Plancknumbercountsat100,143and217GHz.
S S range dN/dS S2.5 N> S dN/dS S2.5 N> S dN/dS S2.5 N> S
h νi ν 217 217 217 143 143 143 100 100 100
[Jy] [Jy] [Jy1.5sr 1] [sr 1] [Jy1.5sr 1] [sr 1] [Jy1.5sr 1] [sr 1]
− − − − − −
0.398 0.303 – 0.480 16.5 2.0 54.3 3.54 8.5 1.1 44.3 2.95 0 0
± ± ± ±
0.631 0.480 – 0.762 11.5 1.9 23.0 2.21 13.7 2.1 28.1 2.46 21.9 3.0 43.7 3.14
± ± ± ± ± ±
1.000 0.762 – 1.207 14.6 2.8 12.0 1.58 17.4 3.1 15.0 1.77 29.1 4.4 22.9 2.22
± ± ± ± ± ±
1.585 1.207 – 1.913 13.6 3.6 5.1 1.01 15.4 3.8 6.7 1.17 18.8 4.3 9.1 1.35
± ± ± ± ± ±
2.512 1.913 – 3.032 6.8 3.4 1.8 0.61 13.6 5.0 3.1 0.79 23.2 6.5 4.6 0.95
± ± ± ± ± ±
3.981 3.032 – 4.805 10.1 5.9 1.0 0.45 13.6 6.9 1.4 0.54 16.5 7.5 1.8 0.59
± ± ± ± ± ±
6.310 4.805 – 7.615 13.5 9.6 0.4 0.29 13.6 9.7 0.6 0.35 13.2 9.4 0.8 0.40
± ± ± ± ± ±
10.000 7.615–12.069 0 0 13.6 13.6 0.2 0.20 26.3 18.7 0.4 0.28
± ± ± ±
3. Classificationofgalaxiesintodustyor tionrisessignificantlyifweremovethecompletenesscutdueto
synchrotroncategories the increasing photometricuncertaintiesat lower flux densities
(seeAppendixAfordetails).Examplesofboth“dusty”or“syn-
For the purposes of this paper, we aim for a basic classifica- chrotron”sourceswithsomewhatunusualSEDsarediscussedin
tionbasedonSEDsthatseparatessourcesintothosedominated AppendixB.
by thermal dust emission and those dominated by synchrotron
emission. (Free-free emission does exist, but is not dominant,
e.g., Peel et al. 2011).In orderto classify our sourcesby type,
westartwiththeband-filledcataloguesdiscussedinSection2.5.
Thermaldustemissionisexpectedtoshowspectralindicesinthe 4. Planckextragalacticnumbercountsbetween100
range α 2–4. On the other hand, colder temperature sources
can show∼lower α857, if the 857GHz point is near the spectral and857GHz
545
peak.The presenceofa strongsynchrotroncomponent,orper-
The Planck extragalactic number counts (differential, nor-
haps free-free emission component, would start to flatten the
malisedtotheEuclideanslope,andcompleteness-corrected)are
SED below 353GHz. With such issues in mind, we have set
∼ presentedinFig.9andTables4and5.Theyareobtainedusinga
upthefollowingclassificationalgorithm:
meanofthe3zones(weightedbythesurfaceareaofeachzone).
all sourceswith α857 2,orα545 2 are assigneda “dusty”
• 545≥ 353≥ The error budget in the number counts is made up of: (i)
classification;
Poisson statistics (ii); the 5% uncertainty in the completeness
all sources where both of these spectral indices are lower
• correction(iii);theabsolutephotometriccalibrationuncertainty
than2,includingnon-detections,areassigned“synchrotron”
of2%atandbelow353GHz,and7%above545GHz(Planck
classification.
HFICoreTeam,2011a,b).Accordingtoe.g.Eq.1ofBethermin
TheresultingclassificationissummarisedinFig.7,showingthe etal. (2011),thecalibrationuncertaintiesscale to the powerof
spectral index distributions for each type as a function of ob- 1.5forEuclideannormalisednumbercounts.
servedfrequency.
Notice that our bright counts error budget is dominated by
However,somesourcesaredifficulttoclassify,andcouldbe
sample variance at low redshift and by small-number statis-
part of an “intermediate dusty” or “intermediate synchrotron”
tics. For instance, the small wiggle seen in the counts at the
type.Theseintermediatesourcescanbedefinedasfollows:
three highest frequencies (seen at 600mJy at 353GHz, 4Jy at
beingdusty(accordingtoourcriterionabove)butalsohav- 545GHz and 10Jy at 857GHz) is due to a few tens of local
• ingα857 <1 NGCsourcesinthemediumzone(seeAppendicesBandC).
100
beingsynchrotron(accordingtoourcriterionabove)butalso
• withα70 measurable(usingLFIdata),i.e.sourcesthatshow Integral (i.e. cumulative) combined number counts are
30 showninFig.11.Althougherrorbarsarehighlycorrelated,these
botha significantdustcomponentanda strongsynchrotron
counts provide a useful estimate of the source surface density.
component.
Thecompletenesscorrectionisalsoappliedhere,andweusethe
Among the sources included in the number counts analy- samecutsinfluxdensityasforthedifferentialcounts.Tab.4and
sis, fewerthan10% areclassified as“intermediate”.Thisfrac- 5alsogivetheN >S values.
8 PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz
Table6.Plancknumbercountsofdustygalaxies(only)between217and857GHz.
S S range dN/dS S2.5 dN/dS S2.5 dN/dS S2.5 dN/dS S2.5
h νi ν 857 857 545 545 353 353 217 217
[Jy] [Jy] [Jy1.5sr 1] [Jy1.5sr 1] [Jy1.5sr 1] [Jy1.5sr 1]
− − − −
0.398 0.303 – 0.480 0 0 0 3.2 0.7
±
0.631 0.480 – 0.762 0 0 23.6 3.1 2.3 0.8
± ±
1.000 0.762 – 1.207 0 0 18.5 3.2 1.8 0.9
± ±
1.585 1.207 – 1.913 0 129.0 21.1 12.5 3.4 0
± ±
2.512 1.913 – 3.032 463.6 70.2 100.3 19.8 11.7 4.5 1.7 1.7
± ± ± ±
3.981 3.032 – 4.805 609.0 96.0 151.5 32.5 13.3 6.7 0
± ± ±
6.310 4.805 – 7.615 573.4 103.4 129.4 37.5 0 0
± ±
10.000 7.615 – 12.069 755.2 150.3 119.5 47.8 13.2 13.3 0
± ± ±
15.849 12.069–22.801 837.1 200.5 136.2 70.4 26.4 26.5 0
± ± ±
Table7.Plancknumbercountsof(only)synchrotrongalaxiesbetween100and857GHz.
S S range dN/dS S2.5 dN/dS S2.5 dN/dS S2.5 dN/dS S2.5 dN/dS S2.5 dN/dS S2.5
h νi ν 857 857 545 545 353 353 217 217 143 143 100 100
[Jy] [Jy] [Jy1.5sr 1] [Jy1.5sr 1] [Jy1.5sr 1] [Jy1.5sr 1] [Jy1.5sr 1] [Jy1.5sr 1]
− − − − − −
0.398 0.303 – 0.480 0 0 0 13.3 1.7 8.5 1.1 0
± ±
0.631 0.480 – 0.762 0 0 7.0 1.4 9.2 1.6 13.5 2.1 21.9 3.0
± ± ± ±
1.000 0.762 – 1.207 0 0 7.9 2.0 12.8 2.6 16.9 3.1 28.6 4.3
± ± ± ±
1.585 1.207 – 1.913 0 2.4 1.7 5.0 2.1 13.6 3.6 15.4 3.8 18.8 4.3
± ± ± ± ±
2.512 1.913 – 3.032 2.6 2.6 4.3 3.1 6.7 3.4 5.1 3.0 13.6 5.0 23.2 6.5
± ± ± ± ± ±
3.981 3.032 – 4.805 4.5 4.6 8.7 6.2 10.0 5.8 10.1 5.9 13.6 6.9 16.5 7.5
± ± ± ± ± ±
6.310 4.805 – 7.615 0 0 0 13.5 9.6 13.6 9.7 13.2 9.4
± ± ±
10.000 7.615 – 12.069 0 0 0 0 13.6 13.6 26.3 18.7
± ±
15.849 12.069–22.801 0 0 26.4 26.5 0 0 0
±
5. Results&Discussion 5.2.PlanckNumberCountsComparedwithOtherDatasets
The number counts are in fairly good agreement at lower fre-
5.1.NatureoftheGalaxiesatsubmillimetreandmillimetre
quencies (100 to 217GHz) with the counts published in the
wavelengths
Planckearlyresults,basedona30GHzselectedsample(Planck
Collaboration XIII, 2011), represented as diamonds in Fig. 9
and10.Theeffectofthecompletenesscorrectionisseeninthe
The change in the nature of sources (synch vs dusty) with
fainterfluxdensitybins,belowabout500mJy.However,weno-
frequency was first observed in the Planck data in Planck
ticeaslightdisagreementaround400mJyat217GHz,whereour
Collaboration VII (2011). Our new sample allows a more pre-
counts of synchrotron galaxies exceed the Planck early counts
cisequantificationbecauseofitscompleteness.Thestatisticsof by a factor of 1.7 (13.7 1.5vs. 8.2 0.9Jy1.5sr 1). This dis-
synchrotronvs.dustygalaxiesaresummarisedinFig.8,show- ± ± −
crepancycanbeeasilyunderstood:ourcurrentselectionofsyn-
ing the fraction of galaxy type as a function of frequency. We
chrotronsourcesisnotthesameastheoneadoptedinthePlanck
estimatetheuncertaintyintheclassificationtobeoftheorderof
EarlyresultspaperPlanckCollaborationXIII(2011),inwhicha
10% (see AppendixA). The striking resultis the almost equal morerestrictivecriterionwasused(α217 < 0.5).Ifweadoptthe
contribution of both source types near 300GHz. The high fre- 143
samecriterionasinthatpaperwefindnostatisticallysignificant
quencychannels(545and 857GHz) are, ynsurprisingly,domi-
differencebetweenthetwoestimatesofthenumbercounts.
nated (> 90%) by dusty galaxies.The low frequencychannels
are,unsurprisingly,dominated(>95%)bysynchrotronsources
OurcountsestimatealsoseemconsistentwiththeHerschel
at100and143GHz.At217GHz,fewerthan10%ofthesources
ATLASandHerMEScounts(Clementsetal.,2010;Oliveretal.,
showadust-dominatedSED.
2010)athighfrequency(545and857GHz)as wellasBLAST
atthesametwowavelengths(Betherminetal.,2010b),although
Allthesourcesfromourcompletesamplehaveanidentified thereisnodirectoverlapinfluxdensityandsmallnumberstatis-
spectral type (by construction), and we can compute the num- ticsaffectthebrightestHerschelpoints.
bercountsforsynchrotronanddustygalaxies.Fig.9,10and11
show the differential and integral number counts by type, also TheACT148GHzdata(Marriageetal.,2011)andSPT150
given in Tables 6 and 7. We note that at 353GHz, about two and 220GHz (Vieira et al., 2010) are also plotted in Fig. 10,
thirds of the number counts are made-up by dusty sources. At together with SCUBA and LABOCA data at 353GHz (Borys
217GHz(545GHz)thereisaminorcontribution(10%orless) etal.,2003;Coppinetal.,2006;Scottetal.,2006;Beelenetal.,
of the dusty (synchrotron) sources contributing to the counts. 2008;Weissetal.,2009),coveringmorethanfourordersofmag-
These number counts of extragalactic dusty and synchrotron nitudeinfluxdensity.
sources are an important step to further constrain models of
galaxy SED and to include them in more general models of Wefinallycheckedthatourcountsareinagreementwiththe
galaxyevolution(seebelow). PlanckSkyModel(Delabrouilleetal.,2012).
PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz 9
Figure9.Planckdifferentialnumbercounts,normalisedtotheEuclideanvalue(i.e.S2.5dN/dS),comparedwithmodelsandother
datasets.Planckcounts:total(blackfilledcircles);dusty(redcircles);synchrotron(bluecircles).Fourmodelsarealsoplotted:de
Zottietal.(2005,dealingonlywithsynchrotronsources–solidline);Tuccietal.(2011,dealingonlywithsynchrotronsources–
dots);Betherminetal.(2011,dealingonlywithdustysources–longdashes);Serjeant&Harrison(2005,dealingonlywithlocal
dustysources–shortdashes).Otherdatasets:Planckearlycountsfor30GHz-selectedradiogalaxies(PlanckCollaborationXIII,
2011)at100,143and217GHz(opendiamonds);HerschelATLASandHerMEScountsat350and500µmfromOliveretal.(2010)
andClementsetal.(2010);BLASTatthesametwowavelengths,fromBetherminetal.(2010b),allshownastriangles.Leftvertical
axesareinunitsofJy1.5sr 1,andtherightverticalaxisinJy1.5.deg 2.
− −
5.3.PlanckNumberCountsandModels Thede Zottiet al. (2005)modelfocussesonradio sources,
flat-andsteep-spectrum,thelatterhavingacomponentofdusty
spheroidals and GPS (GHz peaked spectrum) sources. Its cos-
5.3.1. Models
mologicalevolution for extragalactic radio sources is based on
analysisof all the mainsource populationsat GHz frequencies
Fig. 9 and 10 display our present estimates of number counts
and currently providesa good fit to all data on number counts
of extragalactic point sources, based on ERCSC data, together
andonotherstatisticsfrom 5GHzandupto 100GHz.This
with predictions from recent evolution models. These models ∼ ∼
modeladoptsasimplepower-law,withanalmost“flat”spectral
focus either on radio-selectedsources – i.e. sources with spec-
index(α 0.1),for extrapolatingthe spectra of the brightest
tra dominated by synchrotron radiation at mm/submm wave- extragalac≃tic−sources(essentially“blazar2 sources”)atfrequen-
lengths(“synchrotronsources”):deZottietal.(2005)andTucci
ciesabove100GHz.
et al. (2011)– or on far-IR selected sources– i.e. sourceswith
The Tucci et al. (2011) models provide a description of
spectradominatedbythermalcolddustemissionatmm/submm
three populations of radio sources: steep-, flat-, and inverted-
wavelengths(“dustysources”):Serjeant&Harrison(2005)and
spectrum. The flat-spectrum population is further divided into
Bethermin et al. (2011). Many other models exist in the lit-
Flat-Spectrum Radio Quasars (FSRQ), and BLLacs. The main
erature, among which are Le Borgne et al. (2009), Negrello
etal.(2009),Pearson&Khan(2009),Rowan-Robinson(2009),
2 Blazarsarejet-dominatedextragalacticobjects,observed withina
Valiante et al. (2009), Franceschini et al. (2010), Lacey et al.
smallangleofthejetaxisandcharacterizedbyahighlyvariable,non-
(2010),Marsdenetal.(2011),Wilmanetal.(2010),andRahmati thermalsynchrotronemissionatGHzfrequenciesinwhichthebeamed
&vanDerWerf(2011).Acomparisonisgivenwiththesemod- component dominates the observed emission (Angel & Stockman,
elsinFig.12for857GHz. 1980).
10 PlanckCollaboration:PlanckstatisticalpropertiesofIRandradioERCSCextragalacticsources100–857GHz
Figure10.SameasFig.9,butonawiderfluxdensityscaleandwiththeadditionofACT(Marriageetal.,2011)andSPT(Vieira
etal.,2010)dataassquaresat143and217GHz,andSCUBAandLABOCAdataasstarsat353 GHz(Borysetal.,2003;Coppin
etal.,2006;Scottetal.,2006;Beelenetal.,2008;Weissetal.,2009).
Figure12. Planck differential number counts at 857GHz, nor-
malised to the Euclidean (i.e. S2.5dN/dS). Planck counts: to-
Figure11.Planck integralnumbercounts(filledcircles);dusty tal (black filled circles); dusty galaxies (red circles). Models:
(red);synchrotron(blue).Verticalaxesareinnumberperstera- Bethermin et al. (2011) (long red dashes – dusty); Serjeant &
dian; right axis is in number per square degree. Counts are Harrison (2005) (short green dashes – dusty); Rahmati & van
completeness-corrected.The same faint flux density cut as for Der Werf (2011)(black dash-dot-dot-dotline). Other: Lagache
differentialcountsisapplied. et al. (2004); Negrello et al. (2007); Rowan-Robinson (2009);
Valiante et al. (2009); Pearson & Khan (2009); Franceschini
etal.(2010);Laceyetal.(2010);Wilmanetal.(2010).
noveltyofthesemodelsisthestatisticalpredictionofthe“break”
frequency, ν , in the spectra of blazar jets modeled by classi-
M