Mon.Not.R.Astron.Soc.000,000–000(0000) Printed5February2008 (MNLATEXstylefilev2.2) Deep 610-MHz GMRT observations of the Spitzer extragalactic First Look Survey field – I. Observations, data analysis and source catalogue ⋆ Timothy Garn , David A. Green, Sally E. G. Hales, Julia M. Riley, Paul Alexander 7 0 AstrophysicsGroup,CavendishLaboratory,19J.J.ThomsonAve.,CambridgeCB30HE 0 2 n 5February2008 a J 8 ABSTRACT 1 ObservationsoftheSpitzerextragalacticFirstLookSurveyfieldtakenat610MHzwiththe GiantMetrewaveRadioTelescopearepresented.Sevenindividualpointingswereobserved, 1 coveringa totalareaof 4deg2 witha resolutionof5.8 4.7,PA 60 .Ther.m.s.noiseat ′′ ′′ ◦ v thecentreofthepointing∼sisbetween27and30m Jybefore×correctionfortheGMRTprimary 4 beam. The techniques used for data reduction and productionof a mosaicked image of the 3 regionare described,and the final mosaic, along with a catalogueof 3944sourcesdetected 5 above 5s ,arepresented.Thesurveycomplementsexistingradioandinfrareddataavailable 1 ∼ 0 forthisregion. 7 Keywords: surveys–catalogues–radiocontinuum:galaxies 0 / h p - o 1 INTRODUCTION gion,madewiththeVeryLargeArray(VLA)inB-arrayconfigura- r tion,andtheWesterborkSynthesisRadioTelescope(WSRT).The st The Spitzer First Look Survey was the first major scientific pro- VLAsurveycovered 4deg2witharesolutionof5 ,toa1s depth a gram carried out by the Spitzer Space Telescope (Werneretal. of 23m Jy(Condon∼etal.2003)andcontains3565′′sources,while v: 2004),observingfor∼110hoursofDirector’sDiscretionaryTime the∼WSRTsurvey(Morgantietal.2004)concentratedon 1deg2 i inDecember2003.Theaimoftheextragalacticcomponentofthe ofthexFLSregiontoadepthof8.5m Jybeam 1withare∼solution X FirstLookSurvey(xFLS)wastostudyaregionwithlowGalactic − of14 11 ,PA0 ,andcontains1048sources. r backgroundtoasignificantlydeeperlevelthananypreviouslarge- ′′× ′′ ◦ a There is a tight correlation between the far-infrared and ra- area infrared survey, in order to accurately characterise the dom- dio luminosities of galaxies, which has been known for many inant infrared source populations. Observations were made over four square degrees centred on 17h18m00s, +59 3000 (J2000 years (see e.g. Helouetal. 1985; Condon 1992). The correla- ◦ ′ ′′ tion applies to both the local (Hughesetal. 2006) and global coordinates, which are used throughout this paper) with two in- (Murphyetal. 2006) properties of galaxies in the local and dis- struments–theInfraredArrayCamera(IRAC,Fazioetal.2004), usingthe3.6,4.5,5.8and8 m mbands, andtheMultibandImag- tantuniverse(Garrett2002;Gruppionietal.2003;Chapmanetal. 2005;Luo&Wu2005).Therehavebeen twoprevious studiesof ingPhotometerforSpitzer(MIPS,Riekeetal.2004)at24,70and 160m m.Thedatahavebeenprocessedandmapsandsourcecata- theIR/radiocorrelationinthexFLSregion(Appletonetal.2004; loguesarecurrentlyavailable(IRAC–Lacyetal.2005;24 m m– Frayeretal.2006).Appletonfoundnoevidenceforavariationof Faddaetal.2006;70and160m m–Frayeretal.2006). the correlation with redshift, while Frayer, using more sensitive infrared data, found a decrease in the infrared/radio flux density Complementary observations have been taken at a range of ratio with z. To study the luminosity variation it is necessary to wavelengths to fully exploit the new deep infrared data. Deep accuratelyk-correcttheobservedradiofluxdensitiestotheirrest- optical surveys of the region have been completed in the R- framevaluesusingthespectralindexa (wherea isheredefinedso band (KPNO 4 m, Faddaetal. 2004), and in the u - and g- ∗ thatfluxdensityscaleswithfrequencyasS(cid:181) n a ).Bothprevious bands(CFHT3.6m,Shimetal.2006).ThexFLSregionwascov- − worksinthisregionassumedthatallsourceshavethesameradio ered by the early data release of the Sloan Digital Sky Survey spectralindex(takentobea =0.7and0.8respectively),butinor- (Stoughtonetal.2002).Afurtherredshiftsurveytargetingselected 24-m msourceswasmadewiththeMMT/Hectospecfiberspectro- der toperformthiscorrectionaccuratelyitisnecessary tohavea detectioninatleasttworadiofrequencies. graph(Papovichetal.2006),andatotalof1587redshiftsarepub- liclyavailable. Thelargeamountofexistingdataonthisfieldmakesitagood Therearetwoexisting1.4-GHzradiosurveysofthexFLSre- candidate for a further deep radio survey. We have imaged the xFLS region at 610 MHz with the Giant Metrewave Radio Tele- scope (GMRT), reaching an r.m.s. noise level of around 30 m Jy ⋆ E-mail:[email protected] beforeprimarybeamcorrection.Oursurveyhasacomparableres- 2 T. Garnet al. mosaic,anddiscusstheartefactsseennearbrightsources.InSec- tion4wecompareourcataloguewiththeVLAsurvey,inorderto testthepositionalaccuracyofourdata.Furtheranalysesofthedata, 60 30 includingdetailedcomparisionswithSpitzerdatawillbemadein forthcoming papers. Appendix A presents some technical details ofcorrectionsmadetotheobservedGMRTdata,andAppendixB 00 givesdetailsofacorrectionmadetotheintegratedfluxdensitiesof ON (J2000) sources. DECLINATI59 30 2 OBSERVATIONSANDDATAREDUCTION 00 ObservationsofthexFLSregionweremadeat610MHzwiththe GMRT (Ananthakrishnan 2005), located near Pune, India, over four days in March 2004. Seven pointings were observed, in a 58 30 hexagonal grid centred on 17h18m00s, +59 3000 as shown in ◦ ′ ′′ Fig.1.Thepointingswerespacedby43arcmin,approximatelythe half-power bandwidth of the GMRT primary beam at 610 MHz, 17 25 20 15 10 RIGHT ASCENSION (J2000) whichgivesnearlyuniformnoisecoverageovermostofthexFLS region. Figure1.ThesevenpointingsobservedwiththeGMRT.Pointingcentres Observationsof 3C48or 3C286weremadeatthebeginning arespacedby43arcmin.Thegreyscalerepresentsther.m.snoiselevelcal- culatedbySourceExtractor(seeSection3),aftercorrectingeachpointing and end of each observing session in order to calibrate the flux fortheprimarybeamresponseoftheGMRTandmosaickingthepointings density scale. The AIPS task SETJY was used to calculate flux together.The45m Jycontour(equivalent tothe23m Jynoiselevelofthe densities at 610 MHz of 29.4 and 21.1 Jy respectively, using an VLA 1.4 GHz survey, for a source with spectral index of 0.8) has been extensionoftheBaarsetal.(1977)scale.Twosidebandswereob- plotted. served, eachof 16MHzwith128spectral channels, anda16.9s integrationtimewasused.Eachfieldwasobservedfor 200min ∼ over four 10 hour observing sessions made between March 23rd and27th. Theobservations consisted ofinterleaved20minscans ofeachpointing,inordertoimproveuvcoverage.Anearbyphase calibrator, J1634+627, was observed for three minutes between the scans of each pointing to monitor any time-dependant phase andamplitudefluctuationsofthetelescope.Themeasuredphaseof J1634+627 variedsmoothlybetweenobservations, withatypical variation between calibrator observations of lessthan 40 for the ◦ longbaselineantennasandbelow10 fortheshortbaselineanten- ◦ nas. Standard AIPStasks wereused toflag bad baselines, anten- nas, channels that were suffering from narrow band interference, andthefirstandlast16.9sintegrationperiodofeachscan.Aband- passcorrectionwasappliedusingthefluxcalibrators,foreachan- tenna.Apseudo-continuumchannelwasthenmadebycombining thecentraltenchannelstogether,andanantenna-basedamplitude andphasecalibrationcreatedusingtheobservationsofJ1634+627. Thiscalibrationwasappliedbacktotheoriginal128channeldata set,which was compressed into 11channels, eachcontaining ten oftheoriginalspectralchannels(sothefirstandlastfewchannels, whichtendedtobethenoisiest,werediscarded).Thesmallwidthof thesenewchannelsensuresthatbandwidthsmearingisnotaprob- Figure2.Theuvcoverageforthecentralpointing.Baselineslessthan1kl lem,andall11channelscouldbeindividuallyinspectedtoremove werenotusedinimagingandhavebeenomitted. additionalinterference.Aftertheflaggingandcalibrationwascom- pletethetwosidebandswerecombinedintoasingledataset(see Appendix A1) to improve the uv coverage. The coverage for the olution (5.8 4.7 compared with 5 ) to the VLA 1.4-GHz sur- centralpointingisshowninFig.2. ′′ ′′ ′′ × vey,allowingadirectcomparisonbetweensourcepropertiesatthe ThelargefieldofviewoftheGMRTleadstosignificanterrors twowavelengthstobemade.Ourobservationsaredeeperthanthe ifthewholefieldisimageddirectly,duetothenon-planarnatureof VLAsurveyovermostoftheimage,foratypicalradiosourcewith thesky.Tominimisetheseerrors,eachpointingwasbrokendown a = 0.8. into 19 smaller facets which were imaged separately, with a dif- Section2describestheobservations,anddatareductiontech- ferentassumedphasecentre,andthenrecombinedtodealwiththe niques. In Section 3 we present the mosaicked image and source transformationfromplanestothesphere.Imagesweremadewith catalogue.Weshowaselectionofextendedobjectsvisibleinour an elliptical synthesised beam with size 5.8 4.7, position an- ′′ ′′ × Deep 610-MHzGMRTobservationsof theSpitzerextragalacticFirstLookSurveyfield– I. 3 Figure3.The610-MHzimageoftheSpitzerextragalactic FirstLookSurveyfield.Thegreyscalerangesbetween 0.2and1mJybeam 1.Artefactsare − − visiblearoundthebrightestfewsources. gle+60 ,withapixelsizeof1.5toensurethatthebeamwaswell (ii) The GMRT primary beam centre was shifted by 2.5 ar- ◦ ′′ ∼ oversampled. cmintocompensateforaposition-dependantfluxdensityerror. TheGMRThasalargenumberofsmallbaselines,duetothe Thetheoretical r.m.s.noiseof eachpointing, before primary clusterof14antennasinthecentral1kmofthearray.Thisdomi- beamcorrection,is natestheuvcoverageandaffectsthebeamshape,sobaselinesless than1kl wereomittedintheimaging.Theimageswentthrough s = √2Ts (1) threeiterationsofphaseself-calibrationat10,3and1minutein- Gpn(n 1)NIFD nt tervals, and then a final round of self-calibration correcting both − phase and amplitude errors. The overall amplitude gain was held where Ts 92 K is the system temperature, G 0.32 K Jy−1 is constantinordernottoalterthefluxdensityofsources.Theinitial theantenn≈again–valuestakenfromtheGMRTw≈ebsite1–nisthe self-calibrationstepcorrectedthephaseerrorsbyupto10 ,with number of working antennas, which was typically 28 during our ◦ later self-calibration making much smaller changes and having a observations,NIF=2isthenumberofsidebands,D n =13.75MHz smallereffectonther.m.s.imagenoise. istheeffectivebandwidthofeachsideband,andt 12000sisthe ≈ Two problems were identified and corrected during imaging averageintegrationtimeperpointing. (seeAppendixAforfurtherdetails). The final images have an r.m.s. noise of between 27 and (i) Anerrorinthetime-stampsoftheuvdata,andhencetheuvw coordinates,wascorrectedusingacustom-madeAIPStask. 1 http://www.gmrt.ncra.tifr.res.in/gmrthpage/Users/Help/help.html 4 T. Garnet al. 59 00 00 1 0.8 58 58 00 Declination 56 00 fraction of pixels 00..46 0.2 54 00 0 0 20 40 60 80 100 120 140 160 180 200 Noise level / microJy Figure 8. Pixel noise distibution, from the Source Extractor r.m.s. map. Thehistogramshowsthefractionofpixelsatthatnoise,whilethedotted 52 00 17 23 15 00 22 45 30 15 00 lineshowsthecumulativefractionofpixelswithnoisebelowaparticular Right Ascension value. Figure 4. Artefacts visible around a 294 mJy source near 17h22m40s, +58◦56′00′′.Thegreyscalerangesbetween−0.2and1mJybeam−1. 1800 1600 30 m Jy beam 1 before primary beam correction, which is very − closetothetheoreticallimitof 26m Jy,althoughdynamicrange 1400 ∼ issues limit the quality of the images near bright sources where 1200 the noise is greater. The seven pointings were corrected for the 1000 primary beam of the GMRT, taking into account the offset beam N position as discussed in Appendix A2. The beam correction was 800 performed using an 8th-order polynomial, withcoefficientstaken 600 fromKantharia&Rao(2001).Thepointingswerethenmosaicked together,weightingthefinalimagebyther.m.s.noisesofeachin- 400 dividual pointing, and the mosaic was cut off at the point where 200 theprimary beam correction factor dropped to20% of itscentral 0 value.Figure1illustratesthevariationinnoiseacrossthemosaic. -4 -3.5 -3 -2.5 -2 Thenoiselevelissmoothandaround30m Jyacrosstheinteriorof log10(Peak brightness / Jy) Figure9.Sourcenumbercountsfromthe610-MHzSpitzerextragalactic themap,andincreasestowardstheedgestoabout150m Jywhere FirstLookSurveyfield.Thesolidboxesrepresentuncorrectedcounts,while theprimarybeamcorrectionwas20%. theopenboxesrepresentthecountsaftercorrectionforthesurveysensitiv- itylevelforeachsolidangle. 3 RESULTSANDDISCUSSION andnoiselevelontheimage.Phaseerrorsnearthebrightsources ThefinalmosaicisshowninFig.3.Therearephaseartefactsvisi- leadtoanincreaseinnoise,butbyusingaboxof16 16pixelsto × blearoundthebrightestsources,whichhavenotbeensuccessfully estimatethelocalnoise, thenumber of spuriousdetectionsinthe removedduringself-calibration.Anenlargedimageofoneofthe finalcataloguewasreducedconsiderably–thismeansthatthefitis brightsourcesisshowninFig.4.Itisthoughtthattheartefactsare lessdeepnearthebrightestsources. due to an elevation-dependant error inthe position of the GMRT Table 1 presents a sample of 60 entries in the catalogue, primarybeam(seeAppendixA2),whichwillleadtoimagedistor- which is sorted by right ascension. The full table, and radio tionnearthebrightestobjectssincetheobservationsofeachpoint- image of the Spitzer extragalactic First Look Survey field will ingweretakeninaseriesofscanswithvaryingelevations. beavailableviahttp://www.mrao.cam.ac.uk/surveys/.Col- AsmallportionofthemosaicisshowninFig.5,demonstrat- umn1givestheIAUdesignation ofthesource, intheformFLS- ing the quality of the image away from the bright sources, with GMRT Jhhmmss.s+ddmmss, where J represents J2000.0 coordi- the VLA map of the same area shown in Fig. 6 for comparison. nates, hhmmss.s represents right ascension in hours, minutesand ThegreyscalehasbeensetonFigs5and6sothatanobjectwith truncatedtenthsofseconds,andddmmssrepresentsthedeclination aspectral indexof 0.8willappear equally bright inbothimages. indegrees,arcminutesandtruncatedarcseconds.Columns2and3 Mostsourcesareunresolvedinthe610-MHzimage,althoughthere give the right ascension and declination of the source, calculated aresome objects present withextended structures –wepresent a byfirstmomentsoftherelevantpixelfluxdensitiestogiveacen- sampleoftheseinFig.7. troidposition. Column4givesthebrightnessofthepeakpixelin A catalogue of 3944 sources was created using Source Ex- each source, in mJybeam 1, and column 5 gives thelocal r.m.s. − tractor(Bertin&Arnouts1996)withpeakbrightnessgreaterthan noiseinm Jybeam 1.Column6givestheintegratedfluxdensityin − 5s –seeAppendixBforfurtherdetails.SourceExtractorhasa mJy,calculatedfromthemosaicandapplyingthefluxdensitycor- ∼ significantadvantageoverAIPStaskssuchasSADwhencreating rectionfactordescribedinAppendixB.Column7givestheerror asourcelist,inthatitiscapableofcalculatingthelocalbackground inintegratedfluxdensity,calculatedfromthelocalnoiseleveland Deep 610-MHzGMRTobservationsof theSpitzerextragalacticFirstLookSurveyfield– I. 5 Figure5.AsamplesectionoftheGMRTimage.Thegreyscalerangesbetween 0.2and1mJybeam 1,andtheresolutionoftheimageis5.8 4.7arcsec2, − − × PA+60◦.Mostobjectsareunresolved,thoughoneclearlyextendedobjectisvisiblenearthemiddle. Figure6.Theexisting1.4-GHzVLAmapforthesameregion(Condonetal.2003).Thegreyscalerangesbetween 0.1and0.5mJybeam 1,equivalentto theGMRTimageforobjectswithaspectralindexof0.8.Theimageresolutionis5arcsec2. − − 6 T. Garnet al. 59 15 00 60 04 30 59 47 45 14 45 15 30 30 00 15 15 03 45 00 00 30 17 12 52 50 48 46 44 42 17 14 46 44 42 40 38 36 17 14 42 40 38 36 34 32 (a) FLSGMRTJ171247.5+591428.3 (b) FLSGMRTJ171440.2+600401 (c) FLSGMRT J171437.0+594711 and J171435.1+594728 59 37 35 60 08 10 30 05 25 00 20 07 55 15 50 10 45 05 40 00 35 36 55 30 17 23 42 40 38 36 34 17 17 38 36 34 32 30 (d) FLSGMRTJ172340.4+593721andJ172336.4+593715 (e) FLSGMRTJ171735.9+600752andJ171731.4+600746 58 45 55 58 52 50 58 45 40 50 45 35 45 40 30 40 35 25 35 30 20 30 25 15 25 20 10 20 15 05 15 10 00 17 13 48 47 46 45 44 43 42 41 40 17 19 58 57 56 55 54 53 52 51 50 17 15 14 13 12 11 10 09 08 07 06 (f) FLSGMRT J171344.4+584533 and (g) FLSGMRT J171954.3+585234 and (h) FLSGMRTJ171509.0+584526 J171341.9+584540 J171953.2+585222 59 48 30 60 04 45 59 21 00 15 30 20 45 00 30 15 47 45 15 00 30 17 14 18 17 16 15 14 13 12 11 10 17 24 42 41 40 39 38 37 36 35 34 17 21 20 19 18 17 16 15 14 13 12 (i) FLSGMRT J171413.8+594800 and (j) FLSGMRT J172439.4+592044, (k) FLSGMRT J172115.4+600432 and J171414.6+594750 J172435.7+592053andJ172437.3+592017 J172116.7+600406 Figure7.Aselectionofextendedobjects–contoursareplottedat 100m Jybeam 1 1,2,4,8,...Negativecontoursarerepresentedbydashedlines.The ± − × resolutionofthebeamisshowninthebottomleftofeachimage,andthedesignationsofeachsourceintheGMRT610-MHzcataloguearegivenbelow.All sourcesalsoappearextendedintheVLA1.4-GHzcatalogue,andimagesareavailableinCondonetal.(2003). Deep 610-MHzGMRTobservationsof theSpitzerextragalacticFirstLookSurveyfield– I. 7 59 57 30 00 56 30 00 55 30 00 54 30 17 12 55 50 45 40 35 30 25 20 15 (l) The extended source has been separated into FLSGMRT J171240.3+595659, J171229.1+595639, J171223.0+595551 and J171222.2+595524.ThetwocompactsourcesareFLSGMRTJ171248.1+595528andJ171216.4+595711. 60 09 00 60 37 30 08 45 15 30 00 15 36 45 00 30 07 45 15 30 00 15 35 45 00 30 06 45 15 17 11 54 52 50 48 46 44 42 40 38 36 17 18 50 45 40 35 30 (m) FLSGMRT J171147.7+600836, J171145.4+600808 and (n)FLSGMRTJ171841.3+603630 FLSGMRTJ171140.8+600731 59 16 00 15 45 30 15 00 14 45 30 17 16 55 50 45 40 35 30 25 20 (o) The extended object has been separated into FLSGMRT J171647.4+591458, J171639.4+151513, J171631.1+591544 and J171628.4+591527.ThedoublesourcecomponentsareFLSGMRTJ171622.4+591536andJ171623.1+591523. Figure7–continued 8 T. Garnet al. sourcesize,andtakingintoaccountthecorrectionfactor.Columns 8and9givetheX,Y pixelcoordinatesfromthemosaicimageof thesourcecentroid.Column10istheSourceExtractordeblended objectflag–0formostobjects,but1whenasourcehasbeensplit upintotwoormorecomponents.Fordeblendedobjectsitisneces- sarytoexaminetheimageinordertodistinguishbetweenthecase wheretwoastronomicallydistinctobjectshavebeensplitup, and whenoneextendedobjecthasbeenrepresentedbymorethanone entry.Thereare211deblendedsourcesinourcatalogue. The non-uniform noise characteristics of our survey make it important to quantify the area that has been surveyed with each noiselevel.Figure8showsthefractionofpixelswithaparticular noise level (taken fromtheSourceExtractor r.m.s.map), and the cumulative fractionofpixelsat that noiselevel. Thepeak bright- nessdistributionhasbeenplottedinFig.9,alongwithadistribu- tion that has been corrected for the varying amount of solid an- glebeingsurveyedtoeachsensitivitylevel.Previousstudieshave shown(Hopkinsetal.2002)thatSourceExtractorhasafalsede- Figure10.Sourcepositionsforobjectsdetectedat610MHzandundetected tectionrateof below 5%,and detectsabove90% ofsources with at1.4GHz.Mostsourcesarelocatednearthecentreofoneofourpointings, peakbrightnessclosetothedetectionthreshold. wheretheGMRTsensitivityisgreatest. 6 4 COMPARISONWITHTHEVLASURVEY 4 TheVLAsurveyofthexFLSregion(Condonetal.2003)hasauni- formnoiselevelof23m Jy.Thisisequivalenttoanoiseof 45m Jy ∼ at 610 MHz, assuming a spectral index of 0.8. Our observations 2 havesignificantlylowernoiselevelsthanthisinthecentreofour c) e s mosaic, withthe noise intheoverlap regions between two point- arc ingsbeingapproximately45m Jy(seeFig.1).Figure8showsthat et ( 0 thenoiseisbelow45m Jyforabouthalftheareaofourmosaic. offs C InordertomakeaquantitativecomparisonbetweentheVLA E D xFLSsurveyandourGMRTsurvey,wehaverunSourceExtractor -2 on our mosaic and the 1.4-GHz image. At the 5s level, there ∼ were 3826 sources detected at 610 MHz, and 3091 detected at 1.4 GHz, in the region covered by both surveys. Thesource lists -4 werematchedusingapairingradiusof6 ,whichisapproximately ′′ theresolutionofthetwocatalogues,and1580uniquematcheswere found.Figure10showsthesourcedistributionofobjectsfoundby -6 the GMRT but not by the VLA – the majority of the unmatched -6 -4 -2 0 2 4 6 RA offset (arcsec) sourcesarefoundinregionswhereourobservationsaredeeperthan theVLAimage. Figure11.SourcepositionsintheGMRTcataloguerelativetothepositions foundintheVLAimage,foruniquematcheswithin6 . It is likely that some of the sources that are not detected at ′′ 610 MHz but are detected at 1.4 GHz are flat spectrum objects, with spectral index 0.5. The VLA noise of 23 m Jy would be equivalent toa610-M≤Hznoiselevelof 35m Jyfora =0.5,so ACKNOWLEDGMENTS ∼ faint flat spectrum objects detectable throughout the VLA image WethankthestaffoftheGMRTwhohavemadetheseobservations wouldonlybedetectableinasmallregionofourmosaic. possible.TGthankstheUKPPARCforaStudentship.TheGMRT Figure 11 shows the position offsets of the matched GMRT isoperated by theNational Centre for Radio Astrophysics of the sources compared with their VLA counterparts. The offsets have TataInstituteofFundamentalResearch,India. anapproximatelyGaussiandistribution,withmeanoffsetinright ascension of 0.4, standard deviation 0.5 and in declination 0.2 ′′ ′′ ′′ withastandarddeviationof0.6.Theseoffsetshavenotbeenap- ′′ REFERENCES pliedtoourcatalogue,sinceitisuncertainastowhichsurveythe errorscomefrom. Ananthakrishnan S.,2005, Proceedingsofthe29thInternational The spectral index distribution of objects detected at both CosmicRayConference,10,125 frequencies is shown in Fig. 12. The integrated flux densities of AppletonP.,etal.,2004,ApJS,154,147 sourceshavebeenusedforthecalculation,wherethefluxdensity Baars J.,GenzelR.,Pauliny-TothI.,WitzelA.,1977,A&A,61, of VLA sources was corrected using the same method as for the 99 GMRTimage(detailsinAppendixB). BertinE.,&ArnoutsS.,1996,A&AS,117,393 Deep 610-MHzGMRTobservationsof theSpitzerextragalacticFirstLookSurveyfield– I. 9 Table1.Asampleof60entriesfromthe610-MHzextragalacticFirstLookSurveycatalogue,sortedbyrightascension. Name RA Dec Peak LocalNoise Int.FluxDensity Error X Y Flags J2000.0 J2000.0 mJybeam 1 m Jybeam 1 mJy mJy − − (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) FLSGMRTJ171751.2+584823 17:17:51.24 +58:48:23.2 0.210 35 0.330 0.070 3486 1752 0 FLSGMRTJ171751.3+594336 17:17:51.31 +59:43:36.2 0.190 34 0.150 0.050 3486 3961 0 FLSGMRTJ171751.3+591436 17:17:51.31 +59:14:36.5 0.210 39 0.140 0.050 3486 2801 0 FLSGMRTJ171751.3+590351 17:17:51.34 +59:03:51.1 0.380 41 0.490 0.080 3486 2371 0 FLSGMRTJ171751.5+590017 17:17:51.57 +59:00:17.8 0.290 42 0.350 0.070 3484 2228 0 FLSGMRTJ171751.6+590931 17:17:51.69 +59:09:31.2 0.310 42 0.350 0.070 3484 2597 0 FLSGMRTJ171751.9+591446 17:17:51.96 +59:14:46.4 0.400 40 0.470 0.070 3483 2807 0 FLSGMRTJ171752.0+603620 17:17:52.00 +60:36:20.1 0.310 57 0.350 0.100 3483 6070 0 FLSGMRTJ171752.4+594157 17:17:52.44 +59:41:57.9 0.280 37 0.310 0.070 3480 3895 0 FLSGMRTJ171752.5+600451 17:17:52.59 +60:04:51.1 0.230 38 0.240 0.060 3480 4811 0 FLSGMRTJ171752.9+591241 17:17:52.95 +59:12:41.7 0.440 43 0.480 0.070 3478 2724 0 FLSGMRTJ171753.0+584608 17:17:53.01 +58:46:08.4 0.320 38 0.610 0.080 3477 1662 0 FLSGMRTJ171753.1+601508 17:17:53.11 +60:15:08.1 0.200 33 0.260 0.060 3477 5222 0 FLSGMRTJ171753.1+592330 17:17:53.14 +59:23:30.3 0.180 32 0.300 0.070 3477 3157 0 FLSGMRTJ171753.3+585502 17:17:53.37 +58:55:02.3 0.290 35 0.370 0.070 3475 2018 0 FLSGMRTJ171753.5+594912 17:17:53.54 +59:49:12.2 0.350 40 0.410 0.070 3475 4185 0 FLSGMRTJ171753.7+594428 17:17:53.74 +59:44:28.2 0.230 41 0.180 0.060 3474 3995 0 FLSGMRTJ171753.7+592117 17:17:53.75 +59:21:18.0 0.180 34 0.190 0.060 3474 3069 0 FLSGMRTJ171753.9+594300 17:17:53.98 +59:43:00.5 0.520 35 0.570 0.060 3473 3937 1 FLSGMRTJ171754.7+593533 17:17:54.73 +59:35:33.4 0.200 35 0.170 0.050 3469 3639 0 FLSGMRTJ171754.7+600914 17:17:54.73 +60:09:14.8 0.670 32 0.890 0.070 3469 4986 0 FLSGMRTJ171754.8+601939 17:17:54.84 +60:19:39.4 0.160 29 0.180 0.050 3469 5403 0 FLSGMRTJ171755.0+585933 17:17:55.03 +58:59:33.7 0.290 44 0.560 0.110 3467 2199 0 FLSGMRTJ171755.0+594300 17:17:55.05 +59:43:00.5 0.190 35 0.130 0.040 3467 3937 1 FLSGMRTJ171755.2+604141 17:17:55.29 +60:41:41.4 0.420 76 0.270 0.100 3467 6284 0 FLSGMRTJ171755.2+585512 17:17:55.29 +58:55:12.8 0.360 33 0.410 0.060 3465 2025 0 FLSGMRTJ171755.3+592846 17:17:55.32 +59:28:46.7 0.280 30 0.440 0.060 3466 3368 0 FLSGMRTJ171755.6+594815 17:17:55.62 +59:48:15.4 0.220 38 0.180 0.050 3464 4147 0 FLSGMRTJ171755.6+593421 17:17:55.63 +59:34:21.6 0.180 29 0.120 0.040 3464 3591 0 FLSGMRTJ171755.8+600519 17:17:55.83 +60:05:19.3 0.190 34 0.260 0.070 3464 4829 1 FLSGMRTJ171756.0+594431 17:17:56.08 +59:44:31.9 0.310 40 0.460 0.080 3462 3998 0 FLSGMRTJ171756.4+592839 17:17:56.40 +59:28:39.6 0.170 31 0.170 0.050 3460 3363 0 FLSGMRTJ171756.4+591744 17:17:56.41 +59:17:44.2 0.310 41 0.400 0.070 3460 2926 0 FLSGMRTJ171756.4+590005 17:17:56.49 +59:00:05.4 0.450 43 0.560 0.080 3459 2220 0 FLSGMRTJ171756.5+584809 17:17:56.51 +58:48:09.1 0.230 39 0.300 0.070 3459 1743 0 FLSGMRTJ171756.8+593631 17:17:56.85 +59:36:31.1 0.470 32 0.580 0.060 3458 3677 0 FLSGMRTJ171756.9+601401 17:17:56.90 +60:14:01.1 0.190 33 0.210 0.050 3459 5177 0 FLSGMRTJ171757.5+602115 17:17:57.54 +60:21:15.7 0.200 32 0.290 0.060 3456 5467 0 FLSGMRTJ171757.6+594247 17:17:57.61 +59:42:47.4 0.310 38 0.310 0.060 3454 3928 0 FLSGMRTJ171757.6+584555 17:17:57.65 +58:45:56.0 0.670 37 0.930 0.080 3453 1654 0 FLSGMRTJ171757.7+593136 17:17:57.80 +59:31:36.3 0.380 33 0.390 0.050 3453 3481 0 FLSGMRTJ171757.7+591527 17:17:57.80 +59:15:27.4 0.200 38 0.140 0.050 3453 2835 0 FLSGMRTJ171757.9+585604 17:17:57.93 +58:56:04.3 0.240 39 0.260 0.070 3452 2059 0 FLSGMRTJ171758.1+604204 17:17:58.18 +60:42:04.6 0.450 77 0.420 0.110 3453 6299 0 FLSGMRTJ171758.2+585811 17:17:58.26 +58:58:11.9 0.230 39 0.360 0.080 3450 2145 0 FLSGMRTJ171758.3+602850 17:17:58.37 +60:28:50.8 0.300 48 0.400 0.090 3452 5770 0 FLSGMRTJ171758.7+590407 17:17:58.76 +59:04:07.4 0.260 45 0.230 0.060 3448 2382 0 FLSGMRTJ171758.9+602307 17:17:58.93 +60:23:07.7 0.260 36 0.410 0.070 3449 5542 0 FLSGMRTJ171758.9+593454 17:17:58.96 +59:34:54.5 0.150 29 0.130 0.040 3447 3613 0 FLSGMRTJ171759.0+592438 17:17:59.08 +59:24:38.9 1.910 33 2.280 0.070 3446 3202 0 FLSGMRTJ171759.1+584435 17:17:59.11 +58:44:35.8 0.440 39 0.680 0.080 3445 1600 0 FLSGMRTJ171759.3+594325 17:17:59.38 +59:43:25.7 0.210 39 0.460 0.100 3445 3954 0 FLSGMRTJ171759.4+590132 17:17:59.40 +59:01:32.1 0.260 38 0.230 0.050 3444 2278 0 FLSGMRTJ171759.3+595246 17:17:59.40 +59:52:46.7 0.980 42 1.250 0.090 3446 4328 0 FLSGMRTJ171759.4+590256 17:17:59.43 +59:02:56.4 0.210 40 0.290 0.070 3444 2334 0 FLSGMRTJ171759.8+594210 17:17:59.84 +59:42:10.8 0.220 37 0.230 0.060 3443 3904 0 FLSGMRTJ171800.0+584604 17:18:00.04 +58:46:04.5 0.250 40 0.260 0.060 3440 1660 0 FLSGMRTJ171800.1+594925 17:18:00.16 +59:49:25.1 0.250 42 0.360 0.080 3442 4193 0 FLSGMRTJ171800.3+590207 17:18:00.31 +59:02:07.8 1.220 40 1.510 0.090 3440 2302 0 FLSGMRTJ171800.3+585142 17:18:00.35 +58:51:42.7 0.190 37 0.330 0.080 3439 1885 0 10 T. Garnet al. 140 2 120 1.8 100 N 6800 Correction factor 11..46 40 1.2 20 0 1 -1 -0.5 0 0.5 1 1.5 2 4 6 8 10 12 14 16 18 20 alpha Peak signal-to-noise ratio Figure12.Radiospectralindexbetween610MHzand1.4GHz FigureA1.Fluxdensitycorrectionfactorforpointsources ChapmanS.,BlainA.,SmailI.,IvisonR.,2005,AJ,622,772 data,sothatitwasonthesamefrequencyscaleastheUSBpoints, CondonJ.,1992,ARA&A,30,575 andsothesidebandscouldbecombinedbeforeimaging. Condon J., Cotton W., Yin Q., Shupe D., Storrie-Lombardi L., (ii) The coordinates of sources originally found in our GMRT HelouG.,SoiferB.,WernerM.,2003,AJ,125,2411 imageswereseentobeslightlyrotatedneartheedgeofeachpoint- Fadda D., Jannuzi B., Ford A., Storrie-Lombardi L., 2004, AJ, ingcomparedwiththeirVLApositions.Thiswasduetoincorrect 128,1 time-stampsbeing usedinthe GMRTonlinesoftware, leading to FaddaD.,etal.,2006,AJ,131,2859 a slight error in the uv data. We wrote a customised AIPS task, FazioG.,etal.,2004,ApJS,154,10 UVFXT,toincreasethetime-stampsby7sandcorrectlyrecom- FrayerD.,etal.,2006,AJ,131,250 putetheuvwco-ordinates.Thiswasperformedbeforethefinalim- GarrettM.,2002,A&A,384,19 ageswerecreated. Gruppioni C., Pozzi F., Zamorani G., Ciliegi P., Lari C., Cal- abreseE.,LaFrancaF.,MatuteI.,2003,MNRAS,341,1 A2 Primarybeamcorrection HelouG.,SoiferB.,Rowan-RobinsonM.,1985,AJ,298,7 Hopkins A., Miller C., Connolly A., Genovese C., Nichol R., During the preliminary analysis of the data, we compared the WassermanL.,2002,AJ,123,1086 propertiesofseveral hundred relativelybright sourceswithpeaks Hughes A., Wong T., Ekers R., Stavely-Smith L., Filipovic M., greaterthan1mJybeam 1,visibleintheoverlappingregionsbe- − MaddisonS.,FukuiY.,MizunoN.,2006,MNRAS,370,363 tweentwopointings.Thiscomparisonrevealedasystematicdiffer- KanthariaN.,&Rao,A.,2001,GMRTTechnicalNoteR00185 encebetweentheapparentbrightnessofsourcesinadjacentpoint- LacyM.,etal.,2005,ApJS,161,41 ings;sourceslocatedtothenorth-westofapointingwereconsis- LuoS.,WuX.,2005,ChJAA,5,448 tentlybrighterthanthesamesourcewhenviewedinthesouth-east Morganti R., Garrett M., Chapman S., Baan W., Helou G., ofanadjacentpointing.Wewereabletomodelthisastheeffective, SoiferT.,2004,A&A424,371 averagepointingcentreofthetelescopebeingoffsetby 2.5ar- MurphyE.,etal.,2006,AJ,638,157 cmininanorth-westdirection,comparedwiththenomin∼alpoint- PapovichC.,etal.,2006,AJ,132,231 ingcentre. Theamount and directionof theoffset wasconsistent RiekeG.,etal.,2004,ApJS,154,25 between all pairs of pointings. After applying the correction, the Shim H., Im M., Pak S., Choi P., Fadda D., Helou G., Storrie- systematiceffectswereremovedandther.m.s.fluxdensityerrors LombardiL.,2006,ApJS,164,435 werebelow10%neartheedgeoftheprimarybeam,comparedwith StoughtonC.,etal.,2002,AJ,123,485 nearer20%beforethecorrectionwasapplied. WernerM.,etal.2004,ApJS,154,1 Itisthoughtthattheprimarybeamoffsetdependsontelescope elevation.Ourscansofeachsourcewerenottakensymmetrically about sourcetransit, andthedetected offset islikelytobeanav- erageovertheindividualoffsetsofeachscan.Sourceswillhavea slightlydifferentpositionandfluxdensityineachscan,duetothis APPENDIXA: CORRECTIONSTOGMRTDATA elevation-dependanterror.Self-calibrationassumesthatthesource A1 Correctionstotheuvdata remainsconstantwithtime,andthisislikelytobethereasonwhy residualartefactsareseennearthebrightestsources. TwocorrectionsweremadetotheGMRTuvdatabeforethefinal imagesweremade. (i) Theuvcoveragewasincreasedbycombining thetwoside- APPENDIXB: SOURCEEXTRACTOR bandsintoasingledatafile.Thefrequencychannelsinthelower sideband(LSB)haveanegativefrequencyincrement,whiletheup- SourceExtractor calculates thefluxdensity of an object by sum- persideband(USB)hasapositiveincrement–thismeansthatthe ming all pixels greater than some user-defined threshold. For an two channels can not be combined directly using standard AIPS object to be included in our catalogue, we required it to have at tasks.WewroteanAIPStaskcalledUVFLPtore-ordertheLSB least fiveconnected pixels withbrightness above 2s , and a peak