HI study of extremely metal-deficient dwarf galaxies. I. The Nancay Radio Telescope observations of twenty-two objects PDF
Preview HI study of extremely metal-deficient dwarf galaxies. I. The Nancay Radio Telescope observations of twenty-two objects
Astronomy&Astrophysicsmanuscriptno.6137 (cid:13)c ESO2008 February5,2008 H study of extremely metal-deficient dwarf galaxies. I. The Nanc¸ay Radio Telescope observations of twenty-two objects S.A.Pustilnik1,2 andJ.-M.Martin2,3 1 SpecialAstrophysicalObservatoryRAS,NizhnijArkhyz,Karachai-Circassia,369167Russia 2 ObservatoiredeParis,GEPI,5,placeJ.Janssen,92195MeudonCedex,France 7 3 ObservatoiredeParis,USN,routedeSouesmes,18330Nanc¸ay,France 0 0 Received27.07.2006;Accepted 2006 2 n ABSTRACT a J Aims.The goal of this study is to measure parameters of the integrated H emission for twenty-two dwarf galaxies with oxygen abundance 12+log(O/H) in the range of 7.42 to 7.65, which are representatives of the eXtremely Metal-Deficient (XMD) galaxy 7 group.Someofthemareexpectedtobesimilartothewell-knowncandidatesforlocalyounggalaxies,IZw18andSBS0335−052 1 thathavemostoftheirbaryonmassintheformofneutralgas.Therefore,theH21-cmlineobservationsarecrucialtounderstanding theirgroupandindividualproperties. 1 Methods.TheNanc¸ayRadioTelescope(NRT)withtheupgradedfocalreceiverwasusedforobservationsofthe21-cmH-line.This v permittedthedetectionofthefaintestsourceswithrmsof∼1mJyper10.5kms−1resolutionelement. 5 Results.For eighteen detected galaxies we present the parameters of their integrated H line emission and describe the data on 0 individualobjectsinmoredetail.ForfourundetectedXMDgalaxies,wegiveupperlimitsontheir M(H).For70%ofthetwenty 5 studiednon-LSB(lowsurfacebrightness)XMDgalaxies,wefindevidence(bothfromHandopticaldata)fortheirinteractionwith 1 neighboringobjects.InthebriefdiscussionofthegroupHpropertiesoftheobservedsubsample(thetotalO/Hrangeisof0.23dex, 0 orafactorof1.7),weunderlinethebroaddistributionsoftheHmass(rangeisof2ordersofmagnitude),oftheratioM(HI)/L (of 7 B 1orderofmagnitude),andoftheblueluminosity(rangeisof2ordersofmagnitude).WealsoobtainedHparametersofsixgalaxies 0 thatdonotbelongtotheXMDsample.Asaby-product,wehavedetectedanH-objectwithV =188kms−1,whichisprobablya / hel h partofanewhigh-velocityHcloud(HVC234.3-16.8+208)nearourGalaxy,inthedirectionofAM0624–261. p Conclusions.ThesedataincreasethenumberofXMDgalaxieswithknownintegratedHparameters(orupperlimits)byafactorof - two.Thisallowsustoaddressstatisticalpropertiesofthisgroup,whichwillbepresentedinaforthcomingpaper. o r Keywords.galaxies:dwarf–galaxies:H–galaxies:starformation–galaxies:abundances–radiolines:galaxies t s a : v 1. Introduction and the luminosity-metallicity relations for other BCGs (e.g., i Pustilnik et al. 2003a; Kniazev et al. 2003). In addition, they X The oxygen abundances in the great majority of actively star- haveveryhighgasmass-fractionsµ = M /(M +M ):upto r forminglow-massgalaxies(BCGs-bluecompactandHgalax- 0.95–0.99(vanZeeetal.1998;Papagderosgetal.g2002;staPrustilnik a ies)correspondtometallicitiesof(0.1–0.5)Z 1 (e.g.,Terlevich ⊙ etal.2004b).ThesepropertiescanbeunderstoodiftheseBCGs et al. 1991; Izotovet al. 1992; Ugryumovet al. 2003; Kniazev areyoungobjects,withnon-cosmologicalages. et al. 2004; Salzer et al. 2005; Pustilnik et al. 2006c, 2006b, The number of known eXtremely Metal-Deficient (XMD, amongothers).SomeBCGs,duetotheirlowmetallicities,were 12+log(O/H)≤7.65)galaxieshasgrownto∼50duringthelast originally considered as probable young galaxy populations. decadethankstodedicatedsearches,conductedmainlywithob- However,ithasbeenshownthatthegreatmajorityofBCGscon- jective prism surveys of Emission Line Galaxies (e.g., Izotov tainameasurablenumberofoldstarsandthus,areoldsystems et al. 1992; Kniazev et al. 1998; Melbourne & Salzer 2004; (e.g., Papaderoset al. 1996; Doublier et al. 1999; Cairos et al. Ugryumov et al. 2003) and with spectroscopy of large num- 2003). Only a few BCGs with the lowest known metallicities beroffaintgalaxiesfromtheSloanDigitalSkySurvey(SDSS, (12+log(O/H)of7.12–7.54,orZ=Z /34–Z /13)ofthosewith ⊙ ⊙ presented, e.g., by Kniazev et al. 2003, 2004 and Izotov et al. wellstudieddeepphotometryshownoevidenceforstellarpop- 2006).Such‘unevolved’objectscompriseonlyasmallfraction ulationsolderthanoneGyr.TheyincludeIZw18(Papaderoset al.2002;Izotov&Thuan2004b;O¨stlin&Mouhcine2005);SBS of known galaxies in the nearby Universe. Most of the known XMDgalaxiesareoftheBCGtype. 0335–052EandW(Papaderosetal.1998;Pustilniketal.2004b; Izotov&Thuan2005),Tol65(Papaderosetal.1999),andprob- Differentevolutionaryscenarioscanleadtoverylowmetal- ablyDDO68(Pustilniketal.2005a).TheseBCGsshowanad- licityofgalaxyISM.Theyinclude:a)thelossofmetalsdueto ditional blue color excess and a luminosity excess, which are superwindsrelatedtopowerfulstarformation(SF)events(e.g., notaccountedforbythecolor-metallicity(Gusevaetal.2003a) Mac Low & Ferrara 1999); b) primordial gas inflow or inter- galacticcloudaccretion(e.g.,Matteucci&Chiosi1983;Combes Sendoffprintrequeststo:S.Pustilnike-mail:[email protected] 2005); c) the very slow astration and related metal production, 1 UpdatedZ corresponds to12+log(O/H)=8.66,accordingtothe characteristicofconditionsinsomeLSBgalaxies(LowSurface ⊙ recentdatabyAsplundetal.(2004). Brightness–e.g.,Legrandetal.2000),andfinally,d)trulyyoung 2 S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies galaxies in which the first SF episode took place less than ∼1 and SDSS J1121+0324.The formerwas suggested as a proba- Gyr ago. In particular, for a few LSB dwarfs with sufficiently blemerger,sowetriedtofindsomeadditionalindicationsinits bright H regions the measurements gave very low values of H profile. The latter was serendipitouslydetectedin H at the 12+log(O/H) of 7.3 to 7.5 (Ronnback & Bergvall 1995; van VLAbyHibbard&Sansom(2003).Themainparametersofall Zee et al. 1996; van Zee 2000). Several dwarf irregular galax- observed galaxies with references to the sources are shown in ies(DIG)fromtheLocalVolumewereclaimedasXMDobjects Table1.SomeB-magnitudeswerederivedfromthephotometry longago(e.g.,Skillmanetal.1989).Therecentdataconfirmed presentedintheSloanDigitalSkySurvey(SDSS)DataRelease theXMDclassificationforGR8andLeoA(vanZeeetal.2006), 4byAdelman-McCarthyetal.(2006),withtheuseofthetrans- DDO53(Pustilniketal.2003c)andSextansAandB(Kniazev formationequationfromgandr magnitudesasgivenbySmith etal.2005).AllnearbyXMDLSBdwarfsandDIGsstudiedwith etal.(2002). theColor-MagnitudeDiagram(CMD)methodappearedtohave a substantial fraction of stars belonging to an old stellar pop- ulation. Thus, they are old systems. In contrast to these “non- 3. Observationsandreduction active” dwarfs, the CMD data for the second most metal-poor The H-observations with the Nanc¸ay2 radio telescope (NRT) BCG I Zw 18 imply that it is probablya younggalaxy(Izotov with a collecting area of 200×34.5 m are character- & Thuan2004b; O¨stlin &Mouhcine2005;thisresulthasbeen ized by a half-power beam width (HPBW) of 3.7′ (East- questioned, however, by Momany et al. 2005). Therefore, low West) × 22′ (North-South) at declination δ=0◦ (see also ISMmetallicityis anecessarybutnotsufficientpropertyto as- http://www.obs-nancay.fr/nrt). The data were acquired signanobjectasacandidateforayounggalaxy. duringtheyears2002–2004.We usedthenewantenna/receiver The high-z dwarf galaxies are near still or below the limit system F.O.R.T. (e.g., Granet et al. 1997; Martin et al. 2002) of detectability with the most powerfulinstruments(e.g., Pello with improvedoverall sensitivity. The system temperature was et al. 2005; Yan et al. 2005; Bouwens & Illingworth 2006). ∼35 K for both the horizontal and vertical linear polarizations Local XMD galaxy properties are similar to those of young of a dual-polarization receiver. The gain of the telescope was dwarfgalaxypopulationsathighredshift,sotheirdetailedstudy 1.5K Jy−1 atdeclinationδ=0◦. Thenew8192-channelcorrela- is an effective way to probe early galaxy evolution. Since the torwasusedcoveringatotalbandwidthof12.5MHz.Thetotal cold/warmgasshouldbethemaincomponentofbaryonicmat- velocityrangecoveredwasabout2700kms−1,withthechannel terinunevolvedgalaxies,the21-cmHlineisoneoftheprimary spacingof1.3kms−1beforesmoothing.Theeffectiveresolution meanstostudytheirproperties. afteraveragingoffouradjacentchannelsandHanningsmooth- Ourmulti-wavelengthstudyoftheXMDgalaxysamplein- ing was ≈10.5. km s−1. The observationsconsisted of separate cludes long-slit spectroscopy, optical/NIR imaging, photome- cycles of ‘ON’ and ‘OFF’ integrations, each of 40 seconds in try, and H 21-cm spectral observations. We present the mea- duration. ‘OFF’ integrations were acquired at the target decli- surements of integrated H parameters for 22 galaxies with nation,usuallywiththeEastR.A.offsetof∼15′ ×cos(δ).Fora 12+log(O/H) = 7.42–7.65 here. These data are used for H few cases, when we clearlysuspected confusionfrom a galaxy mappingofsomeXMDgalaxiesattheGiantMeterwaveRadio nearan‘OFF’position,weselectedadifferent‘OFF’positionto Telescope (GMRT, India). In Sect. 2 we describe the sample. avoidpossibleconfusion. ObservationsanddatareductionaredescribedinSect.3.There- For the flux calibration we used a noise diode. Its power sults are presented in Sect. 4. In Sect. 5 we discuss the results was regularly monitored throughoutthe observationsby point- and draw preliminary conclusions. The analysis of ∼40 XMD galaxieswithavailableHdataisdeferredtoaforthcomingpa- ing at known continuum and line sources. The comparisonsof ourmeasuredfluxeswithindependentmeasurementsofthesame per. objectsbyothertelescopesindicatestheconsistencyoftheflux scaletowithin10%. 2. Thesample Withanrmsnoiseof∼1to7mJyperresolutionelementaf- ter smoothing(10.5kms−1),we obtaineda S/N ratioof 30–40 The list of XMD galaxies was compiled by Kunth & O¨stlin fortheHIlinepeakfluxdensitiesF ofthebrightestdetected peak (2000,hereafterKO).AsignificantnumberofnewXMDgalax- objects,whileforthefaintestoneswehaddetectionswithaS/N ies was addedin the recentpaperscited below.Metallicities of ratioof∼3–4.Totalintegrationtimespergalaxy(‘ON’+‘OFF’) thegalaxiesfromtheKOlist,basedonold,loweraccuracydeter- variedbetween1 and10 hours.Forfourof the twenty-twoob- minations,wererevisited,andfourofthemwereexcludedfrom served XMD galaxies, we obtained only upper limits on their theXMDgroup(Kniazevetal.2001;Pustilniketal.2002a). F ,andestimatedlimitsontheirintegratedHflux. peak Three of the 22 galaxiesare well known.AM 0624–261is The data was reduced using the NRT standard programs classified as a blue LSB galaxy (Ronnback & Bergvall 1994, NAPS and SIR, written by the telescope staff (see description 1995).Twoothers-Tol21andTol65areBCGsfromtheTololo on http://www.obs-nancay.fr/nrt/support). Horizontaland verti- survey(Masegosaetal.1994;Thuan&Izotov1997;Papaderos calpolarizationspectrawerecalibratedandprocessedindepen- et al. 1999; Fricke et al. 2001, and references therein). For dentlyandthenaveragedtogether.Theerrorestimateswerecal- thesethree,onlyAM0624–261haspublishedHdata(S/N∼4, culated following Schneider et al. (1986). The baselines were Gallagher et al. 1995). The remaining 19 galaxies were found generallywell-fitbyathirdorderorlowerpolynomialandwere recentlyasXMDobjects.SevenofthemarefromtheHSSsam- subtractedout.Forafewnearbyextendedgalaxieswithanangu- ple(Kniazevetal.1998;Pustilniketal. 1999,2006b),fourare lar size comparableto the NRT horizontalHPBW, a correction from the HSS-LM sample (Ugryumovet al. 2003; Pustilnik et forresolutionoftheobservedHfluxhasbeendone.Havingno al. 2006a), and one is from the SBS sample (Izotov & Thuan 1999). Six more galaxies were found in the SDSS (Kniazev et 2 The Nanc¸ay Radioastronomy Station is part of the Observatoire al.2003).OnegalaxyisfromtheKISSsample(Leeetal.2004). de Paris and is operated by the Ministe`re de l’Education Nationale No data onH parametersof these galaxieswere knownin and Institut des Sciences de l’Univers of the Centre National de la the periodofourNRT observations,exceptforHS0122+0743 RechercheScientifique. S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies 3 Table1.ParametersoftheXMDsampleandadditional(belowtheline)galaxies IAUstyle Other Type Coord.(2000.0) Angul. V B ‡ M0∗ O/H† Alternative opt tot B name nameor R.A. Dec. sizea kms−1 mag mag name prefix h m s ◦ ′ ′′ a ×b 25 25 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) 0017+1055 HS BCG 002021.35 +111220.7 ... 5630±30 18.21 –16.7 7.63 0122+0743 HS IP 012534.18 +075922.2 59×30 2926±30 15.51 –17.7 7.63 UGC993 J0133+1342 SDSS BCG 013352.56 +134209.4 ... 2599±20 18.15 –14.4 7.60 J0205–0949 SDSS SmPec 020549.13 –094918.1 96×48 1908±18 15.55 –16.6 7.60 KUG0203-100 0624–261 AM LSB 062616.99 –261556.3 49×26 495±17 16.01 –12.5 7.49 ESO489-056 0846+3522 HS BCG 084940.26 +351139.2 ... 2177±27 18.15 –14.6 7.65 0937+2949 KISSB23 dIrr 094012.85 +293528.8 36×12 543±11 16.33 –13.6 7.65 KUG0937+298 0940+4025 HS BCG 094336.34 +401118.1 ... 5310±40 18.55 –16.0 7.61 1013+3809 HS BCG 101624.59 +375444.5 ... 1185±20 16.05 –15.4 7.59 KUG1013+381 1033+4757 HS BCG 103625.11 +474152.3 ... 1541±21 17.75 –14.3 7.65 1059+3934 HS BCG 110209.81 +391845.3 ... 2995±70 17.25 –16.1 7.62 J1105+6022 SDSS BCG 110553.62 +602228.5 30×18 1340±20 16.65 –15.1 7.62 SBS1102+606 J1121+0324 SDSS BCG 112152.80 +032421.2 43×12 1223±38 16.95 –14.6 7.62 J1201+0211 SDSS BCG 120122.32 +021108.5 ... 986±20 17.94 –13.5 7.54 1159+545 SBS BCG 120202.36 +541550.1 ... 3537±40 18.75 –14.9 7.49 J1215+5223 SDSS LSB 121546.56 +522313.9 72×40 162±15 15.45 –12.9 7.42 CGCG269–049 1214–277 Tol BCG 121717.09 –280232.6 ... 7795±50 18.46 –16.9 7.54 Tol21 1223–359 Tol BCG 122546.96 –361400.6 ... 2811±30 17.67 –15.6 7.54 Tol65 1442+4250 HS BCG 144411.35 +423734.7 60×12 660±15 15.68 –14.9 7.63 UGC9497 1704+4332 HS BCG 170545.39 +432848.9 ... 2076±12 18.48 –14.5 7.55 2134+0400 HS BCG 213658.95 +041404.1 ... 5070±50 19.31 –15.1 7.44 2236+1344 HS BCG 223831.15 +140028.6 ... 6160±20 17.91 –16.9 7.50 1059+395 KUG BCG 110200.14 +391906.4 ... 3067±60 16.63 –16.8 8.11 1102+6038 MCG LSB 110534.05 +602206.6 57×34 – 16.5:5 –15.1 MCG+10-16-056 J1106+6015 MCG dIrr 110647.99 +601546.8 51×17 1273±3 16.35 –15.3 MCG+10-16-062 J1202+0215 LSB/dI 120220.94 +021558.5 ... – 18.85 – 1342+4210 HS BCG 134459.49 +415505.4 ... 3780±50 18.35 –15.5 7.78 2236+136 KUG GPair 223921.95 +135255.8 ... 5184±15 15.45 –19.0 (†)–Inunits12+log(O/H).DataarefromRonnback&Bergvall(1995);Leeetal.(2004);Kniazevetal.(1998); Kniazevetal.(2003);Frickeetal.(2001);Gusevaetal.(2003b),Pustilniketal.(2006a),Pustilniketal.(2006b); (‡)–Photometricdataarefrom:1Pustilniketal.inprep.;2Ronnback&Bergvall(1994);3Salzeretal.(2002);4Kniazev etal.(2003);5SDSSDR4;6Frickeetal.(2001);7Papaderosetal.(1999);8GildePazetal.(2003) (a)–Opticaldiameters(in′′,measuredatthe25B-mag.arcsec−2level); (∗)–CorrectedforA accordingtoSchlegeletal.(1998)andwiththedistancesfromTable2. B informationontheHspatialdistribution,weusedthegalaxies’ (6)givestheobservedH flux(the areaunderthe profile).The opticalsizesandfollowedtheproceduredescribedbyThuanet errorsforthevaluesinCols.4,5,and6arecalculatedsimilarly al. (1999). The latter takes into account the statistical relation tothemethodsuggestedbySchneideretal.(1986).Whereitis between the BCGs’ optical size and their characteristic size in appropriate,weshowinCol.(7)Hflux,correctedforthebeam H. resolution.Therespectiveformula(seebelow)wasdiscussedin detail,e.g.,byThuanetal.(1999).Thecorrectionforsourceex- tensioninthenorth-southdirectionisnegligiblesincetheNRT 4. Resultsandpreliminaryanalysis verticalbeam(FWHM=22′)ismuchlargerthanthelargestsize TheH-profilesofthestudiedgalaxies,smoothedto10.5kms−1 ofanyofthetargetgalaxies.Thecorrectionforextensioninthe are shown in Fig. 1. The related parameters are presented in east-westdirection(forNRTbeamof3.′7)isnecessaryforsome Table2.Thevelocityprofilesshowsignificantdiversityinform ofthetargetswithopticaldiameterslargerthan0.′5.We follow andwidth.Theyaredescribedindividuallyinmoredetailbelow. theworkbyThuan&Martin(1981),andassumethatneutralgas In Sect. 5 we briefly discuss some properties of the observed inthesegalaxiesisdistributedlikeanellipticalGaussian.Then: galaxiesasagroup. F =F [1+(a2sin2PA+b2cos2PA)/θ2]1/2, In Table 2 we present the following data. In Col. (1) – the c H galaxyIAUtypename.InCol.(2)-thecentralheliocentricve- where a and b are the FWHM diameters (in arc minutes) of locity of the H-profile, derived as the mid-point of profiles at theassumedellipticaldistributionof H,whichareadoptedfor 50%of thepeak,with itsrmsuncertainty.The relatedestimate BCGs/H galaxies (following Lee et al. 2002) as 2.0 times of of the distance, given in Col. (3), follows that of Karachentsev theopticalsizesa andb inTable1.PAisthepositionangle 25 25 etal.(2004)fortheirCatalogofNeighboringGalaxies,withthe ofthemajoraxis,andθistheHPBWoftheNanc¸aytelescopein adoptedvalueof H = 72 km s−1 Mpc−1. The observedwidths theEast-Westdirection.InCol.8wepresentthelogarithmofthe 0 of H-profiles at the levels of 50 and 20 % of the peak value totalHmassinsolarunits,derivedaccordingtothewell-known with their r.m.s. errors are given in Cols. (4) and (5). Column formulabelow. 4 S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies M =2.36×105F D2, Kniazevetal.2003;thelatterislikelybasedonthelightpicked HI c upbyanopticalfiberofonly3′′ aperture,thusmissingalotof where MHI is in M⊙, Fc is the object’s integrated flux in the light). With its absolute magnitude of M0 = −16.64, it is Jy km s−1, and D is the distance to the object in Mpc. Finally, somewhatbrightforagalaxywithsuchalowBmetallicity.While in Col. 9 we show the ratio M(HI)/LB (in solar units, MB⊙ = itcouldbeananalogoftheluminousXMDBCGHS0837+4717 5.48,orLB⊙ =2.12×1033ergs−1),whereM(HI)isfromCol.8, (Pustilnik et al. 2004a), it is worth noting that its cited uncer- andLBcorrespondstotheabsolutemagnitudeMBfromTable1. taintyofO/H(0.09dex)isquitelarge.Hence,itsXMDclassifi- Besides the target galaxies, we detected H emission from cationshouldbe confirmedwithhigherqualityspectraldata.It several galaxies appearing either in the ‘ON’ beam, or in the isworthwhilenotingthehighratioM(HI)/L =2.57,whichisthe B ‘OFF’ beam at the radial velocities outside the uncertainty secondhighestinthissample. ranges for target galaxies. We examined the respective sky re- gionsinNEDandformanycasesfoundacandidategalaxythat probablyappearedinourHspectrum.TheirobservedHfluxes, 4.4.AM0624–261=ESO489-056 correctedfortheoffsetattenuation,aregiveninCol.7.Belowwe givemoredetailsontheHemissionofindividualgalaxiesfrom The first H detection for this LSB galaxy was presented by Gallagher et al. (1995) with the nominal S/N ratio of 4.3. We thisprogram,aswellasotherinformationrelatedtothediscus- achieved an S/N ratio of ∼36. All but one of the profile pa- sionoftheirpropertiesinSect.5. rameters of the old and new data are consistent. Only the in- tegrated H-flux is higher on our data by a factor of 1.35. The 4.1.HS0122+0743=UGC993 mostreasonableexplanationofthisdifferenceisthe ratherlow S/Nratiooftheolddata.ThemeasuredW =27kms−1 (after This object is considered by some authors as a galaxy pair. Its 50 smoothingto5.2kms−1)translates(accordingtoStaveley-Smith morphologyindeedsuggeststhatwearewitnessingmergingof et al. 1992, with the adopted isotropic velocity dispersion σ = theEandWcomponents,whichseemalreadytobeincontact. 10 km s−1) to the disk maximal rotation velocity of 8 km s−1. We aimed to find more evidence for two different components An inclination correction is applied for i = 44◦ (Ronnback & in its H profile and probable high-velocity tails. Previous H Bergvall1994).ThetotalHmassofthisgalaxyof1.6×107M observations with a lower S/N ratio (Garcia et al. 1994; Lu et ⊙ isthelowestinthissubsample.However,thisiscompatiblewith al. 1993) indicatedthe separatefaintcomponentin the H pro- whatoneexpectsforitsverylowblueluminosity,whichisalso filewithδV =–90kms−1 relativetothatofthemainemission. thelowestinthesubsample.Theratio M(H)/L =0.85istypi- Therewas a hintin the Lu etal. (1993) profileof the verylow B calofLSBgalaxies(Schombertetal.2001).Wedetectanaddi- contrast broad velocity component. In our profile, the fainter tionalHsource,seenatV sufficientlyclosetothatoftheMilky componentwithδV =–75kms−1,W ∼35kms−1 iswellde- r 50 Way.ItisdiscussedinSect.5.3asapossiblehigh-velocitycloud tected,butnobroadcomponentis visible.The widthofthe H (HVC). profile of this object (W =50 km s−1) is rather narrow for its 50 high neutralgas mass (M(H)=2.7×109 M ). Accordingto the ⊙ optical morphology of this system, the narrow profile is prob- 4.5.HS0846+3522 ably not due to the object being seen close to face-on. The re- centGMRTHmappingofthissystemshowsclearevidenceof This galaxy is detected near the correct optical velocity of merginginthisXMDobject. 2177 km s−1 with a S/N ratio of only 3.4. Its velocity in the discovery paper (Pustilnik et al. 1999) is mistaken due to a misprint (given as 2481 km s−1 instead of the correct value of 4.2.SDSSJ0133+1342 2184km s−1). Parameterspresented in Table 2 have uncertain- ThegalaxyismarginallydetectedinH.ItsnarrowHprofileis ties of ∼50%. HS 0846+3522 has a disturbed morphology on consistentwithitslowopticalluminosityanda smallHmass. theDSS-2images,probablyduetointeractionwithanirregular galaxy at 50′′ NEE (∼7 kpc in projection, B-mag ∼19.6 from On the DSS-2 image this galaxy is compact (maximum extent is ∼16′′ or ∼3 kpc),with a brightcentralknot. Its morphology SDSS).Itsredshiftisunknown. isverydisturbed.Thegalaxyisalmostconnectedbyabridgeto a ∼3.5-mag (in B-band)fainter reddish galaxy at ∼13′′ N, and 4.6.KISSB23=KUG0937+298 lookslikeamergingsystem. This galaxy is situated at ∼6 Mpc, in the inner rim region of the nearby Lynx-Cancer void described in Pustilnik et al. 4.3.SDSSJ0205−0949=KUG0203−100 (2003b). Two more XMD galaxies are found within the same This object has a double-hornprofile, typical of disk galaxies, void (Pustilnik et al. 2003b, 2005a). While the blue luminos- with W , corresponding to the rotation velocity amplitude of ity of KISSB 23 is only a factor of 2.8 higher than that of 20 ∼70 km s−1, which is consistent with its optical classification AM 0624–261, this galaxy, with a similar H mass has a sig- in NED as SB(s)m pec. Optical morphology, especially in the nificantly broader H profile: W50 ∼70 km s−1. Even for an outerparts,isratherdisturbed,suggestingsignificantinteraction. edge-on case this translates to an amplitude of Vrot (according The disk is warped on both edges. A potential candidate for a toStaveley-Smithetal.1992)ofabout35kms−1incomparison disturber is a LSB galaxy, 3m.5 fainter (with M0 ∼ −12.5 if at to 8 km s−1 for AM 0624–261.Despite having similar param- thesamedistance),at∼1.′4SW(∼11.5kpcinpBrojection).This eters of Lopt and M(H), KISSB 23 is significantly more DM galaxy is the only candidate within the NRT beam that might dominatedthan AM 0624–261.This is related to the enhanced be responsiblefora hintof emission in the H profileat ∆V ∼ stabilityofitsISMtolocalintrinsicperturbations.However,this +100kms−1. galaxy shows enhanced SFR. The EW(Hβ) of its bright non- The total B magnitude of SDSS J0205−0949 (as recalcu- centralknotof ∼30Å (Lee etal. 2004), forthe adoptedmetal- latedfromitsSDSSphotometry)is15m.46(not B=18m.38,asin licityofz=0.001correspondstoanageofaninstantaneousstar- S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies 5 Fig.1.The NRT HprofilesS (in mJy)vs. V (kms−1) of allstudiedgalaxieswith the resolutionof 10.5km s−1. The vertical ν hel arrowsatthebottomwiththehorizontalbarsshowtheopticalvelocityandits±1σuncertaintyforweakdetectionsornon-detected objects. burst of ∼10 Myr (Leitherer et al. 1999). The origin of such a a factor of 2.5 lower than that of the outstanding XMD BCG non-centralSF“burst”isunclear.Thepossibilityofatriggerby Tol65. an external perturber can be investigated through H mapping of thisgalaxy.Its ratio M(H)/L = 0.98is ratherhigh fordIrr B galaxies. 4.8.HS1013+3809andUGC5540 4.7.HS0940+4025 This BCG presumably comprises a physical pair with a ∼2m.3 brighter(M = −17.7)Sc galaxyUGC5540at8′ tothe South B Thisgalaxyismarginallydetected(S/N ratioof2.5),so itspa- and∼0.′5totheWest(intotal∼50kpcinprojection).Thelatter rameters have large uncertainties. The object is very compact was observed in the H line with the 91m Green Bank (here- andelongated,with the totalextentonthe DSS-2 ∼9′′ (3 kpc). after GB) telescope and the NRT (Haynes & Giovanelli 1991; The SDSS image shows two very blue knots almost in contact Theureauetal.1998).Takingintoaccounttheconfusionofthe (∼3′′separation),verysimilarinappearancetoHS2236+1344, twogalaxiesintheNRTbeam,someusefulestimatesofthein- for which there are clear indications of a merger. The same tegrated H flux of HS 1013+3809 can be obtained. If we ac- can be attributed to HS 0940+4025.Thus, more detailed study count for flux attenuation due to an offset of UGC 5540 from of this BCG is necessary. For its rather small profile width the NRT pointing, we can estimate from the observed flux the (∼50 km s−1), this galaxy has one of the largest M(H), only residual that should be attributed to the H emission related to 6 S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies HS 1013+3809. In particular, the value of its H mass can be thesegalaxies(inpreparation)showsdisturbedHmorphology derived. inbothgalaxies,implyingstronginteractioninthesystem. The two cited measurements of UGC 5540 (F(H,GB)= 5.37±0.56andF(H,NRT)=6.20±1.20Jykms−1)arenotsig- 4.11.SDSSJ1105+6022anditsneighbors nificantly different. However, the 0.83 Jy km s−1 higher value of the NRT flux, in comparison to that measured at GB, may ThereissomeHsignalintheoff-sourceposition,whichlooks indicatesomecontributionfromHS1013+3809tothepointing as a negativeamplitude profile centered at V ∼1260 km s−1. hel towardUGC5540.DuetothenarrowerbeamforGB,thisisnot Wehavecheckedpossiblecandidategalaxies,andfoundthatthe thecaseforF(H,GB).Therefore,forfurtheranalysisweaccept mostprobableis MCG+10-16-062,a galaxyof B=16.3,which thatthecorrectF(H)ofUGC5540isgivenbyGBdata.Taking isat∆α = 7s.0(52′′)and∆δ=6.′7fromtheoff-sourcepointing. intoaccounttheattenuationfactor(of0.655)duetothepointing ItsradialvelocityinSDSSisgivenas1273±3kms−1.Itsoffset offsetbetweenthetwogalaxies,theexpectedcontributionfrom results in the attenuationfactor of 0.75.This galaxyis situated UGC 5540 at this position is F(H) = 3.52±0.37 Jy km s−1. at9.′4SEfromourtargetobject. F(H) = 3.23±0.34Jy km s−1 when the horizontal NRT beam Another potential source of confusion is a disturbed resolution effect on the extendedH emission of UGC 5540 is DIG/LSB galaxy of a comparable size and total magnitude, applied.Subtractingthisexpectedfluxfromthemeasuredvalue MCG+10-16-056(m ≈16.5), which is only at 2.′4 west from B of4.74±0.19Jykms−1,weestimatethefluxforHS1013+3809 the target XMD galaxy. Its redshift is unknown. Since this ofF(H)=1.51±0.39Jykms−1.Thederivedvalueisconsistent can also contribute to the measured H flux, we conducted an with the NRT measured flux of UGC 5540. Indeed, the con- NRTobservationwithpointingtoMCG+10-16-056.Itwasde- tribution of HS 1013+3809at the NRT pointing to UGC 5540 tected ata velocityof 1265km s−1, whichis veryclose to that willbe0.99±0.25,andafteradditionwithF(H)=5.37±0.56for of MCG+10-16-062. The contribution of the latter in the off- UGC5540fromGBdataresultsin6.36±0.61,veryclosetothe sourcepositionforthisobservationwasalsosignificant.Having NRTmeasuredvalueof6.20Jykms−1.TheestimatedfluxofHS these 3 results, we solve this system by iterating the fluxes for 1013+3809comprisesonly32%ofthemeasuredHflux.Tore- all three galaxies. This method converges within the observa- solvethisgalaxypairinHrequiresmappingwithabeam-width tionalerrors.Thederivedparametersofthesegalaxiesaregiven of∼1′. in Table 2. The XMD galaxy and these two new H detected dwarfs(attheprojectiondistancesof14and55kpc)presumably form an isolated ‘dwarf’ group. The preliminary results of the 4.9.HS1033+4757 GMRT mapping (in preparation) reveal H-line emission from allthreedescribedabovegalaxies,withHmorphologyofSDSS Thisverycompactandlowluminositygalaxyhasabrightknot J1105+6022andMCG+10-16-056typicalofamergingpair. thatissomewhatdisplacedfromthecenterofthemainbody.The outermorphologyisirregularwithmanyfilaments.ItsHprofile is rather broadand asymmetric,suggestingpossible confusion. 4.12.SDSSJ1121+0324=NGC3640A We have checked all sufficiently bright galaxies falling within This galaxy is well detected with a S/N ratio of ∼10. Its pro- themainNRTbeam.AllthoseforwhichthereareSDSSspectra file, resembling a double-horn one, is consistent with its op- appearedasbackgroundobjects. tical appearance as an inclined disk. This disk has two bright regions displaced from the center and rather disturbed outer 4.10.HS1059+3934andKUG1059+395 isophotes. Hibbard & Sansom (2003) show this galaxy (called NGC 3640A) to be a probable companion of elliptical galaxy The measured H flux for the NRT pointing at the position of NGC3640(ataprojecteddistanceof∼70kpc),whichbelongs HS1059+3934is1.39Jykms−1.ItsHprofileisquitenarrow togalaxygroupLGG233.TheH-lineVLAmapofthisregion (W =59 km s−1), indicating a small rotation velocity. The in- showsnoconfusinggalaxies.Accordingly,itsintegratedparam- 50 clination correction (i=62◦, sin i=0.88) is not large, at least as eters,derivedinthatwork,areconsistentwiththoseobtainedon it can be estimated from the DSS-2 image, where the axial ra- theNRTdata. tiob/a ≈0.5.Thenearestgalaxy,whichcouldpotentiallyaffect TheB-magnitudeof17m.91fromKniazevetal.(2003)relates theHsignalofHS1059+3934,isKUG 1059+395,anirregu- only to emission from the bright H region. Our independent largalaxywithabrightcentralknot,1m.2brighterthanthetarget estimate of its magnitude (from SDSS) of B=17m.88 is consis- galaxy,at1.′9totheWest(25kpcinprojection).Weobtainedthe tentwiththeirs.TheB-magnitudefortherestofthegalaxy(also opticalspectrumofKUG1059+395withtheSAO6mtelescope fromSDSSphotometry)is17m.18.Summingupthelightofboth andfoundthistobeastarburstgalaxywithradialvelocityclose objectsresultsinBtot=16m.92. tothatofHS1059+3934(Pustilniketal.2006b). To resolvethe potentialconfusion,we carriedoutNRT ob- 4.13.SDSSJ1201+0211 servationsof KUG 1059+395.The observedH profile is cen- teredatavelocityveryclosetothatforHS1059+3934.Asmall This galaxy with rather disturbed morphology is a probable extrafeatureis seen atthe highvelocityend.The derivedinte- merger with the fainter (B ∼20.4) companion at 17′′ NW (1.4 gratedflux(withouttheextrafeature)is1.38Jykms−1.Having kpc in projection). It lies at the southern edge of the Virgo two independentmeasurementsanda knownattenuationfactor Cluster. No known galaxies in NED with the radial velocities duetooffsetpositions(0.475)forbothpointings,wesolvedthis withinourvelocityrangeareexpectedtobeintheNRTbeam. systeminseveraliterations.Thetotalfluxis0.95Jykms−1 for Asignificantsignal,correspondingtoanobjectneartheoff- HS1059+3934and0.93Jykms−1forKUG1059+395.Theex- sourceposition,isdetectedatV of∼2110kms−1. Thelatter hel trafeaturedetectedforKUG1059+395,withthetotalfluxof0.3 falls within the range of the Virgo cluster members in the di- Jykms−1,isprobablyanartifact.TherecentGMRTmappingof rectionclosetoitscenter.However,forthedistantsouthernpe- S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies 7 Table2.Hobservedandderivedparametersofthesampleandadditional(belowtheline)galaxies IAUname V(HI) Dist.1 W W F F Log(M(HI) M(HI)/L 50 20 H c B kms−1 Mpc kms−1 kms−1 Jykms−1 Jykms−1 M M /L ⊙ ⊙ ⊙ (1) (2) (3) (4) (5) (6) (7) (8) (9) 0017+1055 5630±30 81.3 50‡ ... <0.20 <8.29 <0.39 0122+0743 2899±5 42.3 50±6 123±16 6.70±0.26 7.60±0.30 9.47 1.48 J0133+1342 2580±4 38.1 33±7 39±11 0.10±0.05 7.54 0.23 J0205−0949 1885±1 26.8 112±1 133±2 11.07±0.19 12.19±0.22 9.35 2.57 0624−261 491±1 5.0 27±1 42±2 2.69±0.08 2.79±0.08 7.22 0.85 0846+3522 2169±3 29.6 33±6 40±10 0.10±0.03 7.32 0.24 0937+2949 505±2 6.0 77±10 98±15 1.95±0.24 2.05±0.25 7.24 0.98 0940+4025 5353±8 74.2 40±16 53±24 0.20±0.08 8.42 0.70 1013+3809 1169±4 15.8 86±7 130±12 4.74±0.19 1.51±0.39 7.90 0.57 1033+4757 1541±9 21.8 86±7 144±29 1.32±0.15 8.02 2.32 1059+3934 3019±3 41.9 59±6 91±9 1.39±0.06 8.60 1.54 J1105+6022 1333±3 19.9 48±6 74±10 2.40±0.17 2.48±0.18 8.36 1.62 J1121+0324 1171±3 17.0 89±6 112±10 2.49±0.15 2.67±0.16 8.29 1.94 J1201+0211∗ 974±3 17.0 29±7 53±10 0.96±0.09 7.82 1.92 1159+545 3560±15 50.3 ... ... <0.10 <7.72 <0.43 J1215+5223 158±1 4.2 27±1 43±2 5.14±0.14 5.24±0.16 7.34 1.01 1214−277 7795±50 105.8 ... ... <0.10 <8.42 <0.28 1223−359 2790±3 36.3 40±6 56±9 2.13±0.24 8.82 2.70 1442+4250 647±1 10.6 85±2 99±2 6.37±0.14 7.05±0.16 8.24 1.77 1704+4332 2082±8 31.9 33±15 59±24 0.24±0.05 7.76 0.75 2134+0400 5090±4 73.9 25±9 33±14 0.12±0.05 8.19 0.76 2236+1344 6160±20 89.1 50‡ ... <0.15 <8.35 <0.25 HI0624−2614 188±2 0.02 37±5 58±7 0.77±0.06 >1.86 1059+395 3023±5 41.9 62±10 101±16 1.38±0.09 8.59 0.86 1102+6038 1265±8 19.0 43±9 70±15 2.30±0.17 2.58±0.19 8.34 1.52 J1106+6015 1252±8 19.0 57±9 80±15 2.10±0.18 8.25 1.02 J1202+0215 2107±8 27.1 33±16 76±24 0.83±0.10 0.97±0.12 8.23 4.5: HIJ1218−2801 7392±3 100.2 117±6 127±10 0.46±0.06 9.04 1342+4210 3850±21 54.6 70±42 120±67 0.40±0.10 8.32 1.21 2236+136 5122±10 75.8 105±16 120±25 0.30±0.08 0.63±0.17 8.91 0.15 1–DistancesadoptedsimilartoKarachentsevetal.(2004)withH =72kms−1Mpc−1. 0 ‡–Adoptedfromthetotalionizedgasvelocityamplitude ∗–AdoptedtobethedistanceofVirgocluster 0017+1055–σ =1.8mJy.FortheF(HI)upperlimitthevalue∆V=50kms−1isusedfromHαdata. noise J0205−0949–Correctionfactorof1.10,calculatedasforScgalaxyNGC5540,seeSect.4.3. 0624–2614–DistancefromKarachentsevetal.(2004) 0937+2949–DistanceisestimatedasinKarachentsevetal.catalog(2004),wherethisgalaxyisabsent. 1013+3809–ResolvedconfusionwithScgalaxyof∼2mag.brighter,at8′totheS,seetext. 1059+3934–ResolvedconfusionwithKUG1059+395at1.9′,seetext. J1105+6022–Inagroupofthreedwarfs,seetext. 1159+545–σ =1.0mJy. noise J1215+5223–Correctionfactorof1.02,accountingforHstructurefromGMRT(Begumetal.2006) 1214–277–σ =1.0mJy.FLASHJ121809.29−275219.5isprobablyamemberofthesamegroup. noise 1223–359–Agalaxyat2.′1totheEastand4′totheSouth. 2236+1344–σ =1.2mJy.FortheF(HI)upperlimitthevalue∆V=50kms−1isusedfromHαdata. noise HI0624−2614–HVC?AssumedD=20kpc.Unknowncorrectionforangularsize/offset. 1059+395–ObservedtoresolveconfusionforHS1059+3934.Seetext. 1102+6038–MCG+10-16-056,observedtoresolveconfusionforSDSSJ1105+6022. J1202+0215–LSB/dIgalaxynearJ1201+0211OFF-sourceposition.Fluxiscorrectedforoffsetattenuation. HIJ1218−2801–AsourcenearOFF-sourcepositionJ121830.74−280125.NoNEDcandidates. 2236+136–IncasethisisKUG2236+136seenintheOFF-sourceposition,theoffsetattenuationis of∼2.1.However,thiscanbeablendsinceitsvelocitiesinUZC(5188kms−1)andSDSS(5224kms−1) catalogsdiffersignificantlyfromthevaluefoundhere. ripherythisvelocityistoohighandthisHobjectisfarbehind terpart.However,its B ∼18.8(fromSDSS) impliesaveryhigh the Virgo cluster. Its W is ∼30 km s−1, suggesting very slow ratioM(HI)/L ∼4.5.Thisrequiresfurtherstudy. 50 B rotation or large inclination. There are two candidate faint op- ticalgalaxieswithintheNRT beamnearthisposition.Thefirst one is an edge-ondisk, at 100′′ N, 25′′ W, and the second is a face-onLSBG,at288′′ N,25′′ W.Consideringtheverynarrow Hprofile,the‘face-on’LSBGgalaxyisamoreprobablecoun- 8 S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies 4.14.SDSSJ1215+5233=CGCG269−049 G1 (Papaderos et al. 1999) is about 5 magnitudes fainter and wouldgiveonlyaminorcontributiontotheobservedHflux. ThisgalaxybelongstotheCanisVenatici(CnV)Igalaxycloud atadistanceof∼4Mpc(e.g.,Karachentsevetal.2003).Itsnar- row H profile, with W50 =25 km s−1, is consistent with its 4.17.HS1442+4250=UGC9497 very low optical luminosity. The inclination correction for its b/a =0.3(Karachentsevetal.2004)issmall,∼1.03.Itsnearest This galaxyappearsto be an edge-ondisk with two prominent neighbor galaxy UGC 7298 (at the projected distance of 14.3 off-center H-regions and several fainter ones (Kniazev et al. kpc) has a similar radial velocity of 146 km s−1. The distance 1998;Gusevaetal.2003b).ItisoneofthenearestXMD BCG ofUGC7298of4.21Mpcisdeterminedthroughthetip ofthe known.ItsdeepV,Iphotometryandlong-slitspectroscopyshow RGB(Karachentsevetal.2004).Wethereforeconsiderthedis- noevidenceofastellarpopulationolderthan2Gyr(Gusevaet tancetoSDSSJ1215+5233tobeequalto4.2Mpcaswell.These al. 2003b). Its high S/N ratio H double-horn profile is rather galaxiesseemtobeaclosepairandtracersoftherecent/current wide,indicatingsignificantrotation.SomeasymmetryinitsH interactionmightbevisibleintheirHmorphology.Indeed,the profileatthehigh-velocityedgeof∼10%ofthepeakvaluemay H maps presented by Begum et al. (2006), give some hint of indicate a low mass companion, or a galaxy within the NRT disturbanceinbothgalaxiesoftheneutralgasmorphologyand beam. However, a better S/N ratio profile is necessary to con- itskinematics. firmthis.Thereisafainterredgalaxyalmostincontact,butits radialvelocityisunknown,sothismightbeabackgroundobject. 4.15.Tol1214−277=Tol21 4.18.HS1704+4332 This BCG is the most distant in our sample. Despite the long integration and the low noise (σ ∼1 mJy for the effective ve- This galaxy has quite a narrow profile, consistent with its low locityresolutionof10.5kms−1),no signalwas detectedin the luminosity. The galaxy is situated in the region of low density spectrum at or near the optical velocity.However,we have de- luminousgalaxies(L& L )withthenearestonebeingsituatedat ∗ tected an object near the off-source position (a negative signal 4.7h−1 Mpc.ThenearestknownsubluminousUpdatedZwicky with a ‘double-horn’ profile) with W50 =117 km s−1, at Vhel Catalog(UZC)galaxyisat3h−1Mpc. of 7392km s−1. We have checked knowngalaxiesclose to the off-source beam position (72.5 s in RA, or 16′ to the East) and found no appropriate candidates in NED. Several galaxies 4.19.HS2134+0400 around Tol 21 have V of 7600–7740 km s−1, implying that hel Tol21probablybelongsto aloosegroup.IfthisH objectisa This is the most metal-poorBCG and the faintest galaxy in B- galaxywithM(HI)/L ∼1,itsmeasuredintegratedfluxF(HI)of bandinthissubsample.TheexpectedHfluxishencealsolow. B 0.46Jykms−1 correspondsto Btot ∼18.Accountingforaprob- Thereis a narrowpeak(W50=25km s−1) at a velocityclose to ableoffsetandalowervalueM(HI)/L ,itcanbeevenbrighter. theopticalone,butitsamplitudeisnotverysignificant,oftheor- B TheupperlimitfortheHmassofTol21isderivedassuming derof3σnoise.Itsmorphologyonthedeep B-bandimagelooks thatitsW <100kms−1 andthepeakvalueoftheH profile rather disturbed, with a plume at the W edge that suggests its 50 is lower than 2 mJy. The upper limit for the ratio M(H)/L is probablemergernature(Pustilniketal.2006a),sincenodisturb- B low(<0.28).FormostofotherXMDgalaxiesthisparameteris ing galaxies are present in its environment. The nearby small (significantly)larger.Thisimpliesthatthisobjectbelongstothe irregulargalaxy South of the BCG is a distant backgroundob- lower end of the XMD galaxy H mass function. Probably the ject. neutral gas “deficiency” is related to the detectability of Ly-α emission in this BCG. This galaxy is the only one among four 4.20.Non-detectedXMDgalaxies XMD BCGs in which Ly-α emission was detected in the HST observations(Kunthetal.1994;Thuan&Izotov1997).Thevery Of fourundetectedXMD galaxiesthree are the mostdistantin faintgalaxyG2withcloseVhel(Izotovetal.2004a)ataprojected thesample,withV intherangeof∼5600to7800kms−1.For hel distanceof14.5kpcisprobablyacompanionofTol21. thesefourgalaxies(HS0017+1055,SBS1159+545,Tol21,and HS 2236+1344),the Notes for Table 2 give estimates of σ noise 4.16.Tol1223−359=Tol65 fortheresolutionof10.5kms−1inthespectralregionsneartheir opticalvelocities.Toderiveupperlimitsoftheir M(H),weas- ThisgalaxyisquiteunusualsinceitsHprofileisrathernarrow sumedthattheirpeakfluxdensitiesF arelowerthanthevis- peak for its large H mass of 0.66×109 M⊙. The inclination correc- ible peak flux density near the optical velocity plus 2σnoise (in tionis difficulttoinfersince its opticalmorphologyisstrongly mJy), and that their profiles are boxcar function with W less 50 disturbedatouterisophotes(Papaderosetal.1999).Itresembles than 100 km s−1. Then their integrated H flux F(H), is less SBS0335−052E,whoseHmassisonlyafactorof1.2higher. thanF ×W .ForSBS1159+545thesituationiscomplicated peak 50 TheirtotalHlinewidthsarealsosimilar. sinceclosetotheopticalvelocitywehaveafaintnegativesignal, Tol 65 is situated at the outskirts of a galaxy group. indicatingpossibleconfusionwithanobjectinthe‘OFF’-source There are several galaxies with close radial velocities at an- beam. However,we failed to find a reliable opticalcounterpart gular distances of 38′ to 53′. One irregular galaxy, FLASH forthisnegativesignal.Ifthenegativepeakneartheopticalve- J122530.92−360714.5 (Kaldare et al. 2003), with V = locityistreatedasnoise,a3σupperlimitofthepeakfluxdensity hel 2781±56 km s−1 and m =15.7, at 7.′5 to the NW, could con- forSBS 1159+545is ∼1 mJy.We presentits parametersbased b tributesomeHemissiontotheobservedflux.However,dueto onthisvalueinTable2.ForW inthespectraofHS0017+1055 50 offset attenuation, its measured flux will be ∼9% of the actual andHS2236+1344weusedthefullamplitudeoftheionizedgas value.Evenfor M(H)/L ∼1, thisgalaxywould contributeno velocityasmeasuredintheHα-line(∼50kms−1)fromtheSAO B morethan20%to the totalH fluxofTol65.Theclose LSBD 6mtelescopelong-slitobservations(inpreparation). S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies 9 Theratio M(H)/L <0.25forHS2236+1344isoneofthe B lowestamongtheXMDBCGsobservedinthiswork.Thereare two additionalfactorsthatcouldaffectits value.Theybothare related to its probable ‘atypical’ nature. As the analysis of its morphologyshows (two bright, almost contacting star-forming knotswithdisturbedperipheryandplumes)andtheunusualion- ized gas kinematics (Pramskij et al. 2003; Pustilnik et al. in preparation),thisobjectisverylikelyanadvancedmerger.The strengthsof starburstsin mergersare the largestin comparison toothercasessincethemainfractionofavailablegascanbein- volvedinSF.Thisimpliesthatinmergers,theM(H)/L ratiois B shifteddownfromits‘typical’value.Thesecondreasonisthat conditions in the ISM of merging gas-rich galaxies are favor- ablefortransformingthewholeHcloudpopulationtoH (e.g., 2 Elmegreen1993). Fig.2. Relation betweenM(H) andL forthe XMD galaxies B fromthispaper.Middledashedlinecorrespondstopositionsof 4.21.Non-XMDgalaxies objectswithM(H)/LB=1.0,whiletheupperandlowerlinesfor objectswiththisratioequal3and1/3,respectively.Thearrows Theparametersoftheobservedanddetectednon-XMDgalaxies showtheupperlimitsonM(H). arepresentedatthebottomofTable2.Wedividethemintothree categories. First:thetargetspreviouslyconsideredasXMDgalaxies,but variance of H parameters of the XMD galaxies and new evi- latereliminatedfromthisgroupduetobetterprecisionO/Hde- dence forthe importanceof interactionsto triggerstarbursts in termination.TheonlygalaxyinthiscategoryisHS1342+4210, theXMDBCGsample. withthevalueof12+log(O/H)=7.89.Thisisprobablyinteract- ingwithafaintercompanionat13′′ SEE,asseenontheDSS-2 image. 5.1.VarianceofHandopticalproperties Second:the targetsobservedtoresolvethepotentialorcer- tainconfusion,eitherinthe‘ON’-source,orinthe‘OFF’-source WebrieflysummarizethedatainTables1and2.Thetotalrange beam. They include KUG 1059+395, MCG+10-15-056 and ofoxygenabundancesinthesampleissmall,0.23dex,orafac- MCG+10-15-062. tor of 1.7. However, the total range of B-band luminosities in Third: new H sources which appeared occasionally in the thestudiedXMDgalaxiesismorethantwoordersofmagnitude ‘OFF’beamintherangeof ±1300kms−1 nearthevelocityof (M0B from –12.5 to –17.7, or log(LB/L⊙) = 7.19 to 9.27). The the target XMD galaxy. One of them, near the position of HS range of H mass in the sample XMD galaxies is comparable 2236+1344,couldbe tentativelyidentified (basedon the infor- to the optical luminosity range: log(M(H)/M⊙)=7.22 to 9.47. mationfromNED) withthe catalogedgalaxyKUG 2236+136. The latter implies that for galaxies in the very-low metallicity Anotherfaintgalaxy,J1202+0215,detectednearthepositionof regime this is not tightly connected with their other global pa- SDSSJ1201+0211(seeSect.4.13)isnotcataloged.Forthefaint rameters.Fortheblueluminosity,thiscanbepartlyunderstood sourceHI J1218−2801near Tol21 we did notfind anyoptical sincethestudiedgalaxysampleincludesobjectsofvarioustypes counterpart. -LSBDs,DIGs,BCGs,andevenmergingpairs.Thestrongstar- burstsinBCGscanbrightentheblueluminosityofaprogenitor galaxybyafactorof∼10ormore(e.g.,Kru¨ger1992),depend- 5. Discussionandconclusions ingonthestarburststrength,itsage,andthepreburstevolution- ary state. This brightening, however, could account only for a As mentioned in the introduction,four evolution scenarios can possiblyexplaintheexistenceofXMDgalaxies.Foroldgalax- part of this large range of LB. Figure 2 shows that the LB and ieswithverylowSFR(kindofLSBGs)andtrulyyounggalaxies M(H) of our XMD galaxy sample are correlated, and have a meanM(H)/L of1.15.Thescatterofthedataissmallerthana forwhichoneexpectstheclosed-boxtobevalid,thesmallrange B ofO/Hinoursamplecorresponds,accordingtoestimates,e.g., factorof∼2.4.However,fiveofthesamplegalaxiesshowvalues of Pilyugin & Ferrini (2000) to a gas mass fraction µ > 0.9. ofM(H)/LBthatarebelowthemeanbyafactorof3–5.Thelat- ter canindicateevolutionwithsignificantmetalloss, as forthe However,ifthelowmetallicityofagalaxyISMisacquireddue nearbyoldXMDDIGsSextansAandB,withM(H)/L <0.2. toexchangewiththeintergalacticmedium,thegasmass-fraction B can be significantly lower, e.g., M . M . The estimate of HI stars H mass from the integrated H flux is quite straightforward. 5.2.InteractionsasaSFtriggerinXMDBCGs However, the estimate of the total stellar mass in star-bursting galaxiesrequirestheseparationofthefluxfromtheyoungpop- The majority of known XMD galaxies are BCGs, low-mass ulationandsurroundingHregionsthatoftendominatehiding galaxies with sufficiently strong SF activity. One of the im- anold,underlyingcomponent.Thus,toaddresstheissueofthe portant questions on XMD BCGs is the nature of the trigger- gas-massfraction,weneedgoodsurfacephotometryofthestud- ing mechanisms for their SF bursts. Both external and intrin- iedobjects.Wejustmentionthatinadditiontowell-knownvery sicmechanismsareproposedtoberesponsibleforthestarbursts gas-richXMDgalaxiesIZw18andSBS0335–052,theHdata observedin BCGs in general, e.g., Salzer & Norton (2000)ar- for Tol 65 from this paper, along with the surface photometry gue that BCG starbursts are due to an intrinsic trigger in some fromPapaderosetal.(1999),indicatethatthisgalaxybelongsto specificBCGprogenitors–gas-richgalaxieswithenhancedgas thesameclass.Wepostponetherelatedanalysisbasedonphoto- concentrationindex.However,ifsomeXMDgalaxiesarereally metricdatatoaforthcomingpaper.Belowwediscussthelarge young,theexternaltriggerduetogravitationalinteractionswith 10 S.A.Pustilnik&J.-M.Martin:Hin22extremelymetal-deficientgalaxies other galaxies appears more natural, otherwise it is not clear range well. Note, that for an extendedsource with a character- why they did not start SF much earlier. The interaction trig- istic size of &10′,a largereductionofthe measuredsignalwill gerseemstobesufficientlyimportantforstarburstsinBCGsin occurduetosubtractionofthesourcesignalinthe‘OFF’beam general (e.g., Taylor et al. 1993, 1995; Pustilnik et al. 2001b). (hereat13.′5).Thus,allparametersofHI0624–2614areconsis- Therefore,itisnaturaltocheckwhethersufficientlyclosecom- tentwiththehypothesisthatitisa portionofaHVCsimilarto panions/neighborsofvariousmassesexistinthevicinityofour severalsuchnearbyobjects. targetgalaxies. The majority of galaxies in the studied sample appear to 5.4.Conclusions havevariousindicationsofstrongerorweakerinteraction.First of all there are direct indications of close companions(both in 1. We have measured the integratedH parametersfor 18 ex- terms of projecteddistance and radialvelocity).This relates to tremely metal-deficient (12+log(O/H) ≤7.65) dwarf galax- HS1059+3934andKUG1059+395at25kpc,HS1013+3809 ies, including15measuredforthe firsttime. Forfourmore andUGC5540at50kpc,andSDSSJ1105+6022andMCG10- XMD galaxieswe obtaineduseful upper limits on their H 16-056at 15 kpc. SDSS J1121+0324is a probablecompanion massandratioM(H)/L . B ofE3galaxyNGC3640at74kpcanda memberofthegalaxy 2. For the total range of O/H of the studied XMD galaxiesof group LGG 233. SDSS J1215+5223is paired with UGC 7298 only1.7(0.23dex),the distributionoftheirglobalparame- at 14 kpc.Tol 21has a tinycompanionat 14 kpc (Izotovet al. tersvarywidely.TherangeofM(H),aswellasofbluelumi- 2004a).OnemoreXMDBCGHS0822+3542hasacompanion nosity,istwoordersofmagnitude(from∼2×107to∼3×109 LSBGat11kpc(Pustilniketal.2003b;Chengaluretal.2006). M ,andM from-12.5to-17.7). ⊙ B Several other XMD galaxies HS 0122+0743, HS 2236+1344, 3. More than 2/3 of the twenty non-LSB XMD galaxies from andHS0837+4717(Pramskijetal.2003;Pustilniketal.2004a) this study show evidence of interaction/merger with other show clear evidence for various stages of a merger from both galaxies. theiropticalmorphologyandionizedgaskinematics. 4. InthedirectionofAM0624–261(l=234.3,b=−16.8),an Other XMD BCGs show only indications for interaction- H-objectisdetectedataradialvelocityV =+188kms−1 hel induced SF activity based on their morphology. In particular, (V = +208 km s−1), with an integrated H flux within the lsr HS 0846+3522andSDSS J1201+0211have a disturbedexter- NRT beam of 0.77 Jy km s−1 and W =37 km s−1. A com- 50 nalmorphologyandfaintergalaxiesintheirvicinity,suggesting parisonwithdataonnearbyhigh-velocityHclouds(HVC) ongoingmerging.HS2134+0400hasadisturbedmorphologyof showsthatthisobjectisprobablyapartofanewHVC(ten- theouterpartsandaplumeonthewesternedge.Anotherindica- tativenameHVC234.3-16.8+208). tionofinteractionisanasymmetricHprofile,whichisvisible inHS1033+4757. Acknowledgements. WearegratefultotheNRTComite´deProgrammesfortime allocation for this project. SAP is grateful to OdP and CNRS for support of Insummary,wenotethatthemajorityofXMDBCGsareei- his visits toObservatoire deParis where the mainpart ofthis work was per- thercertainlyorprobablyinteracting.OurHstudiesgavesev- formed. SAPacknowledges the partial support fromtheRFBRgrant No.06- eral new indications for the importance of tidal trigger to SF 02-16617.TheNanc¸ayRadioObservatoryistheUnite´ scientifiquedeNanc¸ay activity.However,moredetailedstudiesof Hmorphologyand oftheObservatoiredeParis,associatedasUnite´deServiceetdeRechercheNo. kinematics,aswellasthoseoftheionizedgasinthesegalaxies, B704withtheFrenchCentreNationaldelaRechercheScientifique.TheNanc¸ay Observatory alsogratefully acknowledges thefinancial supportoftheConseil arenecessarytounambiguouslysupportthisconclusion. Regional delaRe´gionCentreinFrance.Wethanktheanonymousrefereefor usefulcommentsandsuggestionsthatallowedustoimproveourpresentation. WeacknowledgetheuseoftheNEDandLEDAdatabasestochecktheinforma- 5.3.Probablenewhigh-velocitycloud tiononthepropertiesofthestudieddwarfgalaxiesandtheirenvironment,and theSDSSdatabasefortheuseofphotometryformanyofthestudiedgalaxies A well-detected H source in the ‘ON’ beam in the direction andofspectralinformationonsomeobjectsintheirenvironment. of AM 0624–261,at V = 188 km s−1 can be seen in Fig. 1. hel Its parameters are given in Table 2 under the name HI 0624– 2614.NocandidategalaxieswithintheNRTbeamthatcouldbe References counterpartsforthisHsourcearefound. Adelman-McCarthyJ.K.,etal.2006,ApJS,162,38 Therefore, the most probable interpretation of this feature Asplund, M.,Grevesse, N.,Sauval, A.J.,Allende Prieto, C.,&Kiselman, D., is the emission of a high-velocity H cloud (HVC) near the 2004,A&A,417,751 MilkyWay.IntheHVCnamesystem,itwouldbeHVC234.3- Begum, A., Chengalur, J., Karachentsev, I., Kajsin, S. & Sharina, M. 2006, MNRAS,365,1220 16.8+208 (which means galactic longitude and latitude in de- Bergvall,N.,Ro¨nnback,J.,Masegosa,J.,&O¨stlin,G.1999,A&A,341,697 grees and the local standard of rest velocity in km s−1). The Bouwens,R.,&Illingworth,G.2006,NewAstronomyReviews,50,152 nearestknownHVCstothispositionareHVC234.6-17.2+175, Cairos,L.M.,Caon,N.Papaderos,P.,etal.2003,ApJ,593,312 HVC 233.6-17.2+175, and HVC 233.8-18.7+212 from the Chengalur,J.N.,Giovanelli,R.,&Haynes,M.P.1995,AJ,109,2415 Chengalur,J.N.,Pustilnik,S.A.,Martin,J.-M.&Kniazev,A.Y.2006,MNRAS, HIPASS HVC catalog of Putman et al. (2002), at angular dis- 371,1849(PaperIII) tancesof0.5◦,0.8◦,and2.0◦,respectively.TheFWHMsoftheir Combes,F.2005,inStarbursts:From30DoradustoLymanBreakGalaxies,Ed. H profiles are 33–34 km s−1 in comparison to the value of byR.deGrijs&R.M.GonzalezDelgado,ASSLibrary,vol.329.Dordrecht: 37±5kms−1 measuredforHI0624–2614.ThesenearbyHVCs Springer,p.167 haveaxialratiosofa/b∼1.5,withthemajoraxisFWHMof0.8◦ Doublier,V.,Caulet,A.,&Comte,G.1999,A&AS,138,213 to 2.4◦ and peak brightnesstemperaturesT of 0.06 to 0.23 K. Elmegreen,B.G.1993,ApJ,411,170 b Ekta,Chengalur,J.N.,&Pustilnik,S.A.2006,MNRAS,372,853(PaperII) TheirtotalHfluxesvaryintherangeof9to97Jykms−1.The Fricke,K.,Izotov,Y.I.,Papaderos,P.,Guseva,N.G.,Thuan,T.X.2001,AJ,121, NRTbeamwithFWHM=0.37◦×0.06◦ israthersmallforsuch 169 HVCs and due to resolutionwill pickup only a fractionof the GallagherIII,J.S.,Littleton,J.E.,&Matthews,L.D.1995,AJ,109,2003 totalHflux.Thefluxlosscanvaryfromafactorof∼10tosev- Haynes,M.P.&Giovanelli,R.1991,ApJS,77,331 Hibbard,J.E.&Sansom,A.E.2003,AJ,125,667 eral hundred. Therefore, if this source is a part of a HVC, the Garcia,A.M.,Bottinelli,L.,Garnier,R.,Gouguenheim,L.,&Paturel,G.1994, measuredfluxofthisobject0.77Jykms−1 matchestheknown A&AS,107,265
The list of books you might like
