Contents
Images
Upload your image
DSS Images Other Images
Related articles
Outflows in Infrared-Luminous Starbursts at z < 0.5. II. Analysis and Discussion1,
We have performed an absorption-line survey of outflowing gas in 78starburst-dominated, infrared-luminous galaxies. This is the largeststudy of superwinds at z<~3. Superwinds are found in almost allinfrared-luminous galaxies, and changes in detection rate with SFR-windsare found twice as often in ultraluminous infrared galaxies (ULIRGs) asin less-luminous galaxies-reflect different wind geometries. The maximumvelocities we measure are 600 km s-1, though most of theoutflowing gas has lower velocities (100-200 km s-1). (Onegalaxy has velocities exceeding 1000 km s-1.) Velocities inLINERs are higher than in H II galaxies, and outflowing ionized gasoften has higher velocities than the neutral gas. Wind properties(velocity, mass, momentum, and energy) scale with galaxy properties(SFR, luminosity, and galaxy mass), consistent with ram-pressure drivingof the wind. Wind properties increase strongly with increasing galacticmass, contrary to expectation. These correlations flatten at high SFR(>~10-100 Msolar yr-1), luminosities, andmasses. This saturation is due to a lack of gas remaining in the wind'spath, a common neutral gas terminal velocity, and/or a decrease in theefficiency of thermalization of the supernovae energy. It means thatmass entrainment efficiency, rather than remaining constant, declines ingalaxies with SFR>10 Msolar yr-1 andMK<-24. Half of our sample consists of ULIRGs, which hostas much as half of the star formation in the universe at z>~1. Thepowerful, ubiquitous winds we observe in these galaxies imply thatsuperwinds in massive galaxies at redshifts above unity play animportant role in the evolution of galaxies and the intergalacticmedium.Some of the observations reported here were obtained at the MMTObservatory, which is a joint facility of the Smithsonian Institutionand the University of Arizona.Some of the observations reported here were obtained at the Kitt PeakNational Observatory, National Optical Astronomy Observatory, which isoperated by the Association of Universities for Research in Astronomy,Inc. (AURA), under cooperative agreement with the National ScienceFoundation.
| Outflows in Infrared-Luminous Starbursts at z < 0.5. I. Sample, Na I D Spectra, and Profile Fitting1,
We have conducted a spectroscopic survey of 78 starburstinginfrared-luminous galaxies at redshifts up to z=0.5. We usemoderate-resolution spectroscopy of the Na I D interstellar absorptionfeature to directly probe the neutral phase of outflowing gas in thesegalaxies. Over half of our sample are ultraluminous infrared galaxiesthat are classified as starbursts; the rest have infrared luminositiesin the range log(LIR/Lsolar)=10.2-12.0. The sampleselection, observations, and data reduction are described here. Theabsorption-line spectra of each galaxy are presented. We also discussthe theory behind absorption-line fitting in the case of a partiallycovered, blended absorption doublet observed at moderate-to-highresolution, a topic neglected in the literature. A detailed analysis ofthese data is presented in a companion paper.Some of the observations reported here were obtained at the MMTObservatory, a joint facility of the Smithsonian Institution and theUniversity of Arizona.Some of the observations reported here were obtained at the Kitt PeakNational Observatory, National Optical Astronomy Observatory, which isoperated by the Association of Universities for Research in Astronomy,Inc. (AURA), under cooperative agreement with the National ScienceFoundation.
| The distribution of atomic gas and dust in nearby galaxies - III. Radial distributions and metallicity gradients
The radial distribution of dust and gas in 38 nearby galaxies isinvestigated, using a sample of galaxies for which matched resolution(25 arcsec) neutral hydrogen (HI) and 850-μm images are available.Most of these radial profiles are fitted well by an exponential model,and the derived 850-μm scalelengths are proportional to the HIscalelengths. From this relation, it is found that the metallicitygradients of these galaxies are much shallower than previous studies,unless the dust temperature is constant within the disc, or asignificant component of molecular gas exists at large radii that is nottraced by CO observations.
| Revised masses of dust and gas of SCUBA Local Universe Survey far-infrared bright galaxies based on a recent CO survey
Recent CO measurements of an essentially complete subsample of galaxiesfrom the SCUBA Local Universe Survey (SLUGS) are used to examine theirimplications for dust and gas masses in this sample. Estimates of dustmasses are affected by a contribution to the SCUBA brightnessmeasurements by CO(3-2) emission, and molecular gas masses by the use ofa modified value of the CO-to-H2 conversion factor X. Theaverage dust mass is reduced by 25-38 per cent, which has no bearing onearlier conclusions concerning the shape of the dust mass luminosityfunction derived from the SLUGS. The value of X found from the COsurvey, when applied together with the reduction in dust masses, leadsto lower estimates for the mean gas-to-dust mass ratios, where the gasincludes both H2 and H I. For the CO sample, the mean globalratio is reduced from approximately 430 to about 320-360, but is furtherreduced to values near 50 when applied to the nuclear regions relevantto the CO observations. We discuss these results and suggest that thedifferences between the nuclear and outer regions may simply reflectdifferences in metallicity or the existence of considerable amounts ofunobserved cold dust in the outer regions of these galaxies.
| The Superwind Galaxy NGC 4666: Gravitational Interactions and the Influence of the Resulting Starburst on the Interstellar Medium
We present high-resolution observations of the atomic and molecular gasphase of the ``superwind'' galaxy NGC 4666. Deep Very Large Array (VLA)H I observations of NGC 4666 and its surroundings reveal the presence ofprominent tidal arms, which provide clear evidence that the galaxy isinteracting with its neighbor NGC 4668 and a newly discovered dwarfcompanion. This interaction is also evident from a dynamical analysis,which shows that the diffuse H I envelope around NGC 4666 (radii >19kpc) is kinematically altered with regard to the central disk. Thisinteraction has likely caused the starburst activity in NGC 4666. A cutthrough the H I emission distribution perpendicular to the major axisprovides no evidence for the existence of an H I halo: this distributioncan best be described by an inner disk and two outer spiral arms. Ourhigh-resolution Owens Valley Radio Observatory (OVRO) CO mosaic showsthat the molecular gas is distributed relatively uniformly in thecentral ~7 kpc; the total molecular gas mass is~1.0×1010 Msolar. A kinematicallyinteresting CO feature that can be interpreted as an expanding molecularsupershell is detected near the footpoint of one of the Hαoutflows at the turnover of galactic rotation. About 1% of the totalenergy input of the starburst would be needed to create this feature. Astudy of the CO (2-->1) to CO (1-->0) line-transition ratio(obtained with the IRAM 30 m and Swedish-ESO Submillimeter Telescope[SEST] telescopes) shows that there is no significant variation of themolecular gas excitation over the central galaxy disk[S(2-->1)/S(1-->0)~=0.85] on scales of ~2.5 kpc. This can beexplained by the unusually uniform star formation rate across NGC 4666'sdisk, which sets this system apart from most other starburst systems.The total amount of molecular and atomic gas isMmol~=3.0×1010Msolar, whichimplies that NGC 4666 can sustain many similar starburst episodes(consuming ~108 Msolar each) in the future. Twonew dwarf companions in the NGC 4666 group are detected, with adetection limit of 107 Msolar, over an area of~1.6×104 kpc2. We speculate that the huge HI envelope around NGC 4666 may represent a low-redshift counterpart fordamped Lyα systems seen at higher redshifts.
| A Study of the Distribution of Star-forming Regions in Luminous Infrared Galaxies by Means of Hα Imaging Observations
We performed Hα imaging observations of 22 luminous infraredgalaxies to investigate how the distribution of star-forming regions inthese galaxies is related to galaxy interactions. Based on correlationdiagrams between Hα flux and continuum emission for individualgalaxies, a sequence for the distribution of star-forming regions wasfound: very compact (~100 pc) nuclear starbursts with almost nostar-forming activity in the outer regions (type 1), dominant nuclearstarbursts <~1 kpc in size with a negligible contribution from theouter regions (type 2), nuclear starbursts >~1 kpc in size with asignificant contribution from the outer regions (type 3), and extendedstarbursts with relatively faint nuclei (type 4). These classes ofstar-forming regions were found to be strongly related to globalstar-forming properties, such as star formation efficiency, far-infraredcolor, and dust extinction. There was a clear tendency for the objectswith more compact distributions of star-forming regions to show a higherstar formation efficiency and hotter far-infrared color. An appreciablefraction of the sample objects were dominated by extended starbursts(type 4), which is unexpected in the standard scenario ofinteraction-induced starburst galaxies. We also found that thedistribution of star-forming regions was weakly but clearly related togalaxy morphology: severely disturbed objects had a more concentrateddistribution of star-forming regions. This suggests that the propertiesof galaxy interactions, such as dynamical phase and orbital parameters,play a more important role than the internal properties of progenitorgalaxies, such as dynamical structure or gas mass fraction. We alsodiscuss the evolution of the distribution of star-forming regions ininteracting galaxies.
| Dust masses and star formation in bright IRAS galaxies. Application of a physical model for the interpretation of FIR observations
We address the problem of modeling the far-infrared (FIR) spectrum andderiving the star-formation rate (SFR) and the dust mass of spiralgalaxies. We use the realistic physical model of Popescu et al.(\cite{popescu}) to describe the overall ultra-violet (UV), optical andFIR spectral energy distribution (SED) of a spiral galaxy. The modeltakes into account the 3-dimensional old and young stellar distributionsin the bulge and the disk of a galaxy, together with the dust geometry.The geometrical characteristics of the galaxy and the intrinsic opticaland near-infrared spectra are determined by the galaxy's observed K-bandphotometry. The UV part of the spectrum is assumed to be proportional tothe SFR through the use of population synthesis models. By solving theradiative transfer equation, we are able to determine the absorbedenergy, the dust temperature and the resulting FIR spectrum. The modelhas only three free parameters: SFR, dust mass, and the fraction of theUV radiation which is absorbed locally by dense dust in the HII regions.Using this model, we are able to fit well the FIR spectra of 62 brightIRAS galaxies from the ``SCUBA Local Universe Galaxy Survey" of Dunne etal. (\cite{dunne1}). As a result, we are able to determine, amongothers, their SFR and dust mass. We find that, on average, the SFR (inabsolute units), the star-formation efficiency, the SFR surface densityand the ratio of FIR luminosity over the total intrinsic luminosity, arelarger than the respective values of typical spiral galaxies of the samemorphological type. We also find that the mean gas-to-dust mass ratio isclose to the Galactic value, while the average central face-on opticaldepth of these galaxies in the V band is 2.3. Finally, we find a strongcorrelation between SFR or dust mass and observed FIR quantities liketotal FIR luminosity or FIR luminosity at 100 and 850 μm. Thesecorrelations yield well-defined relations, which can be used todetermine a spiral galaxy's SFR and dust-mass content from FIRobservations.
| CO Molecular Gas in Infrared-luminous Galaxies
We present the first statistical survey of the properties of the12CO(1-0) and 12CO(3-2) line emission from thenuclei of a nearly complete subsample of 60 infrared (IR) luminousgalaxies selected from SCUBA Local Universe Galaxy Survey (SLUGS). Thissubsample is flux limited at S60μm>=5.24 Jy with far-IR(FIR) luminosities mostly at LFIR>1010Lsolar. We compare the emission line strengths of12CO(1-0) and (3-2) transitions at a common resolution of~15". The measured 12CO(3-2) to (1-0) line intensity ratiosr31 vary from 0.22 to 1.72, with a mean value of 0.66 for thesources observed, indicating a large spread of the degree of excitationof CO in the sample. These CO data, together with a wide range of dataat different wavelengths obtained from the literature, allow us to studythe relationship between the CO excitation conditions and the physicalproperties of gas/dust and star formation in the central regions ofgalaxies. Our analysis shows that there is a nonlinear relation betweenCO and FIR luminosities, such that their ratioLCO/LFIR decreases linearly with increasingLFIR. This behavior was found to be consistent with theSchmidt law relating star formation rate to molecular gas content, withan index N=1.4+/-0.3. We also find a possible dependence of the degreeof CO gas excitation on the efficiency of star-forming activity. Usingthe large velocity gradient (LVG) approximation to model the observeddata, we investigate the CO-to-H2 conversion factor X for theSLUGS sample. The results show that the mean value of X for the SLUGSsample is lower by a factor of 10 compared to the conventional valuederived for the Galaxy, if we assume the abundance of CO relative toH2, ZCO=10-4. For a subset of 12galaxies with H I maps, we derive a mean total face-on surface densityof H2+HI of about 42 Msolar pc-2 withinabout 2 kpc of the nucleus. This value is intermediate between that ingalaxies like our own and those with strong star formation.
| COLA. II. Radio and Spectroscopic Diagnostics of Nuclear Activity in Galaxies
We present optical spectroscopic observations of 93 galaxies taken fromthe infrared-selected COLA (compact objects in low-power AGNs) sample.These are all galaxies for which we have previously obtainedlow-resolution radio observations and high-resolution (<0.05")Australian Long Baseline Array snapshots. The sample spans the range offar-IR luminosities from normal galaxies to luminous infrared galaxiesand contains a significant number of galaxies involved in galaxy-galaxyinteractions. Of the galaxies observed, 78 (84%) exhibit emission linesindicating that they are either AGNs or actively forming stars(starburst galaxies). Using a newly developed, theoretically based,optical emission line scheme to classify the spectra, we find that 15%of the emission-line galaxies are Seyfert galaxies, 77% are starbursts,and the rest are either borderline AGN/starburst or show ambiguouscharacteristics. We find little evidence for an increase in the fractionof AGNs in the sample as a function of far-IR (FIR) luminosity, incontrast to previous studies, but our sample covers only a small rangein infrared luminosity(1010.5Lsolar<=LFIR<=1011.7 Lsolar), and thus a weak trend may be masked. Instead,as the infrared luminosity increases, so does the fraction of metal-richstarbursts, objects that on more traditional diagnostic diagrams wouldhave been classified as weak, low-ionization, narrow emission lineregions. As a whole the Seyfert galaxies exhibit a small, butstatistically significant, radio excess on the radio-FIR correlationcompared to the galaxies classified as starbursts. Compact (<0.05")radio cores are detected in 55% of the Seyfert galaxies, and thesegalaxies exhibit a significantly larger radio excess than the Seyfertgalaxies in which compact cores were not detected. Our results indicatethat there may be two distinct populations of Seyfert galaxies,``radio-excess'' Seyfert galaxies, which exhibit extended radiostructures and compact radio cores, and ``radio-quiet'' Seyfertgalaxies, in which the majority of the radio emission can be attributedto star formation in the host galaxy. No significant difference is seenbetween the IR and optical spectroscopic properties of Seyfert galaxieswith and without radio cores.
| The IRAS Revised Bright Galaxy Sample
IRAS flux densities, redshifts, and infrared luminosities are reportedfor all sources identified in the IRAS Revised Bright Galaxy Sample(RBGS), a complete flux-limited survey of all extragalactic objects withtotal 60 μm flux density greater than 5.24 Jy, covering the entiresky surveyed by IRAS at Galactic latitudes |b|>5°. The RBGS includes629 objects, with median and mean sample redshifts of 0.0082 and 0.0126,respectively, and a maximum redshift of 0.0876. The RBGS supersedes theprevious two-part IRAS Bright Galaxy Samples(BGS1+BGS2), which were compiled before the final(Pass 3) calibration of the IRAS Level 1 Archive in 1990 May. The RBGSalso makes use of more accurate and consistent automated methods tomeasure the flux of objects with extended emission. The RBGS contains 39objects that were not present in the BGS1+BGS2,and 28 objects from the BGS1+BGS2 have beendropped from RBGS because their revised 60 μm flux densities are notgreater than 5.24 Jy. Comparison of revised flux measurements forsources in both surveys shows that most flux differences are in therange ~5%-25%, although some faint sources at 12 and 25 μm differ byas much as a factor of 2. Basic properties of the RBGS sources aresummarized, including estimated total infrared luminosities, as well asupdates to cross identifications with sources from optical galaxycatalogs established using the NASA/IPAC Extragalactic Database. Inaddition, an atlas of images from the Digitized Sky Survey with overlaysof the IRAS position uncertainty ellipse and annotated scale bars isprovided for ease in visualizing the optical morphology in context withthe angular and metric size of each object. The revised bolometricinfrared luminosity function, φ(Lir), forinfrared-bright galaxies in the local universe remains best fit by adouble power law, φ(L)~Lα, withα=-0.6(+/-0.1) and α=-2.2(+/-0.1) below and above the``characteristic'' infrared luminosityL*ir~1010.5Lsolar,respectively. A companion paper provides IRAS High Resolution (HIRES)processing of over 100 RBGS sources where improved spatial resolutionoften provides better IRAS source positions or allows for deconvolutionof close galaxy pairs.
| The infrared supernova rate in starburst galaxies
We report the results of our ongoing search for extincted supernovae(SNe) at near-infrared wavelengths. We have monitored at 2.2 mu m asample of 46 Luminous Infrared Galaxies and detected 4 SNe. The numberof detections is still small but sufficient to provide the firstestimate of supernova rate at near-infrared wavelengths. We measure a SNrate of SNNIR_r=7.6+/- 3.8 SNu which is an order of magnitudelarger than observed in quiescent galaxies. On the other hand, theobserved near-infrared rate is still a factor 3-10 smaller than thatestimated from the far-infrared luminosity of the galaxies. Amongvarious possibilities, the most likely scenario is that dust extinctionis so high (AV>30) to obscure most SNe even in thenear-IR.The role of type Ia SNe is also discussed within this context. We derivethe type Ia SN rate as a function of the stellar mass of the galaxy andfind a sharp increase toward galaxies with higher activity of starformation. This suggests that a significant fraction of type Ia SNe areassociated with young stellar populations.Finally, as a by-product, we give the average K-band light curve ofcore-collapse SNe based on all the existing data, and review therelation between SN rate and far-infrared luminosity.Based on observations collected at the European Southern Observatory,Chile (proposal 66.B-0417), at the Italian Telescopio Nazionale Galileo(TNG) operated on the island of La Palma by the Centro Galileo Galileiof the INAF (Istituto Nazionale di Astrofisica), and at the StewardObservatory 61'' telescope.
| The distribution of atomic gas and dust in nearby galaxies - I. Presentation of matched-resolution VLA H I and SCUBA 850-μm maps
We present matched-resolution VLA HI and SCUBA 850-μm maps of 20IRAS-bright galaxies. Of the galaxies observed, two were not detected inHI and two were detected in absorption. The HI distributions of thegalaxies have a range of morphologies. Some of the systems appear HIdeficient in the central regions which could be due to a high conversionrate of HI into molecules or HI absorption. In contrast to the HI, the850-μm emission has a smooth distribution which is concentratedtowards the optical centre of each galaxy. We also find evidence for850-μm emission extending to the periphery of the optical disc insome of the galaxies. Finally, we note that the relative lack of850-μm emission when compared with HI does not necessarily mean thatthe atomic gas and dust do not have similar mass distributions.
| Star Formation Rates in Interacting Starburst Galaxies
By narrowband imaging in Hα and in the adjacent red stellarcontinuum we have studied the rate and distribution of star formation in43 systems of luminous and ultraluminous IR galaxies currentlyundergoing interaction and merging. These galaxies are amongst the mostluminous at 60 μm and range in distance from ~50 up to 100 Mpc. Herewe present the Hα and the adjacent red-continuum narrowbandimages, and we compare the star formation rates derived from Hαwith those estimated from the IR luminosity. We find clear evidence forsubstantial extinction and obscuration of star-forming regions in theoptical. Without correction for reddening in the host galaxy orcorrection for [N II] contamination, the star formation rates derivedfor Hα are typically 0.5-1.0 dex lower than those estimated fromthe IR flux, and the scatter in the correlation is very large. However,an unexpected result is that when spectroscopic data are used toeliminate objects dominated by an active nucleus, to determine thegalaxian extinction, and to correct the Hα flux for both reddeningand for the contamination by the [N II] emission, a remarkably goodcorrelation emerges between the star formation rates estimated from theHα flux and those derived from the FIR continuum. In addition, astrong correlation is found between the extinction in the line-emittingregion, AHα, and the rate of star formation. Ourresults invalidate the use of Hα imaging as a reliable indicatorof star formation in starburst galaxies unless spectroscopic data arealso available. This has important implications for the determination ofstar formation rates in high-redshift galaxies. Finally, we find nocorrelation between the measured star formation rates, and theinteraction class, suggesting that the enhanced star formation ratestriggered by the interaction continue throughout the whole of themerging sequence.
| Radio-to-Far-Infrared Spectral Energy Distribution and Photometric Redshifts for Dusty Starburst Galaxies
As a logical next step in improving the radio-to-submillimeter spectralindex as a redshift indicator, we have investigated a technique of usingthe entire radio-to-far-infrared spectral energy distribution (SED) forderiving photometric redshifts for dusty starburst galaxies at highredshift. A dusty starburst SED template is developed from thetheoretical understanding of various emission mechanisms related tomassive star formation processes, and the template parameters areselected by examining the observed properties of 23 IR-selectedstarburst galaxies: Td=58 K, β=1.35, andfnth=1. The major improvement in using this template SED forderiving photometric redshifts is the significant reduction in redshiftuncertainty over the spectral index technique, particularly at higherredshifts. Intrinsic dispersion in the radio and far-infrared SEDs aswell as absolute calibration and measurement errors contribute to theoverall uncertainty of the technique. The derived photometric redshiftsfor five submillimeter galaxies with known redshifts agree well withtheir spectroscopic redshifts within the estimated uncertainty.Photometric redshifts for seven submillimeter galaxies without knownspectroscopic redshifts (HDF 850.1, CUDSS 14.1, Lockman 850.1, SMMJ00266+1708, SMM J09429+4658, SMM J14009+0252, and FIRBACK J1608+5418)are derived.
| First Results from the COLA Project: The Radio-Far-Infrared Correlation and Compact Radio Cores in Southern COLA Galaxies
We present the first results from the COLA (compact objects in low-powerAGNs) project, which aims to determine the relationship between onefacet of AGN activity, the compact radio core, and star formation in thecircumnuclear region of the host galaxy. This will be accomplished bythe comparison of the multiwavelength properties of a sample of AGNswith compact radio cores to those of a sample of AGNs without compactcores and a matched sample of galaxies without AGNs. In this paper wediscuss the selection criteria for our galaxy samples and present theinitial radio observations of the 107 southern(δ<0deg) galaxies in our sample. Low-resolution ATCAobservations at 4.8, 2.5, and 1.4 GHz and high-resolution,single-baseline snapshots at 2.3 GHz with the Australian Long BaselineArray (LBA) are presented. We find that for the majority of the galaxiesin our sample, the radio luminosity is correlated with the far-infrared(FIR) luminosity. However, a small number of galaxies exhibit a radioexcess causing them to depart from the FIR-radio correlation. Compactradio cores are detected at fluxes greater than 1.5 mJy in nine of the105 galaxies observed with the LBA. The majority (8/9) of these galaxiesexhibit a radio excess, and 50% (7/14) of the galaxies that lie abovethe radio-FIR correlation by more than 1 σ have compact radiocores. The emission from the compact cores is too weak to account forthis radio excess, implying that there are radio structures associatedwith the compact cores that extend farther than the 0.05" resolution(corresponding to a linear scale 11-22 pc) of the LBA. There is noevidence that the radio luminosity of the compact cores is correlatedwith the FIR galaxy luminosity, indicating that the core contributeslittle to the overall FIR emission of the galaxy. The galaxies withcompact cores tend to be classified optically as AGNs, with two-thirds(6/9) exhibiting Seyfert-like optical emission line ratios, and theremaining galaxies classified either as composite objects (2/9) orstarburst (1/9). The galaxies classified optically as AGNs also exhibitthe largest radio excesses, and we therefore conclude that a large radioexcess on the radio-FIR correlation is a strong indication of an AGNwith a compact radio core.
| The SCUBA Local Universe Galaxy Survey - II. 450-μm data: evidence for cold dust in bright IRAS galaxies
This is the second in a series of papers presenting results from theSCUBA Local Universe Galaxy Survey. In our first paper we provided850-μm flux densities for 104 galaxies selected from the IRAS BrightGalaxy Sample and we found that the 60-, 100-μm (IRAS) and 850-μm(SCUBA) fluxes could be adequately fitted by emission from dust at asingle temperature. In this paper we present 450-μm data for thegalaxies. With the new data, the spectral energy distributions of thegalaxies can no longer be fitted with an isothermal dust model - twotemperature components are now required. Using our 450-μm data andfluxes from the literature, we find that the 450/850-μm flux ratiofor the galaxies is remarkably constant, and this holds from objects inwhich the star formation rate is similar to our own Galaxy, toultraluminous infrared galaxies (ULIRGs) such as Arp 220. The onlypossible explanation for this is if the dust emissivity index for all ofthe galaxies is ~2 and the cold dust component has a similar temperaturein all galaxies [formmu3](Tc~20-21K). The 60-μmluminosities of the galaxies were found to depend on both the dust massand the relative amount of energy in the warm component, with a tendencyfor the temperature effects to dominate at the highest L60.The dust masses estimated using the new temperatures are higher by afactor of ~2 than those determined previously using a singletemperature. This brings the gas-to-dust ratios of the IRAS galaxiesinto agreement with those of the Milky Way and other spiral galaxieswhich have been intensively studied in the submm.
| The SCUBA Local Universe Galaxy Survey - I. First measurements of the submillimetre luminosity and dust mass functions
This is the first of a series of papers presenting results from theSCUBA Local Universe Galaxy Survey (SLUGS), the first statistical surveyof the submillimetre properties of the local Universe. As the initialpart of this survey, we have used the SCUBA camera on the James ClerkMaxwell Telescope to observe 104 galaxies from the IRAS Bright GalaxySample. We present here the 850-μm flux measurements. The 60-, 100-,and 850-μm flux densities are well fitted by single-temperature dustspectral energy distributions, with the sample mean and standarddeviation for the best-fitting temperature beingTd=35.6+/-4.9K and for the dust emissivity indexβ=1.3+/-0.2. The dust temperature was found to correlate with60-μm luminosity. The low value of β may simply mean that thesegalaxies contain a significant amount of dust that is colder than thesetemperatures. We have estimated dust masses from the 850-μm fluxesand from the fitted temperature, although if a colder component ataround 20K is present (assuming a β of 2), then the estimated dustmasses are a factor of 1.5-3 too low. We have made the first directmeasurements of the submillimetre luminosity function (LF) and of thedust mass function. Unlike the IRAS 60-μm LF, these are well fittedby Schechter functions. The slope of the 850-μm LF at lowluminosities is steeper than -2, implying that the LF must flatten atluminosities lower than we probe here. We show that extrapolating the60-μm LF to 850μm using a single temperature and β does notreproduce the measured submillimetre LF. A population of `cold' galaxies(Td<25K) emitting strongly at submillimetre wavelengthswould have been excluded from the 60-μm-selected sample. If suchgalaxies do exist, then this estimate of the 850-μm flux is biased(it is underestimated). Whether such a population does exist is unknownat present. We correlate many of the global galaxy properties with theFIR/submillimetre properties. We find that there is a tendency for lessluminous galaxies to contain hotter dust and to have a greater starformation efficiency (cf. Young). The average gas-to-dust ratio for thesample is 581+/-43 (using both the atomic and molecular hydrogen), whichis significantly higher than the Galactic value of 160. We believe thatthis discrepancy is probably due to a `cold dust' component atTd<=20K in our galaxies. There is a surprisingly tightcorrelation between dust mass and the mass of molecular hydrogen,estimated from CO measurements, with an intrinsic scatter of ~=50percent.
| Dust and gas in luminous infrared galaxies - results from SCUBA observations
We present new data taken at 850μm with SCUBA at the James ClerkMaxwell Telescope for a sample of 19 luminous infrared galaxies.Fourteen galaxies were detected. We have used these data, together withfluxes at 25, 60 and 100μm from IRAS, to model the dust emission. Wefind that the emission from most galaxies can be described by anoptically thin, single temperature dust model with an exponent of thedust extinction coefficient(kλ~λ-β) of β~=1.4-2. Alower β~=1 is required to model the dust emission from two of thegalaxies, Arp 220 and NGC 4418. We discuss various possibilities forthis difference and conclude that the most likely is a high dustopacity. In addition, we compare the molecular gas mass derived from thedust emission, M850μm, with the molecular gas mass derivedfrom the CO emission, MCO, and find that MCO is onaverage a factor 2-3 higher than M850μm.
| The Structure of Infrared-luminous Galaxies at 100 Microns
We have observed 22 galaxies at 100 μm with the Kuiper AirborneObservatory in order to determine the angular size of their FIR-emittingregions. This one-dimensional array data constitutes the highest spatialresolution ever achieved on luminous galaxies in the far-infrared. Mostof these galaxies are very luminous far-infrared sources, withLFIR>1011 Lsolar. We clearlyresolved six of these galaxies at 100 μm and have some evidence forextension in seven others. Those galaxies that we have resolved can havelittle of their 100 μm flux directly emitted by a pointlike activegalactic nucleus. Dust heated to ~40 K by recent bursts of nonnuclearstar formation provides the best explanation for their extreme FIRluminosity. In a few cases, heating of an extended region by a compactcentral source is also a plausible option.
| Multiwavelength Observations of Dusty Star Formation at Low and High Redshift
If high-redshift galaxies resemble rapidly star-forming galaxies in thelocal universe, most of the luminosity produced by their massive starswill have been absorbed by dust and reradiated as far-infrared photonsthat cannot be detected with existing facilities. This paper examineswhat can be learned about high-redshift star formation from the smallfraction of high-redshift galaxies' luminosities that is emitted ataccessible wavelengths. We first consider the most basic ingredient inthe analysis of high-redshift surveys: the estimation of star formationrates for detected galaxies. Standard techniques require an estimate ofthe bolometric luminosity produced by their massive stars. We review andquantify empirical correlations between bolometric luminosities producedby star formation and the UV, mid-IR, sub-mm, and radio luminosities ofgalaxies in the local universe. These correlations suggest thatobservations of high-redshift galaxies at any of these wavelengthsshould constrain their star formation rates to within ~0.2-0.3 dex. Weassemble the limited evidence that high-redshift galaxies obey theselocally calibrated correlations. The second part of the paper assesseswhether existing surveys have found the galaxies that host the majorityof star formation at high redshift even though they directly detect onlya small fraction of the luminosities of individual galaxies. We describethe characteristic luminosities and dust obscurations of galaxies atz~0, z~1, and z~3. After discussing the relationship between thehigh-redshift populations selected in surveys at different wavelengths,we calculate the contribution to the 850 μm background from each andargue that these known galaxy populations can together have produced theentire observed background. The available data show that a correlationbetween star formation rate and dust obscurationLbol,dust/LUV exists at low and high redshiftalike. The existence of this correlation plays a central role in themajor conclusion of this paper: most star formation at high redshiftoccurred in galaxies with moderate dust obscurations1<~Lbol,dust/LUV<~100 similar to those thathost the majority of star formation in the local universe and to thosethat are detected in UV-selected surveys.
| Nearby Optical Galaxies: Selection of the Sample and Identification of Groups
In this paper we describe the Nearby Optical Galaxy (NOG) sample, whichis a complete, distance-limited (cz<=6000 km s-1) andmagnitude-limited (B<=14) sample of ~7000 optical galaxies. Thesample covers 2/3 (8.27 sr) of the sky (|b|>20deg) andappears to have a good completeness in redshift (97%). We select thesample on the basis of homogenized corrected total blue magnitudes inorder to minimize systematic effects in galaxy sampling. We identify thegroups in this sample by means of both the hierarchical and thepercolation ``friends-of-friends'' methods. The resulting catalogs ofloose groups appear to be similar and are among the largest catalogs ofgroups currently available. Most of the NOG galaxies (~60%) are found tobe members of galaxy pairs (~580 pairs for a total of ~15% of objects)or groups with at least three members (~500 groups for a total of ~45%of objects). About 40% of galaxies are left ungrouped (field galaxies).We illustrate the main features of the NOG galaxy distribution. Comparedto previous optical and IRAS galaxy samples, the NOG provides a densersampling of the galaxy distribution in the nearby universe. Given itslarge sky coverage, the identification of groups, and its high-densitysampling, the NOG is suited to the analysis of the galaxy density fieldof the nearby universe, especially on small scales.
| The Scatter in the Relationship between Redshift and the Radio-to-Submillimeter Spectral Index
We derive the scatter in the relationship between redshift and theradio-to-submillimeter spectral index,α3501.4, using the observed spectral energydistributions of 17 low-redshift star-forming galaxies. A mean galaxymodel is derived, along with the rms scatter inα3501.4. The scatter is roughly constantwith redshift. Constant rms scatter, combined with the flattening of themean α3501.4-z relationship with increasingredshift, leads to increasing uncertainty for redshift estimates at highredshifts. Normalizing by the dust temperature in the manner proposed byBlain decreases the scatter in α3501.4 formost of the sample, but does not remove outliers, and free-freeabsorption at rest frequencies above 1.4 GHz is not likely to be adominant cause of scatter in the α3501.4-zrelationship. We rederive the cumulative redshift distribution of the 14field galaxies in a recent submillimeter and radio source sample ofSmail et al. The most likely median redshift for the distribution is2.7, with a conservative lower limit of z=2, as was also found by Smailet al. based on the original models. The normalization and shape of theredshift distribution for the faint submillimeter sources are consistentwith those expected for forming elliptical galaxies.
| The QDOT all-sky IRAS galaxy redshift survey
We describe the construction of the QDOT survey, which is publiclyavailable from an anonymous FTP account. The catalogue consists ofinfrared properties and redshifts of an all-sky sample of 2387 IRASgalaxies brighter than the IRAS PSC 60-μm completeness limit(S_60>0.6Jy), sparsely sampled at a rate of one-in-six. At |b|>10deg, after removing a small number of Galactic sources, the redshiftcompleteness is better than 98per cent (2086/2127). New redshifts for1401 IRAS sources were obtained to complete the catalogue; themeasurement and reduction of these are described, and the new redshiftstabulated here. We also tabulate all sources at |b|>10 deg with noredshift so far, and sources with conflicting alternative redshiftseither from our own work, or from published velocities. A list of 95ultraluminous galaxies (i.e. with L_60μm>10^12 L_solar) is alsoprovided. Of these, ~20per cent are AGN of some kind; the broad-lineobjects typically show strong Feii emission. Since the publication ofthe first QDOT papers, there have been several hundred velocity changes:some velocities are new, some QDOT velocities have been replaced by moreaccurate values, and some errors have been corrected. We also present anew analysis of the accuracy and linearity of IRAS 60-μm fluxes. Wefind that the flux uncertainties are well described by a combination of0.05-Jy fixed size uncertainty and 8per cent fractional uncertainty.This is not enough to cause the large Malmquist-type errors in the rateof evolution postulated by Fisher et al. We do, however, find marginalevidence for non-linearity in the PSC 60-μm flux scale, in the sensethat faint sources may have fluxes overestimated by about 5per centcompared with bright sources. We update some of the previous scientificanalyses to assess the changes. The main new results are as follows. (1)The luminosity function is very well determined overall but is uncertainby a factor of several at the very highest luminosities(L_60μm>5x10^12L_solar), as this is where the remainingunidentified objects are almost certainly concentrated. (2) Thebest-fitting rate of evolution is somewhat lower than our previousestimate; expressed as pure density evolution with density varying as(1+z)^p, we find p=5.6+/-2.3. Making a rough correction for the possible(but very uncertain) non-linearity of fluxes, we find p=4.5+/-2.3. (3)The dipole amplitude decreases a little, and the implied value of thedensity parameter, assuming that IRAS galaxies trace the mass, isΩ=0.9(+0.45, -0.25). (4) Finally, the estimate of density varianceon large scales changes negligibly, still indicating a significantdiscrepancy from the predictions of simple cold dark matter cosmogonies.
| New 8-13mum spectroscopy of luminous and ultraluminous infrared galaxies
New moderate-resolution mid-infrared spectroscopy at 10mum of 27infrared galaxies is presented. The galaxies have been chosen from three60-μm selected and one 12-μm selected complete flux-limitedcatalogues of galaxies; 17 of these sources haveL_IR(8-1000mum)>=5x10^11L_solar. A high-resolution spectrum of thesource Arp 299B1 is also presented. Combining these new results withpreviously published results, a nearly complete 60-μm selectedflux-limited subsample, with L_IR(8-1000mum)>=1.6x10^11L_solar, of 25galaxies is defined. Within this subsample, it is found that thedominant power source of infrared galaxies in the luminosity range1.6x10^11=10^12L_solar] withinthis subsample show evidence that an active galactic nucleus provides anenergetically important power source based on the detection of silicateabsorption in their mid-infrared spectra. The physical basis of apossible anticorrelation between the 11.3-μm feature equivalent widthand infrared light to molecular gas mass ratios is discussed.
| The Supernova Rate in Starburst Galaxies
We conducted an optical CCD search for supernovae in a sample of 142bright [m(B) <= 16 mag], nearby (z<=0.03) starburst galaxies overthe period 1988 December to 1991 June, to a limiting R-band magnitude of18. Five supernovae were found, in all cases outside the host galaxy'snucleus. We determine supernova rates (in supernova units or SNU) in theextranuclear regions to be 0.7 h^2 SNU for Type Ia, 0.7 h^2 SNU for TypeIb/c, and ~0.6 h^2 SNU for Type II, with large uncertainties but upperlimits of 2.2 h^2, 2.5 h^2, and 1.7 h^2 SNU, respectively. These ratesare similar to those measured in ``normal'' galaxies. We found noevidence for a supernova-induced brightening in any galactic nucleusand, with a few reasonable assumptions, can place upper limits of 9 h^2,12 h^2, and 7 h^2 SNU on the rates of unobscured supernovae Types Ia,Ib/c, and II, respectively, inside the nuclei.
| The Southern Sky Redshift Survey
We report redshifts, magnitudes, and morphological classifications for5369 galaxies with m_B <= 15.5 and for 57 galaxies fainter than thislimit, in two regions covering a total of 1.70 sr in the southerncelestial hemisphere. The galaxy catalog is drawn primarily from thelist of nonstellar objects identified in the Hubble Space TelescopeGuide Star Catalog (GSC). The galaxies have positions accurate to ~1"and magnitudes with an rms scatter of ~0.3 mag. We compute magnitudes(m_SSRS2) from the relation between instrumental GSC magnitudes and thephotometry by Lauberts & Valentijn. From a comparison with CCDphotometry, we find that our system is homogeneous across the sky andcorresponds to magnitudes measured at the isophotal level ~26 magarcsec^-2. The precision of the radial velocities is ~40 km s^-1, andthe redshift survey is more than 99% complete to the m_SSRS2 = 15.5 maglimit. This sample is in the direction opposite that of the CfA2; incombination the two surveys provide an important database for studies ofthe properties of galaxies and their large-scale distribution in thenearby universe. Based on observations obtained at Cerro TololoInter-American Observatory, National Optical Astronomy Observatories,operated by the Association of Universities for Research in Astronomy,Inc., under cooperative agreement with the National Science Foundation;Complejo Astronomico El Leoncito, operated under agreement between theConsejo Nacional de Investigaciones Científicas de laRepública Argentina and the National Universities of La Plata,Córdoba, and San Juan; the European Southern Observatory, LaSilla, Chile, partially under the bilateral ESO-ObservatórioNacional agreement; Fred Lawrence Whipple Observatory;Laboratório Nacional de Astrofísica, Brazil; and the SouthAfrican Astronomical Observatory.
| Spectroscopy of Luminous Infrared Galaxies at 2 Microns. II. Data for Galaxies with 11.2 <= (L IR/L ) 11.9
Spectra across the infrared K band are presented for a flux-limitedsample of powerful, bright, "infrared" galaxies. The sample in thepresent paper, consisting of 43 systems (47 individual galaxies) withinfrared luminosities LIR in the range 11.2 <~ log (LIR/Lȯ)<~ 11.9, was chosen from the IRAS Bright Galaxy Catalogue. Thespectra have resolving powers of ~340--680. When combined with the 13spectra we have already published for the "ultraluminous" galaxies,those with log (LIR/Lȯ) >~ 12.0, this constitutes the largestdatabase of high-quality infrared spectra yet assembled for awell-defined sample of galaxies. The spectra are, in general, dominatedby emission lines, which are due to the Br gamma hydrogen recombinationline and to several quadrupole transitions of excited molecularhydrogen. Emission from He I also appears frequently. Deep absorptionbands from CO are present in virtually all the spectra, as are a varietyof weaker stellar absorption features. The data are analyzed in acompanion paper (Paper III).
| Spectroscopy of Luminous Infrared Galaxies at 2 Microns. III. Analysis for Galaxies with log (L IR/L ) 11.2
We have obtained spectra across the K window for the first large sampleof luminous galaxies selected from the IRAS survey. This paper containsthe principal analysis of the 43 systems in our sample with luminositiesof 11.2 <~ log (LIR/Lȯ) <~ 11.9. The spectra themselves werepresented in a companion paper by Goldader et al. (Paper II). The Brgamma luminosities are proportional to LIR, at levels similar to thoseof star-forming regions. This strongly suggests that star formationaccounts for the bulk of the energy production in these objects, ingeneral agreement with previous studies. Good agreement is found for thecontinuous star formation models of Leitherer & Heckman with uppermass cutoffs well below 100 Msolar . The models accommodate arange in starburst ages of ~107 to 109 yr. Instantaneous starburstmodels fit the data but imply an unrealistically short range of ages forthe entire sample. It is difficult to avoid concluding that the initialmass functions are deficient in stars of less than ~1 Msolar. Strong emission lines from molecular hydrogen are detected. The H2 v =1--0 S(1) line luminosities are proportional to LIR; the correlationextends through the ultraluminous infrared galaxies. The H2 emission inthe galaxies tends to be more spatially extended than the Br gammaemission. Measured values and upper limits for the ratios of the variousH2 lines visible in our spectra indicate that the H2 seen in emission at2 mu m is consistent with being shock excited. However, othermechanisms, operating at sufficiently high densities that the H2 energylevels are thermalized by collisions, cannot be excluded. Based onenergy considerations, we suggest that the shocks are due to supernovaremnants expanding into the interstellar medium. The frequency of TypeII supernovae necessary to account for the H2 line emission agrees withfrequencies deduced from the starburst models and the radio/far-infraredcorrelation. However, there remain a number of galaxies that cannot bemade to fit this model. A decade after its discovery, a universalexplanation of the strong H2 emission in luminous infrared galaxiescontinues to elude us. No previously unrecognized broad-line activenuclei were discovered in our survey; either they are weak or absent orthe true optical depths at 2 mu m are much higher than indicated byconventional extinction measures. However, there are clear differencesbetween the K-band properties of galaxies that contain broad-line activenuclei and those that do not. The differences seem to be due to thepresence of strong nonstellar continuum emission coming from the activenuclei themselves. With the addition of the 13 ultraluminous galaxieswith log (LIR/Lȯ) >~ 12 from Goldader et al. (Paper I), thenumber of systems observed in this program totals 56. We haveincorporated these ultraluminous galaxies in some parts of the analysisto examine properties across the entire luminosity range of our sample.
| A comparative study of morphological classifications of APM galaxies
We investigate the consistency of visual morphological classificationsof galaxies by comparing classifications for 831 galaxies from sixindependent observers. The galaxies were classified on laser print copyimages or on computer screen using scans made with the Automated PlateMeasuring (APM) machine. Classifications are compared using the RevisedHubble numerical type index T. We find that individual observers agreewith one another with rms combined dispersions of between 1.3 and 2.3type units, typically about 1.8 units. The dispersions tend to decreaseslightly with increasing angular diameter and, in some cases, withincreasing axial ratio (b/a). The agreement between independentobservers is reasonably good but the scatter is non-negligible. In spiteof the scatter, the Revised Hubble T system can be used to train anautomated galaxy classifier, e.g. an artificial neural network, tohandle the large number of galaxy images that are being compiled in theAPM and other surveys.
| Optical Spectroscopy of Luminous Infrared Galaxies. II. Analysis of the Nuclear and Long-Slit Data
A spectroscopic survey of a sample of 200 luminous IRAS galaxies (LIGs:L_ir_^7^ > 3 x 10^10^ L_sun_; H_0_ = 75 km s^-1^ Mpc^-1^) was carriedout using the Palomar 5 meter and University of Hawaii 2.2 m telescopes.Kim et al. (1995) described the data-taking and data-reductionprocedures and presented line and continuum measurements extracted fromthe nucleus of these objects. In this paper, the nuclear data arecombined with circumnuclear measurements on 23 of these galaxies toinvestigate the properties of the line-emitting gas and underlyingstellar population in and out of the nucleus. The nuclear spectra ofthese galaxies were classified as H II region-like" or "AGN-like" usinga large number of line-ratio diagnostics corrected for the underlyingstellar absorption features. This correction is an important source oferrors in some previous studies. The emission-line spectra of many AGNswere found to-be of relatively low ionization level and were thereforeclassified as LINER. We confirm that both the fraction of LIGs with AGNspectra and the fraction of Seyferts among the AGN increase withinfrared luminosity, reaching values of 62% and 54% at the highestobserved luminosities, respectively. The fraction of LINERs, on theother hand, is relatively constant at ~27%. The source of the ionizationof the emission-line gas often is a function of the distance from thenucleus. Based on the emission-line ratios and the strengths of thestellar absorption features, circumnuclear starburst activity is acommon feature of LIGs, regardless of their nuclear spectral types. Theemission-line, absorption-line, continuum, radio, and IRAS properties ofthe LINERs suggest that most of the LINER emission in theseinfrared-selected galaxies is produced through shock ionization ratherthan photoionization by a genuine active nucleus. The nuclear region ofSeyfert LIGs is found to be slightly less reddened than that of theLINERs and H II galaxies. The dust distribution generally isconcentrated toward the nucleus, in agreement with the often peakydistribution of the molecular gas observed in these galaxies. Inverteddust profiles in which the nucleus appears less dusty than thecircumnuclear region are observed in only three LIGs, all of which haveAGN emission-line characteristics (one Seyfert 2 galaxy and two LINERs).Low nuclear dust content appears to favor the detection of activenuclei. This may be due to selection effects or may reflect realphysical differences between these classes of objects: galaxies withSeyfert emission lines may be at a more advanced stage of dustdestruction/expulsion than H II LIGs. Complex optical depth effects mayalso explain these results without invoking a smaller amount of dust inthe nucleus. The Hβ and Mg I b absorption features are stronger inthe nuclei of AGNs (especially among the LINERs) than in H II LIGs,suggesting that AGN LIGs are at a more advanced stage of stellarevolution than H II LIGs. Further support for this scenario comes fromthe fact that AGNs are found more frequently in advanced mergers than HII galaxies (only two Seyfert galaxies are detected in systems withwell-separated nuclei). However, this last result may be a luminosityeffect rather than an effect related to the dominant nuclear source ofionization. Moreover, the absorption-line data may simply reflect thefact that galaxies with powerful H II regions show evidence for youngstars while galaxies with AGNs do not. The radial variations of theHβ and Mg I b absorption features indicate the presence of a strongsource of featureless continuum in the nucleus of nearly all LIGs,regardless of their nuclear spectral types. Contamination by thecircumnuclear starburst prevents us from determining the extent of thiscontinuum source. The [O III] profiles of both Seyfert and LINER LIGswere found to be broader on average than those of H II objects. Nearly20% of the LIGs in our sample have line widths larger than 600 km s^-1^.We find that most of the galaxies in which we could determine the radialvariations of the [O III] line width present broader profiles in thecircumnuclear region than at the nucleus. When combined with publisheddata on a few other well-studied LIGs, these results suggest thatlarge-scale nuclear winds are common in these objects and are anefficient way of getting rid of the obscuring material in the nuclearregion. The spatially extended LINER emission observed in many of theseobjects is probably due to shock ionization resulting from theinteraction of the wind-accelerated gas with the ambient material of thehost galaxy.
|
Submit a new article
Related links
Submit a new link
Member of following groups:
|
Observation and Astrometry data
Catalogs and designations:
|
