a r X i
v
:
a
s t
r o -p h /0307556v 1 31 J u l 2003
Astronomy &Astrophysics manuscript no.LSB˙accept February 2,2008
(DOI:will be inserted by hand later)
Starburst in HS 0822+3542induced by the very blue LSB dwarf
SAO 0822+3545
S.A.Pustilnik 1,4,A.Y.Kniazev 1,2,4,A.G.Pramskij 1,4,A.V.Ugryumov 1,4,and J.Masegosa 3
1Special Astrophysical Observatory RAS,Nizhnij Arkhyz,Karachai-Circassia,369167Russia 2
Max Planck Institut f¨u r Astronomie,K¨o nigstuhl 17,D-69117,Heidelberg,Germany
3Instituto de Astrofisica de Andaluc ´ia,Granada,Spain
4
Isaac Newton Institute of Chile,SAO Branch
Received
February 26,2002;accepted
July 30,2003
Abstract.One of the most metal-deficient blue compact galaxies (BCGs)HS 0822+3542(Z =1/34Z ⊙),is also one of the nearest such objects (D ∼11Mpc).It is in addition well isolated from known bright galaxies.A trigger mechanism for its current star-formation (SF)burst has thus remained unclear.We report the discovery of a very
blue ((B −V )0tot =0.08and (V −R )0tot =0.14)low surface brightness (LSB)(µ0B 23.m 4arcsec
−2
)dwarf irregular (dIrr)galaxy,which we have named SAO 0822+3545.Its small relative velocity and projected distance of only ∼11kpc from the BCG imply their physical association.For this LSB dIrr galaxy,we present spectroscopic results,total B,V,R magnitudes,and the effective radii and surface brightness (SB),and we describe its morphological properties.We compare the very blue colours of this dwarf with PEGASE.2models of the colour evolution of a Z =1/20Z ⊙stellar population,and combine this analysis with the data on the LSBD EW (H α)values.The models best describing all available obser
vational data depend on the relative fraction of massive stars in the IMF used.For a Salpeter IMF with M up =120M ⊙,the best model includes a “young”single stellar population (SSP)with an age of ∼10Myr and an “old”SSP with the age of ∼0.2–10Gyr.The mass ratio of the old to young components should be in the range of 10to 30.If the age of the old component is more than ∼1Gyr,an additional coeval component of “intermediate”age (∼100Myr)with a mass comparable to that of the “young”population,although not required,provided a good fit to the current data.For the two options of a model IMF biased toward the low-mass end,the best match of the observed BV R and EW(H α)is for continuous star-formation rate (SFR)single-component models,with SF durations in the range of ∼0.1to ∼1Gyr.However,only a longer time-scale SF gives the stellar mass,compatible with the LSB galaxy mass estimates.Nevertheless,such a scenario would be inconsistent with the recent encounter of these two dwarfs.The role of interaction between the LSBD and BCG HS 0822+3542in triggering their major SF episodes during the last ∼100–200Myr is emphasized and discussed.For the BCG,based on the results of new spectroscopy with the Russian 6m telescope,we estimate the physical parameters of its SF region and present the first evidence of an ionized gas supershell.This pair of dwarfs lies deep within the nearby Lynx-Cancer void,with the nearest bright (L >L ∗)galaxies at distances >3Mpc.This is probably one of the main factors responsible for the unevolved state of HS 0822+3542.
Key words.galaxies:star formation –galaxies:low surface brightness –galaxies:interaction –galaxies:photometry –galaxies:abundances –galaxies:individual (HS 0822+3542,SAO 0822+3545)–large-scale structure
1.Introduction
A few known blue compact galaxies (BCGs)with ex-tremely low metallicities (1/50to 1/20Z ⊙)are considered to be the best candidates for truly young local low-mass galaxies,in which we are witnessing the first star forma-tion episode,with the oldest stars formed less than ∼100–200Myr ago.The best known examples are SBS 0335−052(Izotov et al.1997;Papaderos et al.1998;Pustilnik et al.2001a),and I Zw 18(Searle &Sargent 1972;Izotov
1
Note that debates on the possible youth of these BCGs
still continue in the literature (,the most recent ¨Ostlin
2000;¨Ostlin &Kunth 2001;Kunth &¨Ostlin 2001).
2S.A.Pustilnik et al.:Starburst in HS0822+3542induced by SAO0822+3545
van Zee et al.1998,Izotov et al.2001).This,presumably, is a key moment of their history.Moreover,it is probably not by chance that in both the cases of Dw1225+0152 and SBS0335−052E the nearest neighbour is very gas-rich,and either is an H i cloud without any hint of past star formation(as in the case of Dw1225+0152,Salzer et al.1991),or is also an extremely metal-deficient,proba-bly truly young,blue compact galaxy(SBS0335−052W, Pustilnik et al.1997;Lipovetsky et al.1999).
Low surface brightness galaxies(LSBGs)comprise a large fraction of the generalfield galaxy population and outnumber by a factor of several times the high sur-face brightness(HSB),McGaugh1996, Dalcanton et al.1997,O’Neil&Bothun2000).Therefore, these LSBGs can be an important factor for interaction-induced star-formation activity in gas-rich galaxies in gen-eral,mostly through distant/weak tidal encounters(see, e.g.,Taylor et al.1995;O’Neil et al.1998;Pustilnik et al. 2001b).In particular,through interaction they can trig-ger thefirst starbursts in the hypothetical population of local protogalactic H i clouds.To check this hypothesis the authors are conducting a systematic study of the lo-cal environment of the most metal-deficient BCGs.We present here new evidence in support of this idea.We re-port the discovery of a LSB dwarf irregular galaxy(named SAO0822+3545)at a projected distance of∼11.4kpc from one of the most metal-deficient BCGs,HS0822+3542 (Kniazev et al.2000).We have used the SAO RAS6m telescope spectru
m in the Hαregion,as well as BV R photometry from the Nordic Optical Telescope(NOT)to study its properties and estimate its tidal effect on the BCG HS0822+3542.We also present new high signal-to-noise(S/N)6m telescope spectra for HS0822+3542which allow us to make more accurate measurements of some of the physical parameters of its star-forming(SF)region and discover thefirst kinematic evidence of an ionized gas supershell.In Section2we describe the observations and their reduction.Results of the data analysis are presented in Section3.We discuss these results in Section4,and summarize ourfindings and draw conclusions in Section5.
2.Observations and reduction
2.1.Nordic Optical Telescope photometry
There are no cataloged galaxies around HS0822+3542in either NED or LEDA databases with V hel<1200km s−1 and projected distances less than 5.8◦.To search for fainter/non-cataloged galaxies in the close vicinity of this BCG we used B,V,R CCD images obtained with the2.5m NOT on May28,1998.These are the same frames which were used to derive B,V,R magnitudes for HS0822+3542. They were acquired with the ALFOSC spectrograph equipped with the LORAL(W11-3AC)CCD,which has a direct imaging mode.Exposure times were900s for the B-band image,and600s f
or both V and R images.For further details of these observations and data reduction we refer to the paper by Kniazev et al.(2000).2.2.Looking for possible companions
The photometric data were reduced with MIDAS2 Command Language programs according to the method described in Kniazev(1997).The MIDAS INVENTORY package was used to classify all objects.Isophotal and total BV R magnitudes were computed using the trans-formation coefficients of Kniazev et al.(2000).
Since the interaction between HS0822+3542and a possible neighbouring galaxy would probably result in some enhanced star formation in the latter,wefirst of all searched for candidate blue galaxies.The second criterion applied was that the brightness of any candidates should be comparable to that of HS0822+3542,since signifi-cantly fainter neighbours could tidally affect the BCG only from very small distances(,estimates in Pustilnik et al.2001b).
Only two blue galaxies were found in the examined field.Thefirst one is HS0822+3542itself.The second is a LSB irregular galaxy(SAO0822+3545)at3.5′to the north-east from HS0822+3542(its coordinates are given in Table5).The Digitized Sky Survey(DSS-II)blue image of thisfield with the2objects of interest and Anon J0825+3534is presented in Fig.1.All other galaxies in the NOTfield were either significantly fainter or redder.
2.3.Analysis of photometric data
We performed the reduction of photometric data for SAO0822+3545using the IRAF3package ELLIPSE.The growth curve(GC)of the galaxy was constructed by sum-ming up the pixel values from the center outwards in the circles of successive radius.
The total magnitudes(B tot,V tot,R tot)were estimated by asymptotic extrapolation of the respective radial GCs. Model-independent parameters were derived from each growth curve.The effective radii(r eff)were read on each GC at one half the asymptotic intensity,and effective SBs (SB eff)were determined as the mean brightness within a circle with the effective radius.These values are summa-rized in Table2.
2.4.Long-slit spectroscopy
Spectroscopic data were obtained with the6m tele-scope of the Special Astrophysical Observatory of Russian Academy of Sciences(SAO RAS).The long-slit spec-trograph(LSS)(Afanasiev et al.1995)was used with a Photometrics1024×1024pixel CCD detector with a 24µm pixel size.The wavelength ranges of the spec-tra obtained for different hardware configurations are
S.A.Pustilnik et al.:Starburst in HS0822+3542induced by SAO0822+35453
Fig.1.The Digitized Sky Survey(DSS-II)blue image of thefield in the vicinity of BCG HS0822+3542.North is up, East is to the left.At the adopted distance of HS0822+3542(11Mpc)the scale is∼54pc=1′′.The two marked galaxies–LSBD SAO0822+3545and Anon J0825+3534have been checked as possible companions of the BCG.On our data the LSBD has V hel∼700km s−1,close to th
at of HS0822+3542.Its projected distance to the BCG is213′′or 11.4kpc.The galaxy Anon J0825+3534,very close to the position of2MASXi J0825440+353459(NED),is a distant background object with V hel=15590km s−1.
Table1.Journal of the6m telescope spectroscopic observations
Object Date Exposure Wavelength Dispersion Seeing Airmass PA
time[s]Range[˚A][˚A/pixel][arcsec][degree]
(1)(2)(3)(4)(5)(6)(7)(8)
4S.A.Pustilnik et al.:Starburst in HS 0822+3542induced by SAO
0822+3545
Fig.2.Top panel:2D spectrum of SAO 0822+3545with the H αline at λ6579˚A .Positive Y corresponds to the NW direction for the slit position shown in Fig.3.H α-emission is clearly seen on top of the continuum in the region of ±∼5′′from the central position.Furthermore,some very faint H α-emission (with S/N ratio of ∼1)can be traced out to ±∼10′′,in the regions without detectable continuum.Bottom panel:1D spectrum of the central region (±3′′),summed along the slit,EW (H α)=17˚A .grating,giving a sampling of 1.2˚A pixel −1and an effec-tive resolution (with the 2′′slit width)of 3.5˚A .Seeing
during these observations was 1.7′′
.
For the reduction of 2D spectra we obtained biases,flat fields and illumination correction images.The primary re-duction consisted of standard steps and was done using the IRAF package CCDRED .The IRAF package LONGSLIT was used to perform wavelength calibration,background subtraction,extinction correction and flux calibration.Straightening of the 2D spectra was performed using APALL ,another IRAF package.After that,aperture ex-tractions,continuum determination,flux and equivalent width measurements of spectral lines were performed in MIDAS (for details,see Kniazev et al.2000).The reduced 2D spectrum and extracted averaged 1D spectrum of SAO 0822+3545are shown in Fig.2.
Fluxes and equivalent widths of blended lines were measured using Gaussian decomposition fitting.An av-erage sensitivity curve was produced for each night s.deviations of 5%in the whole spectral range.The sensitivity curve and line intensity errors have been prop-agated in calculating elemental abundances.To construct the P–V diagram,the methodology described in Zasov et al.(2000)was used,which allowed us to measure points with sufficiently bright H αemission s.errors on the level of 2–4km s −1.
3.Results
3.1.Spectrum of SAO 0822+3545
The 2D spectrum of SAO 0822+3545in Fig.2shows H αemission spanning ∼10′′,comparable to the appar-ent extent of the galaxy continuum.Some fainter H αfea-tures can be detected in the peripheral regions,where no continuum is seen in our spectrum.The equivalent width of the H αline,measured on the 1D spectrum av-eraged over the central region with the a spatial extent of 6′′,is EW (H α)=17±2˚A .The integrated flux in this
line is 4.65(±0.55)×10−16
erg s −1cm −2.Small variations of EW (H α)along the slit near the center of the galaxy on the level of ∼15−20˚A are within the observa-tional uncertainties.Closer to the edges the EW (H α)can be larger,but there the uncertainty reaches 70–100%,so we do not discuss these regions further.The measured velocity V hel =700±50km s −1is very close to that of HS 0822+3542.
3.2.Photometry and morphology of SAO 0822+3545
We calculated total magnitudes B tot =17.56±0.05,V tot =17.43±0.03and R tot =17.26±0.05for SAO 0822+3545us-ing the growth curve method as described in Sec.2.3.The respective integrated colours
are (B −V )tot =0.13and (V −R )tot =0.17.Relatively large errors originate from the zero-point uncertainties of the transformation equa-tions (Kniazev et al.2000).Accounting for a foreground extinction of E (B −V )=0.047in our Galaxy (Schlegel et al.1998),and applying the extinction curve from Whitford (1958),these colours become somewhat bluer:(B −V )0tot
=0.08±0.06,(V −R )0
tot =0.14±0.06.
The 2D spectrum in Fig.2and the irregular structure of the central part of SAO 0822+3545(see Fig.3)indicate that enhanced star formation took place at least in the inner part of that galaxy.The possibility of better distin-guishing individual regions of higher brightness (possible young superclusters and aged H ii regions)in the central part of the LSBD was the motivation for using the MIDAS package IMRES for deconvolution of the NOT images.This package employs an image restoration scheme devised by Richardson and Lucy and described in Adorf et al.(1992),Hook &Lucy (1992;and references therein).The num-ber of iterations was determined by comparisons between the input image and the output deconvolved images,after convolution with the point-spread function (PSF).
The results of deconvolving the B -band image after 15iterations are plotted in the right panel of Fig.3.
A number of filaments and relatively bright knots in the inner part of this LSB/dIrr galaxy are well resolved.Practically all of them are seen on the original image (left panel of Fig.3),although deconvolution makes the structure more visible.All these structures are easily visible in a deconvolved V -band image,but are less distinct in R -band.
There is a faint tail in the SE direction in the outer part of the LSBD.This tail is visible on both blue DSS-
S.A.Pustilnik et al.:Starburst in HS 0822+3542induced by SAO 0822+3545
5
Fig.3.Left panel
:The
same B -band image of SAO 0822+3545,deconvolved after 15iterations using the FFT based Lucy algorithm (Adorf et al.1992).The angular resolution has improved from 1.′′25to ∼0.′′85(FWHM).Several filaments and many knots are easily visible within the brighter part of the galaxy.Knots are marked by l
etters from “a”to “i”,respectively.II and B -band NOT images.The surface brightness of this tail corresponds to µB =25.m 7arcsec −2,or 2σof the noise level for the B -band image.The tail direction coincides with that of the LSBD major axis.The major axis P A varies from –75◦for the inner part of the galaxy up to –60◦for the outer part.The axial ratio is essentially constant over the body:b/a ∼0.55.
active下载Evidence for recent star-formation activity over the in-ternal part of this dIrr galaxy is also ,in the (B −V )0colour map (Fig.4).This map shows a rather uniform distribution (mainly in the range from –0.05to +0.10)over a large part of the main body.Its maximum (+0.2to +0.25)and minimum (–0.15to –0.10)values are measured in the positions close to the edges,where the S/N is small,and these large colour variations are appar-ently spurious.
SAO 0822+3545is a genuine LSB galaxy with traces of recent SF near its center.Since the central bright knots are seen with the lowest contrast in R -band,we made a rough estimate of the central SB of an underly-ing “disk”in R .From the data in Table 2,its effective surface brightness corrected for extinction (A R =0.12)is
µR eff,0=23.m 63arcsec
−2
.For a purely exponential disk this corresponds to a central brightness of µR 0=22.m 50arcsec −2
.Since the galaxy is significantly inclined to the line-of-sight,we need to make a corresponding correction to µR 0.For an observed axial ratio p =b/a =0.55,assuming an
intrinsic axial ratio of q =0.2,and using the well-known formula cos 2(i )=(p 2−q 2)/(1−q 2),we calculate that i =58.5◦.The inclination correction for surface brightness is then −2.5·log(cos i )=0.m 7,giving a corrected central
brightness of µR
0,c =23.m 20arcsec −2.Even if the underly-ing “disk”is as blue,as the integrated light of this galaxy (that is,(B −R )0∼0.2),its central brightness in B -band
µB 0,c =23.m 40arcsec
−2
is well within the LSB galaxy regime.For an exponential law,r eff=7.94′′in R corresponds to the scalelength αR =4.73′′.
Table 2.Photometric parameters of SAO 0822+3545
Band
Total r effSB eff
mag arcsec mag arcsec −2
(1)
(2)
(3)
6S.A.Pustilnik et al.:Starburst in HS 0822+3542induced by SAO
0822+3545
Fig.4.Grey-scale map of (B −V )-colour for SAO 0822+3545smoothed with 2D median with the win-dow size of 1.′′6×1.′′6.The extinction corrected (E (B −V )=0.047)colour distribution is shown only for regions with S/N ratio larger than 4.The colour scale is shown in the right side column.B -band surface brightness (SB)isolines are superimposed,with the lowest SB level of 25.m 7arcsec −2.The other levels are:24.5,24.0,23.6,23.3,23.1,22.9,22.8and 22.6mag arcsec −2.
tor of two (FWHM=0.′′6),and the restored BCG complex morphology is shown in Fig.5in grey scale,with
a contour indicating the outer isophote µB =25.m 0arcsec −2super-imposed.With the letters “a”to “f”we have marked all bright features well resolved after deconvolution in each of B,V ,and R bands.The brightest region consists of two components,“a”and “b”,separated by ∼1.3′′(∼70pc),roughly in the N–S direction.The contrast in Fig.5is adjusted to show faint features.The real ratio of the intensities of knots “a”and “b”,derived from this image,is ∼8.While deconvolution does not preserve brightness proportions between individual features,we consider this ratio to be indicative of the real value.
In addition to several bright knots,some filamen-tary structures are also visible.Some traces of them are also seen in other filters,but due to their lower surface brightness they are not as easily visible as the knots.Nevertheless,the prominent arc-like structure at the NW edge of the main body (stretching from X =+2,Y =+4to X =+6,Y =+4in Fig.5),is clearly visible on both B and V images,but is more noisy in R -band,presumably due to the lower S/N ratio.This feature probably repre-sents an ionized gas shell with a diameter of ∼4′′(∼200pc),caused by recent active SF in this region,Four more knots are seen in the restored image.One (“d”)is near the geometrical center of the main body.This,perhaps,
might naturally be expected because of the gravitational well in the center of the galaxy.The others knots provide further evidence that the current SF episode in the BCG is spread across the galaxy.
The P–V diagram in Fig.6indicates a supershell with a size of ∼480pc,comparable to the extent of H α-emission in the long-slit direction.The supershell velocity ampli-tude,as seen from our data,is about 30km s −1.Such supershells are well-resolved on high-resolution H αlong-slit spectra for many nearby starbursting galaxies (e.g.,Marlowe et al.1995;Martin 1996,1997,1998).The re-lated supershells of neutral gas are also seen in H i maps of such galaxies (e.g.,Walter &Brinks 1999).For the typ-ical case of a starburst off-set in position relative to the midplane of the gas disk,the asymmetry of the gas den-sity distribution in the z -direction results in the shell’s asymmetric appearance.This was ,in numer-ical simulations by Silich et al.(1996),and by Walter &Brinks (1999)for the observed types of P–V diagrams.That part of the shell propagating out of the plane should have a significantly lower Emission Measure due to de-creasing gas density,and usually appears much fainter in comparison to the part of the shell moving towards the midplane (e.g.,Martin 1996).
Shells (or supershells)are produced by hot bubbles,caused by the injection of the energy from numerous massive star winds and supernova (SN)explosions into the interstellar medium (ISM)(e.g.,Tenorio-Tagle &Bodenheimer 1988).The asymmetric appearance of super-shells in H αemission is easily seen with the high-resolution data mentioned above.When the velocity resolution is not sufficient to distinguish motions on both sides of the shell,we measure the intensity-weighted velocity i
n the H α-line at each slit position.If the H αintensities in the segments on opposite sides of the shell differ significantly,we will see mainly the side approaching the midplane of the gas disk.However,since the observed velocity at each slit position is the weighted mean of the emission from the opposite sides of the shell,a shell velocity derived in this way rep-resents a lower limit of the real value.Depending on the relative strengths of H αemission on the shell sides the expected correction could reach tens of percent.
If we accept the full amplitude of the radial veloc-ity difference between the two edges on the P–V di-agram (∼35km s −1)as the result of rotation in the BCG,then the apparent V rot is 17.5km s −1.With an apparent axial ratio p =b/a =0.5,and an assumed in-trinsic ratio q =0.2,using the same relation as for SAO 0822+3545in section 3.2we get an inclination an-gle i =63.6◦,and thus an inclination-corrected value of V rot =19.5km s −1.The latter value is quite consistent with that expected from the Tully-Fisher relation be-tween galaxy V rot and its blue luminosity,namely,using the relations derived by Karachentsev et al.(1999)for dwarf galaxies in the Local Volume.For HS 0822+3542,M B =–12.5(or L B =1.56×107L ⊙)and so the expected V rot =17km s −1,with a ±1σconfidence range of 12to 24km s −1.
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论