Are the compact star clusters in M82 evolving towards globular clusters
Supermassive Objects as Gamma-Ray Bursters
a rXiv:as tr o-ph/971120v414May1998Supermassive Objects as Gamma-Ray Bursters George M.Fuller and Xiangdong Shi Department of Physics,University of California,San Diego,La Jolla,CA 92093ReceivedABSTRACTWe propose that the gravitational collapse of supermassive objects(M∼>104M⊙),either as relativistic star clusters or as single supermassive stars(which may result from stellar mergers in dense star clusters),couldbe a cosmological source ofγ-ray bursts.These events could provide the seeds of the supermassive black holes observed at the center of many galaxies. Collapsing supermassive objects will release a fraction of their huge gravitational binding energy as thermal neutrino pairs.We show that the accompanying neutrino/antineutrino annihilation-induced heating could drive electron/positron “fireball”formation,relativistic expansion,and associatedγ-ray emission.The major advantage of this model is its energetics:supermassive object collapses are far more energetic than solar mass-scale compact object mergers;therefore,the conversion of gravitational energy tofireball kinetic energy in the supermassive object scenario need not be highly efficient,nor is it necessary to invoke directional beaming.The major weakness of this model is difficulty in avoiding a baryon loading problem for one dimensional collapse scenarios.Subject headings:gamma rays:bursts-cosmology:observations and theory1.IntroductionIn this letter we propose that the collapse of supermassive objects and the associated neutrino/antineutrino annihilation could give rise to high redshift(cosmological)γ-ray bursts(GRBs).This model could alleviate vexing problems associated with the energetics of conventional stellar remnant-based scenarios.We define a supermassive object to be a star or star cluster that suffers the general relativistic Feynman-Chandrasekhar instability during its evolution.This corresponds to objects with initial masses M∼>104M⊙,i.e., those which may leave black hole remnants with masses M∼>103M⊙.Detections of absorption and emission features at a redshift z=0.835in the spectral observation of the afterglow ofγ-ray burst GRB970508(Metzger et al.1997a,b)and at redshift z=3.42in the host galaxy of GRB971214(S.Kulkarni et al.1998)have established that at least some of the GRB sources lie at cosmological distances.Observations show that the total energy in gamma rays associated with a GRB at cosmological distancesis∼1052erg to∼1053erg when a4πsolid angle coverage is assumed(Fenimore et al. 1993;Wijers et al.1997;Kulkarni et al.1998).Catastrophic collapse events,such as neutron-star/neutron-star mergers(Paczy´n ski1986;Goodman1986;Eichler et al.1989), neutron-star/black-hole mergers(Mochkovitch et al.1993),failed supernovae(Wooseley 1993),“hypernovae”(Paczy´n ski1997),collapse of Chandrasekhar-mass white dwarfs(Usov 1992),have been touted as natural candidates for cosmological GRB sources.Fireballs created in these collapse events could accelerate material to the ultra-relativistic regime, with Lorentz factorsΓ=E e/m e c2∼>102(Paczy´n ski1986,Goodman1986,Rees&M´e sz´a ros1992,M´e sz´a ros&Rees1992).The kinetic energy in thesefireballs could then be converted toγ-rays possibly via the cyclotron radiation and/or the inverse Compton processes associated with ultrarelativistic electrons.In these models,the energy loss of the shock(s)propelled by thefireball would produce the afterglow associated with a GRB event(Waxman1997).There are,however,problems for these stellar remnant-based models if the GRBs originate from high redshift events.The total gravitational binding energy released when a ∼1M⊙configuration collapses to a black hole(or into a pre-existing larger black hole)is only∼1054erg.Calculations have shown that it is very difficult to power a GRB of energy ∼1052erg(Wijers et al.1997),or an afterglow with a similar energy(Waxman1997;Dar 1997)with such a collapse scenario,unless theγ-ray emission and the blast wave causing the afterglow are highly collimated(improbably highly collimated in the case of very high redshift events).This energetics problem can be avoided in the supermassive object collapse model suggested here.Collapse of such large mass scale objects could result in prodigious gravitational binding energy release.Some of this gravitational energy is radiated as thermal neutrino/antineutrino pairs(Fuller,Woosley,&Weaver1986,hereafter FWW; Fuller&Shi1997)whose annihilations into electron/positron pairs could create afireball above the core that generatesγ-rays.There is no direct evidence for supermassive stars ever having been extant in the universe.However,it has been argued that their formation could be an inevitable result of the collapse of∼105M⊙to106M⊙primordial clouds(the baryon Jean’s mass at early epochs,see Peebles&Dicke1968,and Tegmark et al.1997)at high redshifts in which cooling was not as efficient as in clouds contaminated with metals, or more likely,as a result of stellar mergers associated with∼>107–108M⊙relativistic star cluster collapse(Hoyle&Fowler1963;Begelman&Rees1978;Bond,Arnett,&Carr1984; FWW;McLaughlin&Fuller1996).Theflow chart for supermassive black hole production suggested by Begelman&Rees(1978)includes several pathways whereby supermassive stars are formed in the central region of the collapsing cluster.Further,supermassive black holes apparently are ubiquitous in the universe.They are invoked as the central engines ofActive Galactic Nuclei(AGNs)and quasars,and are inferred to be in the centers of nearby galaxies(van der Marel et al.1997).We note that Prilutski and Usov(1975)have previously tied GRBs to magneto-energy transfer during collapses of supermassive rotators(∼106M⊙)postulated to power AGNs and quasars.Here we propose a different energy transfer mechanism(neutrinos)based on objects not necessarily tied to AGNs or quasars,but which could possibly be related to the birth of the supermassive black holes that power them.2.Fireballs from Supermassive Object CollapseSupermassive stars will suffer the General Relativistic(Feynman-Chandrasekhar) instability,either at or before the onset of hydrogen burning(c.f.,FWW)in the case of quasi-statically contracting objects,or immediately upon formation as in the case where stellar mergers produce them.As such a star collapses,the entropy per baryon is slightly increased by nuclear burning,but then is reduced by neutrino pair emission.Though initially the whole star can collapse homologously,as the entropy is reduced only an inner “homologous core”can continue to collapse homologously(FWW).It is this homologous core that will plunge through an event horizon as a unit to make a black hole.The≡M HC/105M⊙,can be much smaller(possibly by an mass of the homologous core,M HC5order of magnitude or more)than the mass of the initial hydrostatic supermassive star,≡M init/105M⊙.M init5The collapse to a black hole of a supermassive star with a homologous core massM HC will have a characteristic(prompt)Newtonian gravitational binding energy release of erg.During the collapse,neutrino emission will ensue from e±-annihilation ∼E s≈1059M HC5in the core.The emissivity of this process scales as the core temperature to the ninth power(Dicus1972).As a result,most of the gravitational binding energy removed by neutrinos will be emitted very near the point where the core becomes a black hole,and on a timescale characterized by the free fall time(or light crossing time)of the homologous core near the black hole formation point.We employ a characteristic free fall collapse time scale oft s≈M HC5sec,and a characteristic radius(the Schwarzschild radius)of r s≈3×1010M HC5cm.For a core mass∼>104M⊙the neutrinos will not be trapped in the core and will freelystream out.For a smaller core mass,the neutrino diffusion timescale will be long compared to the free fall timescale and so neutrinos will be trapped in the core.Neutrino emission in this latter case will be from a“neutrino sphere”at the edge of the homologous core.In general it is difficult to estimate the range of initial stellar masses that will give rise to a given range of homologous core masses,though there is a clear hierarchy at each evolutionary stage.We therefore guess that the initial star cluster masses will be in the range105M⊙to109M⊙,while the subsequently produced supermassive stars will have masses M init5≈0.1to∼1000,while the corresponding homologous core masses will lie inthe range M HC5≈10−2to∼10.Figure1shows aflow chart for the collapse of supermassive objects.The neutrino luminosity can be crudely estimated from the product of the neutrino energy emissivity(Schinder et al.1987;Itoh et al.1989)near the black hole formation pointand the volume inside the Schwarzschild radius,i.e.,4×1015(T Schw9)9(4πr3s/3)erg/sec.Here T Schw9is the characteristic average core temperature near the black hole formation point in units of109K.For a spherical non-rotating supermassive star we can show thatT Schw9≈12α1/3Schw 11/2M HC5 1/6 M HC5 −1/2,(1) whereαSchw is the ratio of thefinal entropy per baryon to the value of this quantity in the initial pre-collapse hydrostatic configuration,and g s≈g b+7/8g f≈11/2is the statistical weight of relativistic particles in the core.Since for spherical non-rotating supermassivestars M init5/M HC5≈12π2R4νLνL¯ν E2ν E¯ν .(3)Here G F is the Fermi constant,L is the luminosity of the neutrinos/antineutrinos,and the brackets denote averages of neutrino energy or squared-energy over the appropriate neutrino or antineutrino energy spectra(see Shi&Fuller1998).The phase space and spin factors are K≈0.124(0.027)forν=νe(νµ,ντ),and the radial dependence of the energy deposition rate isΦ(x)=(1−x)4(x2+4x+5),with x=[1−(Rν/r)2]1/2.The characteristic neutrino luminosity Lν¯νin eq.(2)could be an underestimate of the true neutrino luminosity.A detailed numerical calculation(without considering theuncertain gravitational redshift,however)shows that the true average neutrino luminosity can be much higher if there is rapid rotation and/or magneticfields holding up the collapse (Shi&Fuller1998).The neutrino energy loss rate scales steeply as T99,and the temperature distribution in the homologously collapsing core(an index n=3polytrope)follows the Lane-Emden function and so peaks at the pensating this feature will be the R4νdependence of the aboveν¯νenergy deposition rate˙Qν¯ν.Therefore,we will approximate the entire neutrino emissivity of the core as arising from the edge of the core(Rν∼r s), and then take Lν¯νas the characteristic neutrino luminosity from eq.(2).(Note that this equation is appropriate in the case where M HC5∼<0.1and neutrinos diffuse from the core.In this case,the central temperature is irrelevant,though we may get luminosities comparable to the free streaming case because the core will have lower mass and,hence,a generally higher temperature scale.)The expected near-thermal spectrum of the neutrino emission impliesE2ν / Eν = E2¯ν / E¯ν ≈6(M HC5)−1/2MeV(Shi&Fuller1998).Therefore,the neutrino energy deposition rate per unit volume will be roughly˙Q ν¯ν(r)∼4×1022(M HC5)−7.5(r s/r)8erg cm−3s−1.(4)The total energy deposited into thefireball above a radius r isE f.b.(r)=t s ∞r4πr2˙Qν¯ν(r)d r∼2.5×1054(M HC5)−3.5(r s/r)5erg,(5) which is tremendous.Thefireball will undoubtedly lose some of this energy to thermalneutrino emission.But,once the e±pair density is high enough for this,neutrino/electronscattering should deposit even more energy.If M HC5=0.5,the energy deposited in the fireball will be∼1053erg at a radius r∼3r s∼1011cm.This is the total observed energy in a GRB assuming a4πsolid angle and a redshift z∼3.A successful model of GRBs must avoid excessive baryon loading so that a Lorentz factor ofΓ∼>102can be achieved for the baryons accelerated by thefireball.This suggeststhat the region at several Schwazschild radii from the supermassive star core should have extremely low baryon density.This may be satisfied if the whole star collapses homologously into a black hole,and/or substantial rotation causes the star to collapse in aflattened geometry with very little material in the polar directions(an extreme case of this geometry was discussed in Bardeen&Wagoner1969).The homologous collapse of the entire star could be engineered only if the star has substantial centrifugal support from rotation and/or if there is significant magnetic pressure(but not so much that an explosion results). Therefore,rotation could be a crucial factor in this picture.Rotation will also result in a longer collapse timescale,and mildly beamedγ-ray emission.A high angular momentum collapse may therefore be challenged in generating GRBs with durations∼<1second.Another means to avoid excessive baryon loading may be possible in the collapseof a dense star cluster.In this case the whole star cluster can collapse on the General Relativistic instability(Shapiro&Teukolsky1985)and collisions of M∗∼M⊙stars could provide the neutrino“engine”that powersfireballs.During the collapse,the central stars will have relativistic speeds and the typical entropy per baryon produced in zero impact parameter collisions of these will be S∼104Γ1/2(g s/5.5)1/4(M⊙/M∗)1/4(V∗/V⊙)1/4withT9∼1,conditions commensurate with those required for hydrostatic supermassive stars (S≈104(M init/108M⊙)1/2).(HereΓ∼1is an appropriate Lorentz factor,and V∗/V⊙is the ratio of the stellar collision interaction volume to the solar volume.)In fact,most collisions will not be“head-ons,”but rather will involve the tenuous outer layers of the stars.The lower densities involved will translate into larger entropies(effectively,(V∗/V⊙)1/4could be considerably larger),possibly large enough(S∼107)to provide a pairfireball directly.In the collapse,space between moving stars may provide baryon-free“lanes”,and the stellar collisions themselves may cause the neutrino emission to be“spiky”(the overall emission profile,however,should nevertheless follow the free fall collapse profile indicated above for supermassive stars).Both processes are stochastic,possibly contributing to the“spiky”time structure of the GRBs.This direct collapse of relativistic star clusters and the collapse of supermassive stars may well represent two extremes on a continuum of supermassive object collapse.3.Event Rate and Peak Flux DistributionThe rate of supermassive object collapses should be able to match the observed rate of GRB events(several per day)if a substantial fraction of the burst events are to come from this source.Assuming that supermassive objects all form and collapse at a redshift z,the rate of these collapses as observed at the present epoch is4πr2a3z d rM init,(6)where r is the Friedman-Robertson-Walker comoving coordinate distance of the objects,a z is the scale factor of the universe at the epoch corresponding to a redshift z(with a0=1), t0is the age of the universe,ρb≈2×10−29Ωb h2g cm−3≈5×10−31g cm−3(Tytler& Burles1997)is the baryon density of the universe today,h is the Hubble parameter in100 km s−1Mpc−1,and F is the fraction of baryons that were incorporated in supermassive objects.For z∼3we will have r∼3000h−1Mpc.The collapse rate is therefore0.15F(M init5)−1sec−1∼104F(M init5)−1day−1.(7)With F∼0.1%,i.e.,with0.1%of all baryons having been incorporated into supermassive objects of M init5∼10,we should observe(assuming a100%detection efficiency)one collapse per day if they emittedγ-rays into a4πsolid angle.This would constitute a substantial fraction of the observed rate of GRB events.The baryon fraction F=0.1%in∼106M⊙black holes implies a(cumulative)density of7h2such supermassive black holes formed in1 Mpc3.This GRB rate is about two orders of magnitude lower than24Gpc−3yr−1,the rate required if GRBs originate from source populations that do not evolve over time(Fenimoreand Bloom1995).This shortfall in rate results because we have assumed that all GRBsare high redshift collapse events and are therefore seen from a larger volume.In addition,the rate of supermassive object collapses required in our GRB model does not depend onthe mass scale of the collapsing objects,although the fraction F scales linearly with M init.5Observations show that almost all galaxies that have been examined appropriately seem tohave supermassive black holes in their centers(van den Marel et al.1997).It is thereforeintriguing to estimate the rate of supermassive object collapses required by our GRB modelon a per galaxy basis.If such supermassive object collapses occurred only in normal∼L∗galaxies,the rate needed is about350h−1per L∗galaxy.However,this number of eventsper galaxy is much lower,perhaps∼<10h−1per galaxy(based on,for example,the galaxynumber densities of Zucca et al.1997),if dwarf galaxies harbor supermassive objects aswell.Therefore,it may be conceivable that these supermassive object collapse events aretied to the supermassive black holes at the centers of galaxies,if such supermassive blackholes occur in every galaxy-scale object.Such an association of supermassive objects andgalaxy-scale objects may also be born out by considering Lyman limit systems and dampedLyman-αsystems,which are associated with galactic halos and disks at high redshifts.Using a column density N HI distribution per unit column density per unit absorptiondistance of1013.9N−1.74(Storrie-Lombardi,Irwin&McMahon1996),wefind that the HIrate of supermassive object collapse matches that of GRBs if every Lyman-αsystem withN HI∼>1018cm−2harbors a supermassive object.If all GRBs are from z∼>1then theγ-ray burst peakflux distribution(log N-log P)will be very different from models with a homogeneously distributed population of GRBs.The observed log N-log P distribution is a power law with index=−1.5which has a breakat the faint end(Fenimore et al.1993).This would be consistent with homogeneouslydistributed cosmological sources with a cut-offat high redshifts,unless the peakflux ofGRBs,P,cannot be regarded as a standard candle.But since the log N-log P distributionis a convolution of the peakflux and spatial distribution,there is no guarantee that the observed power law requires a homogeneous distribution of sources.For our model,in which supermassive object collapses most likely occur at cosmological distances with z∼>1,we can always invoke variances in the peakflux of GRBs,and/or an evolution of supermassive object co-moving number densities,or invoke another population of GRBs,tofit the observedγ-ray burst peakflux distribution.It is worth noting that even in existing stellar remnant-based models,the sources tend to be more abundant at z∼>1,because the star formation rate was higher then(Totani1997).4.ConclusionThe formation of the supermassive black holes inferred in AGNs,quasars and many galaxies may well involve the collapse of relativistic star clusters which form intermediate phase supermassive stars.We point out here that collapses of these supermassive objects will be accompanied by prodigious thermal neutrino emission which could transport a fraction of the gravitational binding energy of these objects to a region(s)where the baryon loading is low,thus creating“clean”fireballs that generateγ-ray bursts.The major advantage of this model is a huge energy release,and just such an energy scale is required by recent observations of high redshift bursts.We have shown that the collapse timescale and expected collapse event rates are consistent withγ-ray burst parameters.The principal weakness of our model is the baryon loading problem.We have outlined possible ways to circumvent this problem by appealing to high angular momentum andflattened collapses,and by appealing to the stochastic nature of stellar collision-induced supermassive star/black hole build-up in the collapse of relativistic star clusters.We thank David Band and Edward Fenimore for valuable suggestions.This work is supported by NASA grant NAG5-3062and NSF grant PHY95-03384at UCSD.REFERENCESBardeen,J.M.,&Wagoner,R.V.1969,ApJ,158,L65Begelman,M.C.&Rees,M.J.1978,MNRAS,185,847Bond,J.R.,Arnett,W.D.,&Carr,B.J.1984,ApJ,280,825Cardall,C.Y.,&Fuller,G.M.1997,ApJ,486,L111Cooperstein,J.,van den Horn,L.J.,&Baron,E.1987,ApJ,321,L129Dar,A.1997,unpublishedDicus,D.A.1972,Phys.Rev.D,6,941Eichler,D.,Livio,M.,Piran,T.,and Schramm,D.N.1989,Nature,340,126 Fenimore,E.E.et al.1993,Nature,366,40Fenimore,E.E.&Bloom,J.S.1995,ApJ,453,25Fuller,G.M.,Woosley,S.E.,&Weaver,T.A.1986,ApJ,307,675(FWW)Fuller,G.M.&Shi,X.1997,ApJ,487,L25Goodman,J.1986,ApJ,308,L47Goodman,J.,Dar,A.,&Nussinov,S.1987,ApJ,314,L7Hoyle,F.&Fowler,W.A.1963,MNRAS,125,169Itoh,N.,Adachi,T.,Nakagawa,M.,Kohyama,Y.,&Munakawa,H.1989,ApJ,339,354 Kulkarni,S.R.,et al.1998,Nature,393,35McLaughlin,G.C.&Fuller,G.M.1996,ApJ,456,71M´e sz´a ros,P.,&Rees,M.J.1992,MNRAS,257,29pMetzger,R.M.,et al.1997a,Nature,387,879Metzger,R.M.,Cohen,J.G.,Chaffee,M.H.,&Blandford,R.D.1997b,IAU circular No.6676Mochkovitch,R.,Hernanz,M.,Isern,J.,and Martin,X.1993,Nature,361,236Paczy´n ski,B.1986,ApJ,308,L43Paczy´n ski,B.1998,ApJ,494,L45Peebles,P.J.E.&Dicke,R.H.1968,ApJ,154,891Prilutski,O.F.,&Usov,V.V.1975,Ap&SS,34,395Rees,M.J.,&M´e sz´a ros,P.1992,MNRAS,258,41pSchinder,P.J.,et al.1987,ApJ,313,531Shapiro,S.L.&Teukolsky,S.A.1985,ApJ,292,41Shi,X.&Fuller,G.M.1998,ApJ,in pressStorrie-Lombardi,L.J.,Irwin,M.J.,&McMahon,R.G.1996,MNRAS,282,1330 Totani,T.,1997,ApJ,486,L71Tytler,D.,&Burles,S.1997,in”Origin of Matter and Evolution of Galaxies”,eds.T.Kajino,Y.Yoshii&S.Kubono(World Scientific Publ.Co.:Singapore),37 Usov,V.V.1992,Nature,357,472van der Marel,R.P.,de Zeeuw,P.T.,Rix,H.,&Quinlan,G.D.1997,Nature,385,610Waxman,E.1997,ApJ,489,L33Wijers,R.A.M.J.,Bloom,J.S.,Bagla,J.S.,&Natarajan,P.1998,MNRAS,294,L13 Woosley,S.E.,Wilson,J.R.,&Mayle,R.1986,ApJ,302,19Woosley,S.E.1993,ApJ,405,273Zucca,E.,et al.1997,A&A,326,477Figure CaptionFigure1.Aflow chart for the collapse of supermassive objects.。
英语科普：密集星团被证明是二元黑洞的工厂
The coalescence1 of two black holes -- a very violent and exotic event -- is one of the most sought-after observations of modern astronomy. But, as these mergers3 emit no light of any kind, finding such elusive4 events has been impossible so far.Colliding black holes do, however, release a phenomenal amount of energy as gravitational waves. The first observatories5 capable of directly detecting these'gravity signals' -- ripples6 in the fabric7 of spacetime first predicted by Albert Einstein 100 years ago -- will begin observing the universe later this year.When the gravitational waves rolling in from space are detected on Earth for the first time, a team of Northwestern University astrophysicists predicts astronomers8 will "hear," through these waves, five times more colliding black holes than previously9 expected. Direct observations of these mergers will open a new window into the universe."This information will allow astrophysicists to better understand the nature of black holes and Einstein's theory of gravity," said Frederic A. Rasio, a theoretical astrophysicist and senior author of the study. "Our study indicates the observatories will detect more of these energetic events than previously thought, which is exciting."Rasio is the Joseph Cummings Professor in the department of physics and astronomy in Northwestern's Weinberg College of Arts and Sciences.Rasio's team, utilizing10 observations from our own galaxy11, report in a new modeling study two significant findings about black holes:Globular clusters (spherical collections of up to a million densely13 packed stars found in galactic haloes) could be factories of binary14 black holes (two black holes in close orbit around each other); andThe sensitive new observatories potentially could detect 100 merging15 binary black holes per year forged in the cores of these dense12 star clusters. (A burst of gravitational waves is emitted whenever two black holes merge2.) This number is more than five times what previous studies predicted.The study has been accepted for publication by the journal Physical Review Letters and is scheduled to be published today (July 29).词汇解析：1 coalescencen.合并，联合参考例句：It is formed by the coalescence of the first three neuromeres in the embryo .它是由胚胎时的前三个神经原节愈合而成的。
The role of pressure anisotropy on the maximum mass of cold compact stars
(2)
the field equations are obtained as
1 − e−2µ
2µ′e−2µ
ρ=
r2
+ r,
(3)
pr
=
2γ′e−2µ − r
1 − e−2µ
r2
,
(4)
∆e2µ
=
γ ′′
+
γ′2
−
γ ′ µ′
−
γ′ r
−
µ′ r
−
1 − e2µ r2
,
(5)
where we have set p⊥ − pr = ∆. In (3)-(5), ρ is the energy density, pr is the radial pressure, p⊥ is the tangential pressure and ∆ is the measure of pressure anisotropy in this model. To solve this system we use the ansatz [22]
details)
eγ = A
cos[(β + 1)ζ β+1
+
δ]
−
cos[(β − 1)ζ β−1
+
δ]
(9)
where,
β
=
√ Λ + 2,
ζ
=
cos−1 z,
and
A
and
δ
are
constants
which
can
be
determined from the boundary conditions. The physical parameters in this
Formation of Young Star Clusters
a rXiv:as tr o-ph/45579v128May24IAU XXV JD11:Dynamics and Evolution of Dense Stellar Systems ASP Conference Series,Vol.XXX,2004XXX Formation of Young Star Clusters Bruce Elmegreen IBM Research Division,T.J.Watson Research Center,PO Box 218,Yorktown Hts.,NY,10598,USA,bge@ Abstract.Turbulence,self-gravity,and cooling convert most of the in-terstellar medium into cloudy structures that form stars.Turbulence compresses the gas into clouds directly and it moves pre-existing clouds around passively when there are multiple phases of temperature.Self-gravity also partitions the gas into clouds,forming giant regular com-plexes in spiral arms and in resonance rings and contributing to the scale-free motions generated by turbulence.Dense clusters form in the most strongly self-gravitating cores of these clouds,often triggered by compression from local stars.Pre-star formation processes inside clusters are not well observed,but the high formation rates and high densities of pre-stellar objects,and their power law mass functions suggest that turbulence,self-gravity,and energy dissipation are involved there too.In Dynamics and Evolution of Dense Stellar Systems,IAU Joint Discussion 11,Sydney Australia,July 18,2004.1.Many Scales of Star Formation Star formation has many scales.Giant star complexes extend for ∼500pc alongspiral arms and disperse in the interarm regions.The clouds that form them are usually visible in galactic-scale HI surveys,and their cores are visible in CO surveys (Grabelsky et al.1987).Many of these clouds are mildly self-bound by gravity (Elmegreen &Elmegreen 1987;Rand 1993),so they are like any other star-forming clouds:virialized,supersonically turbulent,and capable of produc-ing stars in perhaps several generations with an overall efficiency of ∼10%.The star formation process itself is confined to the densest cores of these clouds,where gravity is strong and thermal pressure is weak.Between these extremes of scale,the gas temperature decreases and the molecular content increases,but the physical processes that cause stars to form in aggregates do not appear to change much.These processes are a combination of multi-scale and repetitive compres-sions from supersonic turbulence and self-gravity,energy dissipation through shocks and magnetic diffusion,and contraction or collapse from overwhelming gravitational forces.Some of the complexity of star-formation dynamics is shown in the simulations by Bate,Bonnell,&Bromm (2003).12Bruce Elmegreen2.Scale-dependent MorphologiesCorresponding to the many scales of star formation,self-gravitating clouds havea wide range of masses,from∼107M⊙to less than1M⊙in our Galaxy.What a cloud produces is called a star cluster only if its mass exceeds∼100M⊙(Lada&Lada2003).Other than this,there is no characteristic or dominant mass forclouds or clusters,only power law distributions,so most star-forming regions are similar except for size.Size determines velocity dispersion and density for acommon background pressure,and density variations lead to important morpho-logical differences through two dimensionless ratios:the dynamical time dividedby the evolution time of stars,and the dynamical time divided by the sheartime in the local galaxy.The largest clouds take a short time,in relative terms, to form most of their stars:just1or2dynamical times like nearly every othercloud.But these largest clouds take a long time,in absolute terms,to do this,∼40My in the case of Gould’s Belt,and by then the oldest populations have lost their most massive members to stellar evolution,making the complexes lookrelatively dull(Efremov1995).The largest clouds are also the most severely af-fected by shear,making them look likeflocculent spiral arms or spiral arm spurs(Kim&Ostriker2002).These morphological differences disguise the fact thatthe physical processes of star and cluster formation are very similar on all scales.Galactic-scale stellar dynamical processes can lead to the collection of gasinto spiral density waves and resonance rings.Then the largest clouds are some-what uniformly distributed along the length of the stellar structure with a char-acteristic separation equal to∼3times the arm or ring thickness.What hap-pens here is that clouds form by asymmetric gravitational instabilities with a convergingflow along the length of the structure.Typically shear and galactic tidal forces are low in these regions,allowing the clouds to form in gas that would otherwise be stable(Rand1993;Elmegreen1994).3.Power SpectraWhen there are no galactic-scale structures,the gas appears completely scale-free,as in the Large and Small Magellanic Clouds(Stanimirovic et al.1999; Elmegreen,Kim,&Staveley-Smith2001).Power spectra of the emission or absorption from this gas have power laws with a slope similar to that for velocity power spectra in incompressible turbulence,namely∼−2.8in two-dimensions (St¨u tzki et al.1998;Dickey et al.2001).Incompressible turbulence has the Kolmogorov spectrum with a slope of−8/3.Why the column density structure in a medium that is supersonically turbulent should have about the same power spectrum as the velocity structure in incompressible turbulence is somewhat of a mystery,unless it is partly coincidence.The power spectrum of turbulent velocities varies by only a small amount,from−8/3to−3(in2D),as the motion varies from incompressible to shock-dominated.Thus even the most extreme cloud formation scenarios,where all clouds are shock fronts,would have a power spectrum similar to incompressible turbulence.In addition,some of the gas structure could result from entrainment of many tiny clouds in the larger-scale turbulent velocityfield(Goldman2000).Entrainment means density is a passive scalar,and then density power spectra are the same as velocity power spectra.Formation of Young Star Clusters3 Third,expanding shells make dense gas,and these introduce a−3component to the power spectrum because of their sharp edges.The result is a mixture of processes and innate power spectra.This is why widely diverse morphologies ranging fromflocculent dust spirals in galactic nuclei(Elmegreen,Elmegreen, &Eberwein2002)to shells and holes in the LMC or SMC(Kim et al.1999; Stanimirovic et al.1999;Elmegreen et al.2001;Lazarian,Pogosyan,&Esquivel 2002)all have about the same overall power spectrum.4.Stars Follow the GasStar formation structures,such as clusters andflocculent spiral arms,have hier-archical geometries(Feitzinger&Galinski1987;Gomez et al.1993;Elmegreen &Elmegreen2001;Zhang,Fall,&Whitmore2001)and power-law power spectra (Elmegreen,Elmegreen,&Leitner2003;Elmegreen et al.2003)that are nearly identical to those of the gas.Star formation also has a duration that scales with the region size in the same way as the turbulent crossing time scales with size (Efremov&Elmegreen1998).These similarities between star formation and turbulent gas imply that star formation follows the gas tofirst order,i.e.,that turbulence controls the star formation density,rate,and morphology.This control apparently extends to small scales too,perhaps down to indi-vidual binary stars(Larson1995),as the protostars in clusters sometimes have their own hierarchical structure(Motte,Andr´e,&Neri1998;Testi et al.2000). The large formation rates and high densities of embedded protostars also suggest that turbulence compresses the gas in which they form(Elmegreen&Shadmehri 2003).5.TriggeringCloser examination also shows a second-order effect:that a fairly high fraction of star formation is also triggered inside pre-existing clouds by external pres-sures unrelated to the clouds and to the pressures of the current generation. These processes are revealed by the wind-swept appearance of many cluster-forming clouds(e.g.,de Geus1992;Bally et al.1987)and by the proximity of cluster-forming cores to external HII regions(Yamaguchi et al.1999;Walborn et al.1999;Heydari-Malayeri et al.2001;Yamaguchi et al.2001a,b;Dehar-veng et al.2003).What is probably happening is that supersonic turbulence and entrainment in a multi-phase ISM produce the basic cloudy structure,and then unrelated pressurefluctuations in the environment trigger star formation in these structures(Elmegreen2002).Presumably there would still be star for-mation without the triggers,but with an average rate per cloud that is less because of the lower cloud densities,and a number density of active clouds that is greater because of the more dispersed nature of the dense sub-regions.The influence of pressurized triggering on the overall star formation rate in a galaxy is not known,but the universal scaling of star formation rate with average den-sity or column density(Kennicutt1998)suggests that any direct influence is weak.Star formation is probably saturated to the maximum rate allowed by the density structure in a compressibly turbulent medium(Elmegreen2002).4Bruce Elmegreen6.Size of Sample EffectsThe stochastic nature of turbulence is also reflected in the formation of star clusters,which show a random size-of-sample effect with regard to maximum mass.This appears in several ways:the most massive stars in a cluster increase with the cluster mass(Elmegreen1983),the most massive clusters in a galaxy increase with the number of clusters(Whitmore2003;Billett et al.2002;Larsen 2002),and the most massive clusters in a logarithmic age interval increase with the age(Hunter et al.2003).In all cases,the slopes of these increases are determined exclusively by the mass function through the size of sample effect: bigger regions sample further out in the tail of the distribution and have more massive most-massive members.There is apparently no physical effect or phys-ical parameter that has yet been found to determine the most massive member of a population.This is true even for individual stars(Massey&Hunter1998; Selman et al.1999)although stellar radiation pressure and winds could limit the stellar mass once it gets large enough(Yorke&Sonnhalter2002;but see McKee&Tan2003).Similarly,the ISM pressure should limit the cluster mass,considering that a cluster is recognized only if its density exceeds a certain value(depending on the sensitivity of the observation),and the density,mass and pressure are related by the virial theorem with a boundary condition.Nevertheless,this pressure limit for massive clusters has not been seen yet.It would appear as a drop-offat the upper end of the cluster mass function in a very large galaxy(sampling lots of clusters)with a low pressure(such as a giant low-surface brightness galaxy). Most galaxies have their sample-limiting mass comparable to or less than their pressure-limiting mass.Dwarf starburst galaxies are an extreme example of this because they have very few clusters overall and yet some high pressure regions. Dwarf galaxies do indeed have an erratic presence of massive clusters,some of which may be related to galaxy interactions(Billett et al.2002).7.SummaryMost stars form in clusters(Carpenter2000;Lada&Lada2003)and many of these clusters are close enough to high-pressure regions to look triggered. Triggering seems necessary because the dynamical pressures inside clusters are several orders of magnitude larger than the ambient interstellar pressure.The high pressure state of a cluster is an obvious remnant of its birth,but clues to the origin of the pressure are lost once the gas disperses and the stellar orbits mix.The primary distinction between the formation of standard“open clusters”and the mere aggregation of stars in a compressibly turbulent medium is probably this last step of triggering.HII regions did not compress gas to make Gould’s Belt,but they did compress gas to make the Trapezium cluster in Orion.The masses and positions of the clouds that are compressed into clusters seem to be the result of interstellar turbulence and shell formation.Turbulence structures the gas in two ways:by directly compressing parts of it through random large-scaleflows,and by moving pre-existing clouds around passively. This duality of processes follows from the multi-phase nature of the ISM and from the presence of bine these with pervasive pressure burstsFormation of Young Star Clusters5 from massive stars and the result is a mode of star formation dominated by dense clusters.ReferencesBally,J.,Langer,W.D.,Stark,A.A.,&Wilson,R.W.1987,ApJ,312,L45 Bate,I.A.,Bonnell,M.R.,&Bromm,V.2003,MNRAS,339,577Billett,O.H.,Hunter,D.A.,&Elmegreen,B.G.2002,AJ,1231454 Carpenter,J.M.2000,AJ,120,3139de Geus,E.J.1992,A&A,262,258Deharveng,L.,Zavagno,A.,Salas,L.,Porras,A.,Caplan,J.,&Cruz-Gonz´a lez,I.2003,A&A,399,1135Dickey,J.M.,McClure-Griffiths,N.M.,Stanimirovic,S.,Gaensler,B.M,&Green,A.J.2001,ApJ,561,264Efremov,Y.N.1995,AJ,110,2757Efremov,Y.N.,&Elmegreen,B.G.1998,MNRAS,299,588Elmegreen,B.G.1983,MNRAS,203,1011Elmegreen,B.G.1994,ApJ,433,39Elmegreen,B.G.2002,ApJ,577,206Elmegreen,B.G.&Elmegreen,D.M.1987,ApJ320,182Elmegreen,B.G.&Elmegreen,D.M.2001,AJ,121,1507Elmegreen,B.G.,Kim,S.,&Staveley-Smith,L.2001,ApJ,548,749 Elmegreen,B.G.&Shadmehri,M.2003,MNRAS,338,817Elmegreen,D.M.,Elmegreen,B.G.,&Eberwein,K.S.2002,ApJ,564,234 Elmegreen,B.G.,Elmegreen,D.M.,&Leitner,S.N.,2003,ApJ,590,271 Elmegreen,B.G.,Leitner,S.N.,Elmegreen,D.M.,&Cuillandre,J.-C.2003, ApJ,in pressFeitzinger,J.V.&Galinski,T.1987,A&A,179,249Goldman,I.2000,ApJ,541,701Gomez,M.,Hartmann,L.,Kenyon,S.J.,&Hewett,R.1993,AJ,105,1927 Grabelsky,D.A.,Cohen,R.S.,May,J.,Bronfman,L.&Thaddeus,P.1987,ApJ, 315,122Heydari-Malayeri,M.,Charmandaris,V.,Deharveng,L.,Rosa,M.R.,Schaerer,D.,&Zinnecker,H.2001,A&A,372,495Hunter,D.A,Elmegreen,B.G.,Dupuy,T.J.,&Mortonson,M.2003,ApJ,in pressKennicutt,R.C.1998,ApJ,498,541Kim,S.,Dopita,M.A.,Staveley-Smith,L.,Bessell,M.S.1999,AJ,118,2797 Kim,W.-T.&Ostriker,E.C.2002,ApJ,570,132Lada,C.J.&Lada,E.A.2003,ARAA,in press.Larsen,S.S.2002,AJ,124,1393Larson,R.B.1995,MNRAS,272,2136Bruce ElmegreenLazarian,A.,Pogosyan,D.,&Esquivel,A.2002,ASP Conf.Proc.276,Seeing Through the Dust:The Detection of HI and the Exploration of the ISM in Galaxies,ed. A.R.Taylor,ndecker,&A.G.Willis(San Francisco:ASP),182Massey,P.&Hunter,D.A.1998,ApJ,493,180McKee,C.F.&Tan,J.C.2003,ApJ,585,850Rand,R.J.1993,ApJ,410,68Selman,F.,Melnick,J.,Bosch,G.,&Terlevich,R.1999,A&A,347,532 Stanimirovic,S.,Staveley-Smith,L.,Dickey,J.M.,Sault,R.J.,&Snowden,S.L.1999,MNRAS,302,417St¨u tzki,J.,Bensch,F.,Heithausen,A.,Ossenkopf,V.,&Zielinsky,M.1998, A&A,336,697Walborn,N.R.,Barb´a,R.H.,Brandner,W.,Rubio,M.,Grebel,E.K.,Probst, R.G.1999,AJ,117,225Whitmore,B.C.2003,in STScI Symp.14,A Decade of HST Science,ed.M.Livio,K.Noll,&M.Stiavelli(Baltimore:STScI),in press Yamaguchi,R.,Saito,H.,Mizuno,N.,Mine,Y.,Mizuno,A.,Ogawa,H.,& Fukui,Y.1999,PASJ,51,791Yamaguchi,R.,Mizuno,N.,Onishi,T.,Mizuno,A.,&Fukui,Y.2001a,ApJ, 553,185Yamaguchi,R.,Mizuno,N.,Onishi,T.,Mizuno,A.,Fukui,Y.2001b,PASJ,53, 959Yorke,H.W.&Sonnhalter,C.2002,ApJ,569,846Zhang,Q.,Fall,S.M.,&Whitmore,B.C.2001,ApJ,561,727。
The Star Clusters in the Starburst Irregular Galaxy NGC 1569
a rXiv:as tr o-ph/9280v118Se p2The Star Clusters in the Starburst Irregular Galaxy NGC 15691Deidre A.Hunter Lowell Observatory,1400West Mars Hill Road,Flagstaff,Arizona 86001USA;dah@ Robert W.O’Connell University of Virginia,Department of Astronomy,PO Box 3818,Charlottesville,Virginia 22903-0818USA;rwo@ J.S.Gallagher Washburn Observatory,University of Wisconsin,475N.Charter St.,Madison,Wisconsin 53706USA;jsg@ and Tammy A.Smecker-Hane University of California,Department of Physics and Astronomy,4129Reines Hall,Irvine,California 92697-4575USA;smecker@ ABSTRACT We examine star clusters in the irregular,starburst galaxy NGC 1569from HST images taken with filters F336W,F555W,and F814W.In addition to the two super star clusters that are well known,we identify 45other clusters that are compact but resolved.Integrated UVI colors of the clusters span a large range,and comparisonwith coeval evolutionary models suggest that the ages range from 2–3Myrs to 1Gyr.Most of the clusters have colors consistent with ages of ≤30Myrs placing them at the tail end of the recent burst of star formation.We examine the radial surface brightness profiles of four of the clusters,and fit King models to three of them.The colors of the clusters are approximately constant with radius.The four clusters have half-light radii and core radii that are in the range observed in present-day globular clusters in our Galaxy.However,they are somewhat less concentrated that the average globular.The two well-known super star clusters have luminosities,and one has a known mass,that are comparable to those of typicalglobular clusters.The other two clusters,and likely numerous others in the sample,are similar to a small globular cluster and to R136in the LMC.The conditions thatproduced the recent starburst,therefore,have also been those necessary for producingcompact,bright star clusters.We examine resolved stars in the outer parts of the super star clusters.Wefind that cluster A contains many bright blue stars.Some of the blue stars are brightenough to be evolved massive stars.There is also a small population of red supergiants.Components A1and A2within cluster A have similar colors and a two-dimensionalcolor map does not offer evidence that one component is dominated by red supergiantsand the other not.The contradiction of the presence of red supergiants with Wolf-Rayetstars may instead not be a contradiction at all since there coexistence in a coevalpopulation is not inconsistent with the evolution of massive stars.Alternatively,there may be a small age spread of several Myrs within cluster A.The stars thatwe resolve around cluster B,on the other hand,contain a small population of morenormal blue massive stars and a large population of red supergiants.The presence ofthe red supergiants is consistent with the view that cluster B is in its red supergiantphase.The presence of the red supergiant stars in clusters A and B is also verified innear-infrared spectra where wefind strong stellar CO absorption features.The variousage indicators are consistent with a picture in which cluster B is of order10–20Myrsold,and cluster A is≥4–5Myrs old.The timescale to form the holes seen in HαandHI is comparable to the age of cluster B.Subject headings:galaxies:irregular—galaxies:star formation—galaxies:individual:NGC1569—galaxies:star clusters1.IntroductionWith the Hubble Space Telescope(HST)people arefinding increasing numbers of super star clusters(see,for example,Holtzman et al.1992;Whitmore et al.1993;Conti&Vacca1994; Hunter et al.1994;O’Connell et al.1994,1995;Barth et al.1995).These are compact,luminous star clusters that have sizes and luminosities(when scaled to a common age)that make them comparable to globular clusters,the most massive star clusters known.However,unlike globular clusters,the super star clusters being found today are often young,in some cases as young as a few Myrs.Therefore,these super star clusters are invaluable in providing us with a unique window on the early stages,evolution,and conditions necessary to form globular-type clusters.They also probe the star formation process at one of its extremes.Young super star clusters as massive as globular clusters are not common in normal disk galaxies,but there are several of these clusters within a few Mpc of the Milky Way.Because of their proximity,these clusters are the best examples for investigating the details of the clustersthemselves.The closest example of a young massive,compact star cluster is R136,at the center of the30Doradus nebula in the Large Magellanic Cloud(LMC).Within a cluster radius of4.7 pc,R136contains300times the concentration of luminous stars in a typical OB association (O’Connell et al.1994,Hunter1995).R136is unique among the globular-like clusters in that it can be resolved into individual stars and one can investigate the mass function resulting from such a concentrated star-forming event.Nevertheless,R136is several magnitudes fainter than other super star clusters when normalized to the same age.R136has an integrated V-band magnitude of only−11.Beyond the LMC,the next closest known super star clusters are among the most extreme in terms of luminosity.These are clusters A and B located in the nearby peculiar irregular galaxy NGC1569(Ables1968).These clusters are so compact that they appear stellar in ground-based images with good seeing,and their nature was controversial(Arp&Sandage1985).However, HST Cycle1images resolved these objects,proving them to be star clusters within NGC1569 with absolute V-band magnitudes of−14and−13(O’Connell et al.1994).These super star clusters are resident in the central region of a galaxy that itself is very unusual.NGC1569has recently undergone a true wide-scale burst of star formation involving most of the optical galaxy(de Vaucouleurs,de Vaucouleurs,&Pence1974;Hodge1974;Gallagher, Hunter,&Tutukov1984;Israel1988;Vallenari&Bomans1996;Greggio et al.1998).The galaxy still today has substantial on-going star formation as seen by the presence of bright H ii regions.In addition Della Ceca et al.(1996)have detected hard X-rays from supernova remnants and binaries associated with the center of the galaxy and soft,diffuse X-rays extending along the minor axis of the galaxy(see also Heckman et al.1995),both consequences of the intense recent star formation. Furthermore,there are largefilaments of ionized gas visible in Hαthat extend to1.9kpc from the center of the galaxy(Waller1991;Hunter,Hawley,&Gallagher1993;Hunter&Gallagher 1997)and a complex velocityfield in the ionized gas(Tomita,Ohta,&Saitstar formation episode in NGC1569.2.The Observations and Data Reduction2.1.The HST DataThe center of NGC1569was imaged with the Wide Field and Planetary Camera2(WFPC2) on HST on1998October21.The galaxy was centered on the PC for maximum resolution of star clusters A and B:0.04555′′per pixel which is0.55pc at the galaxy.The galaxy was imaged through filters F336W,F555W,F814W,and F656N.Exposures through F555W and F814W consisted of a series of three integration times—short,medium,and long.In the longer exposures the centers of clusters A and B are saturated.For the longer exposures there are multiple exposures in order to improve signal-to-noise and remove cosmic rays.The list of observations are given in Table1.Basic data reduction steps were done by the Space Telescope Science Institute“pipeline”processing system.We produced a nebular emission image by combining the medium-exposure F555W and F814W images,shifting,scaling,and subtracting from the F656N image to remove the stellar continuum.We then subtracted nebular emission from the F555W and F814W images, as needed,using a scaled and shifted Hαemission image.A mosaic of the CCD frames is shown in Figure1.A false-color combination of F555W,F814W,and F656N is shown in Figure2.We measured the brightnesses offield stars in the galaxy using the crowded star photometry package DAOPHOT(Stetson1987)as implemented in the Image Reduction and Analysis Facility (IRAF).Here we will discuss the resolved stars in and near the star clusters A and B;we will discuss analysis of thefield star population itself in another paper.We measured integrated photometry of star clusters using simulated aperture photometry.We calibrated the photometry using the zero point constants given by Holtzman et al.(1995b)in their Table9.We also converted the F336W,F555W,and F814W photometry to the Johnson/Cousins UVI system using the conversions of Holtzman et al.The instrumental magnitudes were corrected for reddening and the red leak in F336W,as discussed below,before converting to UVI,as discussed by Holtzman et al.2.2.Near-IR SpectroscopyWe obtained longslit near-infrared spectra of clusters A and B in order to examine the stellar CO features,which are an age diagnostic through sensitivity to the presence of red supergiants.The spectra were obtained over3nights in1995December with the Ohio State Infrared Imager/Spectrometer(OSIRIS)on the1.8m Perkins telescope.The spectra cover21900˚A to23900˚A at8˚A per pixel.The clusters were observed in alternating positions on the slit,10′′apart.One position was used as a sky observation for the other.We also alternated a25–70 minute series of observations of a cluster with a pair of observations of HR1440,an A1V star without the CO spectral features,chosen to be close to NGC1569in airmass.We also observed a series of bright stars of various spectral types for comparison to the clusters.We corrected the data for non-linearities using dark andflatfield observations.Theflat sequence consisted of a series of increasing exposure times intertwined with1secondflat exposures to remove drifts in the lamps.Dark observations preceeded and followed theflatfield observations. Regular domeflats were also taken and the data wereflatfielded.Sky frames were subtracted from object frames,one-dimensional spectra were extracted,and subgroups of spectra were combined. Night sky lines in spectra with the sky not subtracted were used to determine the wavelength scale and linearize along the dispersion axis.The night sky lines were identified using a line list provided by M.Hanson(private communication)which came originally from C.Kulesa(private communication to M.Hanson).The cluster spectra were divided by the nearest spectrum of HR 1440.in order to remove the telluric absorption.For the cluster spectra the continuum was also fit and the spectra divided by thefit.3.Data Analysis Issues3.1.ReddeningThe reddening to and within NGC1569has been estimated by various methods.The reddening in the Milky Way in the direction of NGC1569E(B−V)f is estimated to be0.51by Burstein&Heiles(1984)from the column density of HI.Israel(1988)measured a total reddening E(B−V)t of0.56±0.10from ANS ultraviolet data.Devost,Roy,&Drissen(1997)obtained optical emission spectra of the ionized gas in NGC1569and concluded that foreground E(B−V)f is0.52 and internal E(B−V)i is0.21for R=ing optical spectra Kobulnicky&Skillman(1997) examined the variation of reddening within the galaxy and found that c(Hβ)varies from0.8to1.2 with an average of0.97±0.07.The average translates into an E(B−V)t of0.63for R=3.1.Figure2clearly shows that there is ionized,and hence probably neutral gas and dust, distributed non-uniformly around the galaxy.From thisfigure one can see that clusters A and B themselves are sitting in a large hole in the ionized gas and Israel&van Driel(1990)state that they are sitting in holes in the HI gas as well(see also Swaters1999).Thus,the extinctions to these clusters are likely to be at the low end of the range in NGC1569.However,other clusters are embedded in H ii regions and are likely to have extinctions at the high end of the range. Nevertheless,according to Kobulnicky&Skillman(1997),the total range in E(B−V)t is only0.56 to0.71,where0.51of that is foreground Milky Way reddening.We have used Figure2to classify the clusters in3extinction bins—heavy,medium,and light, according to how much ionized gas is present in the immediate vicinity of the cluster.Clusters6,41,and44are considered to be heavily internally extincted and E(B−V)t is taken to be0.71 for these.Clusters4,7,8,9,10,39,40,and42are considered to have medium extinction and E(B−V)t is taken to be0.63.All other clusters are taken to be lightly internally extincted and an E(B−V)t of0.56is adopted.Although this manner of determining the reddening is rough,the uncertainty in any adopted E(B−V)can be no more than±0.15magnitude,the total range in E(B−V)measured from spectroscopy.We use the Cardelli,Clayton,&Mathis(1989)extinction curve and an A V/E(B−V)of3.1.(Although NGC1569is of low metallicity compared to the Milky Way,this has small effect on the optical extinction curve[Bouchet et al.1985]).Holtzman et al.(1995b)showed that the reddening correction is a function of the spectrum of the object,and at F336W the difference between the extinction of an O6star and a K5star can be large.The colors of the clusters cover a large range,comparable to that of the stellar main sequence.Therefore,we have determined a reddening correction that depends on the integrated colors of the clusters.To determine the extinction as a function of the observed colors of objects, we used the tools in the Space Telescope Science Data Analysis System(STSDAS)to simulate the throughput of the telescope plusfilters for various blackbodies reddened by E(B−V)t’s of0.56, 0.63,and0.71.The extinction corrections for O6and K5type spectra are those given by Holtzman et al.in their Table12;we have simply used STSDAS simulations to determine extinctions for effective temperatures in between these.The cluster photometry was then corrected for extinction according to its observed F555W−F814W color and its E(B−V)t category.3.2.Red Leak in F336WSince we wish to examine the F336W−F555W color of clusters,we need to consider the red leak in the F336Wfilter.Not only are many of these clusters intrinsically red,but they are also highly reddened from foreground extinction.This means that red photons could contribute significantly to the counts in the F336Wfilter.To determine the red leak as a function of the observed F555W−F814W color,we used the simulations in STSDAS and blackbody curves reddened by E(B−V)t’s of0.56,0.63,and0.71.The red leak was taken to be anyflux contribution from≥4000˚A,after the definition of Holtzman et al.(1995b).The F336W cluster photometry was corrected for the red leak based on its observed F555W−F814W color and its reddening category. For an especially red cluster with an observed F555W−F814W of1.5and an E(B−V)t of0.71, the contribution of the red leak to the F336W magnitude is0.06magnitudes,but the correction increases rapidly for redder observed colors.3.3.Distance to the GalaxyIsrael(1988)used a distance to NGC1569of2.2±0.6Mpc.From WFPC1observations O’Connell et al.(1994)measured the brightnesses of the brightest stars in NGC1569,and,assuming an E(B−V)t of0.56from Israel(1988),determined a distance of2.5±0.5Mpc.Greggio et al.(1998)examined the color magnitude diagram of the galaxy from WFPC2images and adopted an E(B−V)t of0.56and distance of2.2Mpc without apparent distress to the analysis of the stellar population.In this paper we will adopt a distance of2.5Mpc.4.Identification of ClustersClusters A and B are the most spectacular star clusters in NGC1569,but they are not the only ones.We have examined the WFPC2images for other compact,but less luminous,clusters. We have included as a cluster any compact object that was resolved compared to an isolated star profile.Contamination of the sample by background galaxies is possible although WF2and WF3, which include little of NGC1569,contain few background galaxies and no objects outside of NGC 1569thatfit our criteria.Generally,we have not included looser,and spatially bigger,OB associations in our list of clusters.The reason for this is that the galaxy is really one big OB association and separating one from another is not feasible.Also,we are interested in the less common compact clusters for comparison with clusters A and B.However,we did include one OB association,cluster number 48on WF4.This OB association is discussed by Drissen&Roy(1994)who discovered a ring nebula surrounding the cluster and broad stellar emission lines that they attribute to a WN star. It is located in the outer part of NGC1569and was easy to isolate.It serves as an example of a certified OB association for comparison to the other clusters.Because the clusters are so numerous,we could not reasonably continue the identification scheme begun by others with clusters A and B.Therefore,we switched to using numbers as identifiers,but began with number3,so clusters A and B are also numbers1and2.The clusters are identified in Figure3,and are listed in Table2.Ten of the newly identified clusters are located in the vicinity of super star cluster A,over a region of about50pc projected on the sky.Together they may form something akin to a small version of the greater30Doradus region in the LMC.30Doradus is a kpc-sized area in which star formation has taken place in small units here and there over a timescale of order10Myrs (Walborn&Blades1997;see also Constellation III,Dolphin&Hunter1998).Otherwise the clusters are distributed throughout the galaxy with no apparent pattern.5.Integrated Cluster PhotometryWe give integrated photometry of the star clusters in Table2.The contributions from the sky and background galaxy were measured in an annulus just beyond the integration aperture for each object.The aperture radius is given in Table2.We show the integrated colors and magnitudes in color-color and color-magnitude diagrams in Figures4and5.We also include cluster evolutionary tracks from Leitherer et al.(1999).We use their tables for an instantaneous burst of star formation and a Salpeter(1955)stellar initial mass function from1to100M⊙.Time steps of1to9Myrs in steps of1Myrs are marked with an “x”along this evolutionary track;time steps of10,20,and30Myrs are marked with open circles. The M V have been adjusted from the models of106M⊙to the mass of3.3×105M⊙determined for cluster A by Ho&Filippenko(1996).For less massive clusters the tracks in Figure5would slide vertically to fainter M V.Kobulnicky&Skillman(1997)give the oxygen abundance of emission nebulae in NGC1569 as8.19±0.02with no evidence for chemical inhomogeneities.This oxygen abundance impliesa metallicity Z of approximately0.004.One would expect that the metallicity of current H ii regions in NGC1569should be a good estimate of the metallicities of the recently formed stars. Therefore,we began with the Z=0.004models of Leitherer et al.(1999).However,we found that the Z=0.004cluster evolutionary tracks did not account very well for the colors of some of the clusters,but that Z=0.008models did.Therefore,in Figures4and5we include cluster models for both Z=0.004and Z=0.008.In Figures4and5we see that cluster A has UVI colors that are consistent with an age of order4–5Myrs,according to cluster evolutionary models for both metallicities.This is consistent with the detection of the signature of Wolf-Rayet stars in the cluster(Delgado et al.1997). However,O’Connell et al.(1994)and De Marchi et al.(1997)have shown that cluster A has two peaks in the light distribution,suggesting two sub-clusters.De Marchi et al.also suggest that these two sub-clusters may have different ages,with one being dominated by Wolf-Rayet stars and the other dominated by red supergiants.However,the integrated UVI colors of this cluster do not reflect that and appear to be dominated solely by the younger component.Many of the other clusters in Figure4likewise fall near the locus of the Z=0.004models. The group lying above cluster A in the diagram and slightly blueward of the locus can be explained if we have made a small overcorrection to their extinction.However,cluster B and8fainter objects are too red in(V−I)to be consistent with theZ=0.004locus.We believe this is a symptom of an important contribution from red supergiants to the U,V,I light.The colors of the Leitherer et al.(1999)models exhibit strong sensitivity to abundance in the range Z=0.004to Z=0.008for ages8–16Myr,as shown in Figure4,because of the influence of red supergiants at these ages.Unfortunately,as emphasized by Origlia et al. (1999),there is considerable uncertainty about the evolutionary tracks for such phases at subsolar metallicities.What we can say is that cluster B and the other objects in its vicinity in Figure4 are roughly consistent with the Z=0.008models of Leitherer et al.(1999)at an age betweeen8 and15Myr.The agreement for cluster B would be particularly good if we have overcorrected for extinction by about0.2mags in E(B−V).However,the E(B−V)that we assumed for cluster B is only0.05magnitude above the estimated foreground extinction,so it is unlikely that we haveoverestimated the extinction by very much.An age of8–15Myrs,however,is consistent with the CO-band evidence for red supergiants discussed in Section6.3.The youngest cluster,occupying the upper left in Figure4,is number48,the OB association in the southeast part of the galaxy.Its colors indicate an age of just2–3Myrs.This ageis consistent with the upper limit of5Myrs placed by Drissen&Roy(1994)based on their observation of an expanding bubble due to a WN-type evolved massive star.The rest of the clusters,with the exception of the group around cluster B,span the range of colors along the cluster evolutionary sequence.Part of the scatter seen in Figure4is caused by stochastic effects.The latter are discussed and simulated by Girardi&Bica(1993;see also,Santos&Frogel1997;Brocato et al.1999). Stochastic effects are due to small number statistics in populating the upper masses of the stellar initial mass function;as the few massive stars in small clusters evolve,the status of a single star can have a profound affect on the integrated colors of a small cluster.Girardi and Bica’s simulated color–color diagram shows a scatter in UBV colors of several tenths magnitude.This problem affects clusters as old as1Gyr as well as young ones(see,for example,Gallagher&Smith1999). Brocato et al.further point out the difficulty in correcting for the red leak in HSTfilters under these circumstances.In addition,differences of several tenths magnitude in colors can be seen between different models depending on the treatment of various evolutionary parameters(see,for example,Brocato et al.1999).Thus,there is considerable uncertainty in assigning an age to a specific cluster from global colors.Keeping the observational,stochastic,and modeling uncertainties in mind,we conclude from Figure4that the clusters roughly span the full range in ages covered by the evolutionary models from the OB association,which is the youngest at a few Myrs,to nearly1Gyr.However,the distribution is very unlike what would be expected for a uniform distribution of ages.Instead, there is a very strong concentration to ages less than30Myr,with a subconcentration at4–6 Myr.Therefore,most of the clusters have been formed in the recent starburst activity of the galaxy.The presence of a few potentially older clusters suggests that the galaxy also formed stars in compact clusters even before the advent of the present starburst.Clusters A and B are by far the most luminous of the clusters.The rest of the clusters in our sample have M V between−7and−12.As shown in Figure5,the model clusters evolve at about constant M V for thefirst7Myrs,and M V steadily declines after that,losing several magnitudes by the time the cluster reaches the colors of the reddest clusters in our sample.Therefore,the bluer clusters with M V near−11and the redder clusters with M V near−8are likely to have had absolute magnitudes of order−11when they were a few Myrs old.This implies that these clusters are comparable in stellar mass to the compact cluster R136in the LMC.R136has an M V of−11 at an age of2Myrs(Hunter et al.1995,Massey&Hunter1998).Since R136is comparable to a small globular cluster in mass and compactness,it appears that NGC1569,besides making two very luminous super star clusters,has also made another half dozen or so clusters comparable tosmall globulars as well.In addition to the more luminous star clusters,however,there are many less luminous clusters. Those with M V near−7likely contain≤15massive stars,and would be considered more like OB associations in terms of stellar content if they were not so compact in size.The OB association, number48,included in the sample of clusters has an M V of−8.8and was encircled by an aperture with radius14pc.Except for clusters A and B,the other clusters were observed with apertures smaller than this:0.2–11pc,with an average of3pc,in radius.Some of these may be comparable to the LMC’s populous clusters.For example,the populous cluster NGC1818in the LMC has an M F555W,0of−9.3now and probably−10.5at an age of4Myrs with a half-light radius of3.2pc (Hunter et al.1997).Thus,it appears that NGC1569in its recent burst of star formation has formed many compact and relatively massive star clusters.These clusters include the extreme super star clusters A and B,as well several clusters comparable to small globular clusters in luminosity and many other comparable to normal OB associations but more compact than is typical.We do not understand what conditions are necessary to produce compact clusters or super star clusters(see, for example,discussion by Dolphin&Hunter1998).However,those conditions appear to have been met recently in NGC1569.6.The Super Star ClustersFrom HST Cycle1data,the super star clusters A and B were found to have half-light radii R0.5of2.2and3.0pc,compared to R0.5of1–8pc for today’s Milky Way globular clusters(van den Bergh et al.1991).The integrated V-band magnitudes of the clusters in NGC1569are−14 and−13,corrected for extinction.As shown in Figure5,both would have had M V∼−14at an age of4Myrs.By contrast,a typical globular cluster,if it had a Salpeter(1955)stellar initial mass function,would be expected to have had a magnitude of−13.7±1.3(Harris1991).Thus,the clusters in NGC1569have luminosities and sizes that are comparable to globular clusters that are seen in our Galaxy today,and they are excellent candidates for being true young globulars.The extreme nature of NGC1569’s super star clusters have long intrigued astronomers,and some properties,particularly their ages,have remained controversial.Ho&Filippenko(1996)used velocity dispersion data to estimate a mass of3.3±0.5×105M⊙for cluster A.Sternberg(1998) combined photometric data with a velocity dispersion and concluded that in cluster A there are stars present down to the hydrogen burning limit of0.1M⊙.Arp&Sandage(1985)obtained integrated spectra of the clusters and found that they have spectra like those of A supergiants. Prada,Greve,&McKeith(1994)obtained spectra of the Ca II infrared lines of the clusters and concluded,as had Arp and Sandage,that cluster A is dominated by A and B stars and is13–20 Myrs old.Cluster B,they concluded,is in a red supergiant phase and is12Myrs old.O’Connell et al.(1994)estimated the ages of both clusters to be roughly15Myrs from integrated V−Icolors obtained from HST images.In optical ground-based spectra Delgado et al.(1997)detected the signature of Wolf-Rayet stars in cluster A as well as the signature of young,massive stars in both clusters.They concluded that each cluster had undergone two episodes of star formation separated by about6Myrs:3and9Myrs in cluster A and2and8Myrs in cluster B.However, De Marchi et al.(1997)from HST images concluded that cluster A is actually two clusters,one that is dominated by red supergiants and one that is younger and contains the Wolf-Rayet stars.Here we examine some additional data on the structures and ages of these super star clusters. In addition we will include in some of the analysis below two other clusters from Table2for comparison.The comparison clusters were chosen to be well-resolved and among the brightest few clusters after clusters A and B:Cluster30has an M V of−11and cluster35has an M V of−10.6.1.StructureIn Figure6we show contour plots of clusters A,B,30,and35.We can see from Figure6 that cluster A has two peaks in its two-dimensional profile,asfirst pointed out by O’Connellet al.(1994)and later analyzed De Marchi et al.(1997).The peaks,designated A1and A2by De Marchi et al.,are separated by0.18′′,which is2.2pc at the galaxy,and the fainter peak is southeast of the brighter peak.Clusters B,30,and35,on the other hand,are relatively round.In the outer parts,cluster B is not entirely symmetrical,being elongated in one dimension,but the deviation is small.Because of interest in structure within clusters A and B,we have produced two-dimensional ratio maps of these clusters using the F555W and F814W images.We subtracted local background galaxy from each cluster in each image,and aligned all of the images to the medium F555W exposure.We combined the short,medium,and long exposures,replacing the saturated pixels with values from the unsaturated exposures.The resulting ratio images are shown in Figure7. One can see that both clusters are relatively blue in the centers.In cluster A the two luminosity peaks A1and A2are very similar in color and both are blue.The reddest regions are locatedin the outer parts of cluster A and are not obviously associated with any one peak.This is not what one would expect if one component is dominated by red supergiants and the other is not as suggested by De Marchi et al.(1997).However,the large overlap of the two components makes it hard to disentangle the outer parts of the two sub-clusters.Cluster B,on the other hand,is more uniform in appearance,and the center is not as blue although there is a bluer clump of stars to the southeast of the center.We have also experimented with deconvolving components A1and A2in cluster A,as had De Marchi et al.(1997).We used point-spread-function deconvolution(DAOPHOT,Stetson 1987)with various combinations of clusters B,30,and35with Gauss,Lorentz,Penny,andMoffatfitting functions to determine the point-spread-function.We alsofit two two-dimensional Gaussians to the peaks.Residuals were fairly high in all cases arguing that clusters B,30,35,and。
Rotational velocities of A-type stars. III. Velocity distributions
2
F. Royer et al.: Rotational velocities of A-type stars. III.
tational rates of coeval stars with the same mass in a single rotational distribution can have a multimodal aspect. Guthrie (1982) found that late B-type stars in clusters have bimodal rotational velocity distributions, while they are unimodal for the same class of field stars. Bimodality was also observed for the rotation of young solar mass stars in Orion (Attridge & Herbst 1992; Choi & Herbst 1996; Herbst et al. 2001; Barnes 2003). For masses intermediate to the above cases, Abt & Morrell (1995) found bimodal-like distributions among A0–F0 objects in the MS, where the component owing to low rotators was ascribed to the chemically peculiar Am and Ap stars. However, a high fraction of objects with the Am phenomenon were found to be binaries (Debernardi 2000). Regarding Ap stars, their peculiarity requires a given time after the ZAMS to appear and thus, it can be considered somehow evolution-dependent (Hubrig et al. 2000; Ste ¸ pie´ n 2000). All these quoted facts can then be summarized as follows: 1. Although it is known that the chemical separation in Am and Ap stars are not produced by the slow rotation, slow rotation favors their appearance, so that both are correlated. In Am stars slow rotation can be due to their binary character through tidal braking, while in Ap stars could be caused in part by magnetic braking; 2. Since in the whole mass interval 1.3 < ∼ M /M⊙ < ∼ 3 bimodality is clearly apparent only in cluster late B-type stars, it may correspond to some star formation characteristics. It is tempting then to review anew the velocity distributions in the main sequence phase (“dwarf” evolutionary state) of stars in the mentioned mass range, in order to detect possible signatures on differences in stellar formation characteristics. It is not excluded, however, that some particular signatures can be present in the rotational velocity distributions of those A-type stars that make the transition between objects with and without convectively unstable envelopes. These last can likely be due to stellar structure properties rather than to formation circumstances. Thus, the aim of this paper is to see: 1) whether single late B- and A-type field stars in the early main sequence evolutionary phases have unimodal or multimodal rotational distributions; 2) whether the rotational distributions of intermediate and late A-type stars bear signatures related with the complexity of their stellar envelope/atmospheric structure. The data used in this paper and the selection of samples are described in Sect. 2. The v sin i distributions are presented in Sect. 3. Section 4 gives details on the statistical processing of stellar samples to get the equatorial velocity distributions from the observed v sin i values and describes the resulting distributions. Finally, the results are discussed in Sect. 5 and summarized in Sect. 6.
高一年级英语天文知识单选题40题
高一年级英语天文知识单选题40题1. Which planet is known as the "Red Planet" because of its reddish appearance?A. EarthB. MarsC. JupiterD. Venus答案：B。
解析：在太阳系中，火星（Mars）因为其表面呈现出红色的外观而被称为“Red Planet（ 红色星球）”。
地球（Earth）是我们居住的蓝色星球；木星（Jupiter）是一个巨大的气态行星，外观不是红色；金星 Venus）表面被浓厚的大气层覆盖，不是以红色外观著称。
2. Which planet has the most moons in the solar system?A. EarthB. MarsC. JupiterD. Mercury答案：C。
解析：木星（Jupiter）是太阳系中拥有最多卫星（moons）的行星。
地球（Earth）只有一颗卫星；火星（Mars）有两颗卫星；水星 Mercury）没有卫星。
3. The planet with the shortest orbit around the Sun is _.A. MercuryB. VenusC. EarthD. Mars答案：A。
解析：水星（Mercury）是距离太阳最近的行星，它的公转轨道是最短的。
金星 Venus）、地球 Earth）、火星 Mars）距离太阳比水星远，它们的公转轨道都比水星长。
4. Which planet has a thick atmosphere mainly composed of carbon dioxide?A. EarthB. MarsC. VenusD. Jupiter答案：C。
解析：金星（Venus）有一层非常厚的大气层，其主要成分是二氧化碳 carbon dioxide）。
地球 Earth）的大气层主要由氮气和氧气等组成；火星（Mars）大气层很稀薄，主要成分虽然有二氧化碳但比例和金星不同；木星（Jupiter）的大气层主要由氢和氦等组成。
第七章星系系统的介绍
17
棒旋星系 (Barred spiral galaxies)
Type SBa
Type SBb
Type SBc
18
银河系主要成分 (1) 银盘 (disk) （旋臂spiral arm）、 (2) 核球 (bulge) 、 (3) 银晕 (halo) 、(4) 银 冕 (corona)
19
星际物质（星际介质，Interstellar Medium) 星系内分布在恒星与恒星之间（～ 6-10 ly）的物质。 包括星际气体、星际尘埃、宇宙线与星际磁场。 星际物质的质量约为银河系恒星质量的10%。 主要分布在距离银道面约1000 ly的范围内。
33
所有这些激烈的物理过程主要是集中在星系的 核心，或者是由核心引发的。 通常也称这类星系为活动星系核（Active Galactic Nuclei,简称AGN） 只要不是专门讨论活动星系本身的结构，两者 之间不再严格加以区分
34
1918年（美）柯 蒂斯（Curtis）发 现星“云”M87的 光学喷流
28
星系质量测量结果
正常旋涡星系质量~ 109 -1012 M⊙ 椭圆星系质量~ 105 -1013 M⊙ 不规则星系质量~ 106 -1010 M⊙ 星系团质量~ 1013 -1014 M⊙ 星系和星系团的引力质量大约是可见质量的10倍。
29
星系的大小变化很大： 不规则星系，只有银河系的1％－25％； 巨椭圆星系，银河系大小的5倍； 矮椭圆星系，银河系大小的1％。
10
哈勃的裁决
1924年，哈勃 (Edwin Hubble) 分解出 “仙女座大星云” (M31) 中的造父变星， 证实它确实是恒星系统。 由造父变星周光关系哈勃估计M31的距 离285 kpc（实际距离778 kpc） > 最远 的球状星团的距离 (100 kpc) 。
Clustered-Massive-Star-Formation-in-Molecular-Clou
CO clouds
Galactic clumps: Mueller et al. (2002)
IRDCs:
Carey et al. (2000);
Kirkland & Tan, in prep.
AV=200 NH=4.2x1023cm-2 =4800 M pc-2
AV=7.5 NH=1.6x1022cm-2 =180 M pc-2
nH~500cm-3
v~10km/s
Galaxy-galaxy collisions
nH~1cm-3
v~200km/s
Converging gas flows more efficient than shocks
from stellar feedback (Elmegreen 2003)
7
Surface density vs. cluster mass
Clustered Massive Star Formation in Molecular Clouds
Jonathan C. Tan
ETH Zurich & University of Florida
1
Some open questions
What are the initial conditions for proto star clusters and how do they arise?
Pressure due to self-gravity:
R M
surface density
pressure
McKee & Tan (2002; 2003)
Local examples in the Galaxy: ~ 0.1 - 1.0 g cm-2
小学上册第十四次英语第六单元期末试卷(含答案)
小学上册英语第六单元期末试卷(含答案)英语试题一、综合题(本题有50小题，每小题1分，共100分.每小题不选、错误，均不给分)1 There are many __________ in the garden.2 What is the capital of Portugal?A. LisbonB. MadridC. BarcelonaD. Porto答案:A. Lisbon3 Which sport involves a racket and a shuttlecock?A. TennisB. BadmintonC. SquashD. Table Tennis答案: B4 The sun is ___ in the sky. (shining)5 The butterfly flaps its ______.6 An electric motor converts electrical energy into _______ energy.7 of __________ (眼泪) refers to the forced relocation of Native Americans. The Trea8 His favorite food is ________.9 What is the term for a young sloth?a. Kitb. Pupc. Calfd. Hatchling答案:c10 The first human to orbit the Earth was ______ (尤里·加加林).11 What do you use to cut paper?A. GlueB. ScissorsC. TapeD. Ruler12 What do we call a group of stars that form a recognizable pattern?A. GalaxyB. ConstellationC. NebulaD. Star Cluster答案:B13 Which insect has colorful wings?A. BeetleB. AntC. ButterflyD. Grasshopper14 The ____ has a long neck and can reach high branches.15 A ____ burrows into the ground and enjoys digging.16 Which animal is known for its long tusks?A. ElephantB. HippoC. RhinoD. Giraffe答案:A17 I enjoy playing with my toy ________ (玩具名称) during holidays.18 My ___ (小兔子) has soft, white fur.19 She wants to be a _____ (doctor/teacher) when she grows up.20 My mom loves __________ (参与当地事务).21 The main purpose of proteins is to build and repair _____.22 What do you call a group of birds?A. FlockB. PackC. SchoolD. Colony答案:A23 The __________ is a famous landmark in the United States.24 A _______ (海马) is often seen near coral reefs.25 What is the name of the famous American author known for his horror stories?A. Edgar Allan PoeB. Mark TwainC. Ernest HemingwayD. F. Scott Fitzgerald答案: A26 In winter, I love to drink __________. (热可可)27 What do you call a person who studies space?A. BiologistB. AstronomerC. GeologistD. Physicist28 The ice cream is ___ (melting/freezing).29 The _____ (种植者) takes care of the garden.30 __________ are found on the right side of the periodic table.31 My best friend is very _____ (聪明) in school.32 I have a ______ robot that talks.33 A _______ can be formed through ionic or covalent bonding.34 My _____ (花园) is full of colors.35 My favorite dessert is ______ (布丁).36 The ancient Egyptians made ______ (木乃伊) as part of their burial practices.37 ts can ______ (适应变化) in their environment. Some pla38 What is the name of the ocean on the west coast of the USA?A. Atlantic OceanB. Indian OceanC. Pacific OceanD. Arctic Ocean答案:C39 My mom enjoys _______ (动词) in the kitchen. 她的手艺很 _______ (形容词).40 The process of distillation is used to separate liquids based on their ______.41 I enjoy creating games with my toy ________ (玩具名称).42 Which animal is known as the "King of the Jungle"?A. TigerB. LionC. ElephantD. Bear答案: B43 The _____ (沙子) is soft.44 I like to listen to ______ (music).45 What do we call a vehicle that flies?A. CarB. BoatC. AirplaneD. Train46 Which animal is known as "man's best friend"?A. CatB. BirdC. DogD. Fish答案：C47 The axolotl can regrow its ______.48 The soup is very _______ (hot).49 The _______ (猴子) is very intelligent and playful.50 I like to ___ (write) in my journal.51 The ______ has a strong sense of smell.52 I like to go ________ (骑马) during summer.53 My grandma grows ________ in her kitchen.54 What do we call a place where we can see art?A. MuseumB. GalleryC. StudioD. Workshop55 What do you call the process of a caterpillar turning into a butterfly?A. MetamorphosisB. TransformationC. EvolutionD. Development答案:A56 What is the name of the phenomenon where light waves are stretched as an object moves away from us?A. Doppler EffectB. RedshiftC. BlueshiftD. Gravitational Lensing57 A reaction that produces bubbles indicates a ______ reaction.58 What is the primary function of leaves?A. Absorb sunlightB. Store waterC. Attract insectsD. Provide shelter59 The chemical formula for sodium fluoride is _____.60 The weather is ___ (perfect) for a picnic.61 The _____ (蜜蜂) is essential for many plants to grow.62 The __________ (全球合作) addresses common challenges.63 Which month comes after April?A. MarchB. MayC. JuneD. July答案：B64 The main function of carbohydrates is to provide _____.65 A _______ is a chemical that can change the color of indicators.66 Dolphins are known for their ______ behavior.67 What is the value of 5 × 5?A. 20B. 25C. 30D. 35答案:B68 How many states are in the USA?A. 50B. 51C. 52D. 5369 A koala eats ______ (桉树) leaves.70 I saw a _____ (鹦鹉) talking at the zoo.71 The chemical symbol for nickel is ______.72 I like to listen to ________ (音乐).73 The chemical formula for lithium carbonate is ______.74 The __________ is a large body of water between Europe and Asia. (黑海)75 What do we call a young horse?A. FoalB. ColtC. FillyD. Both A and B答案: D76 My ___ (小狗) runs in circles.77 The process of making wine involves fermentation of _______.78 A kitten's purring is very ______ (放松).79 What do you call a baby swan?A. CygnetB. GoslingC. DucklingD. Chick答案:A80 The __________ (历史的声音) echoes through time.81 A _______ can be a wonderful addition to any garden.82 I want to _______ (学习) how to play chess.83 I enjoy ______ (玩耍) with my cousins.84 What do we call the act of protecting the environment?A. ConservationB. PollutionC. DeforestationD. Waste答案：A85 restoration ecology) aims to rehabilitate ecosystems. The ____86 My hamster likes to _______ (跑轮) in its cage.87 The Milky Way is a ______ galaxy.88 The kids are _____ in the park. (running)89 Which is the tallest animal?A. ElephantB. GiraffeC. LionD. Bear答案:B90 I like to collect ______ (邮票) from different countries. Each one tells a ______ (故事).91 The city of Funafuti is the capital of _______.92 My teacher encourages us to be polite by calling her ______. (我的老师鼓励我们通过称呼她为____来表现礼貌。
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
arXiv:071.2145v1[astro-ph]11Oct271To appear in proceedings of the Puerto Vallarta Conference on “New Quests in Stellar Astrophysics II:Ultraviolet Properties of Evolved Stellar Populations”eds.M.Chavez,E.Bertone,D.Rosa-Gonzalez &L.H.Rodriguez-Merino,Springer,ASSP series.2Mayya et al.different timescales(see Fall&Zhang,2001;Ma´ız-Apell´a niz,2004;de Grijs &Parmentier,2007,for more details).On short timescales(t∼107yr),the exploding supernovae and the resulting superwinds are responsible for cluster expansion and disruption,a process popularly dubbed as infant mortality.On intermediate timescales(107<t<few×108yr),the mass-loss from evolv-ing stars leads to the disruption of the clusters.On even longer timescales (t>few×108yr),stellar dynamical processes,especially evaporation due to two-body scattering,and tidal effects on a cluster as it orbits around the galaxy,known as gravitational shocks,come into play in the removal of stellar mass from clusters.The GCs represent those objects that have survived all these processes,whereas young SSCs are just experiencing them.Intermedi-ate age SSCs are the ideal objects to investigate the influence of disruption processes on the survival of star clusters.Almost all the star formation in the disk of M82took place in a violent disk-wide burst around100–500Myr ago, following the interaction of M82with the members of M81group(Mayya et al.,2006).Cluster formation is known to be efficient during the burst phase of star formation(Bastian et al.,2005),and hence we expect large number of clusters of intermediate age(∼100Myr)in its disk.Hence,M82offers an excellent opportunity to assess the evolutionary effects on the survival of star clusters,and to look for a possible evolutionary connection between the SSCs and GCs.2Data Extraction,Source Selection,and Simulations The observational data used in this work consisted of images in F435W(B), F555W(V)and F814W(I)filters,that were obtained by the Hubble Her-itage Team(Mutchler et al.,2007)using the ACS/WFC instrument aborad the Hubble Space Telescope(HST).Bias,dark,andflat-field corrections were carried out using the standard pipeline process by the Heritage Team.The final reduced science quality images cover the entire optical disk of the galaxy with a spatial sampling of0.05arcsec pixel−1,which corresponds to0.88par-sec pixel−1at M82’s distance of3.63Mpc(Freedman et al.,1994).The point sources have a size distribution that peaks at a Full Width at Half Maximum (FWHM)of2.1pixels,with the tail of the distribution extending to3.0pixels (or2.6parsec).Very few clusters are expected to have sizes smaller than3par-sec,and hence clusters can be distinguished from the stars on these images.A circle of500pixels(450pc)radius is used to separate the nuclear region from the disk.The clusters inside this radius are associated with strong Hαemitting complexes,and hence are younger than10Myr(Melo et al.,2005). On the other hand,the disk outside the450pc radius shows characteristic signatures of post-starburst conditions,with hardly any Hαemission.We used SExtractor(Bertin&Arnouts,1996)independently on the B,V, and I-band images to construct an unbiased sample of cluster candidates.A source having a FWHM>3pixels and an area of at least50adjacent pixels,Star clusters in M823 each of S/N>5is considered a cluster candidate.All the bright sources sat-isfying these criteria are genuine clusters,but at fainter magnitudes majority of the candidate sources lack the symmetry expected for a physical cluster. These are found to be artificial extended sources formed due to the super-position of stars in this nearly edge-on galaxy.These artificial sources most often are elongated,and are rejected automatically from the sample using the ellipticity parameter of SExtractor.Cluster candidates in eachfilter were then combined,the common sources being counted only once.The resulting list contains653clusters,260of them belonging to the nuclear region.For all the sources in thefinal list,aperture photometry is carried out in all the three bands.The FWHM calculated by SExtractor is used as a measure of the size of the clusters.The observed cluster luminosity function(LF)follows a power-law at the bright end,turning over sharply at faint magnitudes.Similarly,the size distri-bution function peaks at a characteristic value of FWHM∼10pixels.Monte Carlo simulations were carried out to check the effect of incompleteness of cluster detection on the observed functional forms.In the simulations,each cluster is assumed to be round and to have a Gaussian intensity profile of a given FWHM.A power-law distribution function is used to assign a luminosity to each cluster.Two separate simulations are done,one in which a cluster is assigned a size based on a power-law size distribution function,and the other based on a log-normal function.The simulated luminosity function resembles very much the observed one,implying that the observed turn-over of the lu-minosity function is due to incompleteness at the faint end and not intrinsic to the cluster population.Hence,the turn-over in the luminosity function,if any,would correspond to a magnitude fainter than B=22mag.On the other hand,the observed size function points to an intrinsically log-normal size dis-tribution,rather than a power-law function.A more detailed description of the selection process,observed luminosity and size functions,and the Monte Carlo simulations can be found in Mayya et al.(2007).3Physical Parameters of ClustersWe analyzed the color and magnitude of the individual clusters to obtain their reddening and mass,making use of solar metallicity Single Stellar Population (SSP)models of Girardi et al.(2002).These authors provide the evolutionary data on colors and magnitudes for the instrumental HST/ACSfilters,a fact that enables us a direct comparison with the observed data.The Kroupa (2001)initial mass function(IMF)in its corrected version has been used.It has nearly a Salpeter slope(2.30instead of2.35)for all masses higher than 1M⊙.The derived masses depend on the assumption of the lower cut-offmass of the IMF.In the case of standard Kroupa’s IMF,the derived masses would be around2.5times higher.4Mayya et al.3.1The Ages of ClustersColors obtained by the combination of the threefilters that we used suffer from age-reddening degeneracy,and hence it was necessary to assume one of the quantities to obtain the other.We found that the observed range in colors is too large to be explained by evolutionary effects,even for stellar populations as old as10Gyr.On the other hand,age of the principal stellar populations in the nuclear region of M82is determined in innumerable studies(Rieke et al., 1993;F¨o rster Schreiber et al.,2003,and references therein),and it is found to be<10Myr.Based on these studies,we adopt an age of8Myr for the nuclear clusters.Most of the disk stars in M82were formed in a violent burst around500Myr ago.Ages of those clusters for which spectroscopic data are available(Smith et al.,2006)lie in the range between50–500Myr,suggesting that the clusters are formed during or immediately after the disk-wide star formation epoch.3.2Color-Magnitude DiagramsFrom the assumed ages(8Myr for the nuclear clusters,and50–500Myr for the disk clusters),and the very likely hypothesis that the extinction is the main cause of the dispersion in the observed colors,we can estimate the masses of the clusters.The method we have followed is illustrated in Figure1.For a given position in the Color Magnitude Diagram(CMD),we derived the extinction by comparing the observed colors with those of the SSP.Once the extinction is determined,we calculate the mass using the extinction-corrected luminosity and the mass-to-light ratio of such SSP.The disk masses are derived assuming an age of100Myr.The mass estimates would be higher by a factor of3.2, if the clusters are as old as the stellar disk(500Myr).On the other hand,if the clusters are as young as50Myr,the masses would be lower by a factor of1.6.The distribution of the derived visual extinction values is peaked at ∼1mag for the disk clusters,whereas it isflat between1–4mag for the nuclear clusters.3.3Mass Distribution FunctionThe determination of the cluster masses for our complete sample enables us to derive the present-day Cluster Mass Function(CMF).In the left panel of Figure2,we plot the CMF separately for the nuclear and disk clusters.The nuclear CMF is scaled up to match the disk CMF at1.5×106M⊙.Poissonian error bars are indicated.The distribution for both samples follows a power-law over almost two orders of magnitude in mass for cluster masses above ∼2×104M⊙.However,the power-law index for the disk and nuclear cluster populations shows a clear difference,α=1.8±0.1for the nuclear clusters, and1.5±0.1for the disk population.Studies of young star clusters in nearbyStar clusters in M8250122422201816-4-6-8-10-120122422201816-4-6-8-10-120122422201816-4-6-8-10-120122422201816-4-6-8-10-12Fig.1.Observed color-magnitude diagrams (CMDs)for the nuclear (filled circles)and disk clusters,in M82.(Top left )Evolutionary track for an SSP of a cluster mass of 105M ⊙is superposed.Two vectors,placed at 8Myr and 500Myr,show the location of the track reddened by A v =3mag.In the top-right panel,we show the CMD for the nuclear clusters only.The locations of an 8Myr SSP for a range of cluster masses and visual extinctions are shown by the superposed grid.Mass varies vertically along the grid (in solar units),whereas the visual extinction (in magnitude)varies along the diagonal axis.In the bottom panels,we show a similar diagram for the disk clusters,with the superposed grids corresponding to fixed ages of 100Myr (left)and 500Myr (right).In all the panels,tick mark values of the right-vertical axis correspond to the absolute magnitude in the V -band.6Mayya etal.Fig.2.(Left)Mass functions for the nuclear (dotted line)and disk (solid line)cluster samples.Both the samples follow a power-law distribution between 2×104M ⊙and 106M ⊙.The best-fit indices in this mass range are indicated.(Right)Mean size (FWHM)of the clusters as a function of mean mass for three mass bins for the nuclear (young)and disk (old)samples.The error bar denotes the rms dispersion about the mean value.High mass clusters have similar mean sizes irrespective of their evolutionary status.On the other hand,mean size of the low-mass clusters clearly decreases as they become older.Among the young clusters,low-mass ones are more extended than higher mass ones.galaxies yield a value of αclose to 2.0(de Grijs et al.,2003).Hence,α=2.0can be considered as the expected slope of the initial CMF.In general,the cluster size distribution function (CSF)for the nuclear and disk clusters follow a log-normal form.However,the mean,as well as the maximum cluster sizes are systematically smaller for the lower mass bins.This tendency is illustrated in the right panel of Figure 2,where the mean cluster size for each mass bin has been plotted against the mean mass of clusters in that bin,for the young and old ones,separately.For the highest mass bin,the mean sizes of the young and old clusters are similar.The mean size decreases systematically with decreasing cluster mass for the old clusters,whereas the inverse is true for the young clusters.4On the Survival Chances of Star Clusters in M82The observed differences in the CSF for young and old clusters are consistent with the expected evolutionary effects.Both the disruption of the loose OB associations and the dynamical trend towards relaxation would diminish the number of large low-mass systems.Thus,the destruction process is both mass and size dependent,with the most extended clusters in each mass bin being the most vulnerable to disruption.All clusters of masses higher than 105M ⊙areStar clusters in M827 still surviving∼108yr after their formation.In this section,we discuss these observational results in the context of theoretical models of cluster disruption, and their possible survival to become globular clusters.At early times,disruption is caused mainly due to the expulsion of the intra-cluster gas through supernova explosions.This process is ineffective once all the high mass stars in the cluster die,which happens in around∼30Myr. Hence,the observed disk clusters have survived this early mechanism of dis-ruption.On intermediate timescales(107<t<few×108yr),the mass-loss from evolving stars leads to a decrease in the cluster mass from its initial value.Clusters can loose as much as30%of their stellar mass during their evolution.The decreased cluster mass can result in the expansion of the clus-ter,finally leading to its disruption.However,this process of disruption is equally effective for high and low mass clusters,and hence a change in the slope of mass function is not expected.The observedflattening of the mass function at older ages,suggests that the cluster disruption process that is ac-tive in M82selectively destroys low-mass clusters.The tidal effect experienced by the clusters as they move in the gravitationalfield of the parent galaxy is one such process.According to Fall&Zhang(2001),this process becomes important after∼300Myr in normal galaxies.However,in the case of M82, de Grijs et al.(2005)have estimated a disruption timescale as short as30Myr for a cluster of mass104M⊙at1kpc away from the center,with a depen-dence on mass that varies as M0.6.The short timescale in M82implies that the surviving clusters in the disk are presently experiencing the dynamical processes of cluster disruption.If the trend of selective disruption of loose clusters continues,how many of the present clusters will survive for a Hubble time?Can the LF of the sur-viving clusters look like that of the Galactic GCs?In Figure3,we show the evolutionary effects on the LF of the M82clusters.The histogram with dashed lines shows the LF considering the photometric evolution of the clusters for 5Gyr,whereas the solid histogram shows the same,but after taking into ac-count the dynamical effects as well.The latter is implemented in a simplistic way,by imposing the condition that for the clusters to survive the dynamical effects,their half-light radius,R eff,should be smaller than the tidal radius, R t,for that cluster.For a cluster of mass M C at galactocentric radius R G,R t is given by the expression(Spitzer,1987),R t= M C8Mayya et al.which implies that the galactocentric distance of a cluster will change with time.The disruption of a cluster depends on the net tidal force received by its stars as it orbits the galaxy during its lifetime.We found that for an assumed R G=350pc,the future LF of M82will resemble that of the Galactic GCs. Even in this extreme case,85clusters will survive,as compared to the146 GCs in the Milky Way.The number of GCs in a galaxy scale with the mass of the parent galaxy,and considering that M82is an order of magnitude less massive than the Milky Way,only∼15GCs are expected to present in M82. Thus,number of clusters that will survive represent an over-abundance by a factor of5–6.For ages older than this,the distribution is similar except that the peak of the distribution shifts to∼0.5mag fainter.Thus,the compact star clusters in M82will evolve into GCs.Fig.3.Present(dotted histogram)and future luminosity functions of M82star clusters with(solid histogram)and without(dashed histogram)taking into account dynamical effects of evolution at an age of5Gyr.The log-normal function repre-senting the luminosity function of the Galactic Globular Clusters is shown by the dot-dashed curve.Star clusters in M829 5ConclusionsLuminosity and Mass functions of star clusters in M82follow power-law func-tions,with the power law index showing a tendency forflattening of the profile with age.In other words,there is a deficiency of low-mass clusters among the older clusters.We alsofind the mean size of the older clusters to be smaller as compared to the younger clusters for masses<105M⊙.These two results together imply the selective destruction of loose clusters.The tidal forces ex-perienced by the clusters as they orbit around the galaxy lead to exactly such a destruction process.If this process continues in M82,the LF of surviving clusters can mimic the presently observed LF of the Galactic GCs,provided the clusters move around the galaxy in highly elliptical orbits,with perigalac-tic distance as small as350pc.The resulting LF contains85clusters with the function peaking at the same luminosity as for the Galactic GCs at5Gyr age,and fainter by∼0.5mag at10Gyr.On the other hand,if the clusters move in nearly circular orbits,the LF will retain the power-law form,with the number of surviving clusters even higher.This work is partly supported by CONACyT(Mexico)research grants 42609-F and49942-F.We would like to thank the Hubble Heritage Team at the Space Telescope Science Institute for making the reducedfitsfiles available to us.ReferencesBastian,N.,Gieles,M.,Lamers,H.J.G.L.M.,Scheepmaker,R.A.,&de Grijs,R.2005,A&A,431,905Bertin,E.,&Arnouts,S.1996,A&AS,117,393de Grijs,R.,Anders,P.,Bastian,N.,Lynds,R.,Lamers,H.J.G.L.M.,& O’Neil,E.J.2003,MNRAS,343,1285de Grijs,R.,Parmentier,G.,&Lamers,H.J.G.L.M.2005,MNRAS,364, 1054de Grijs,R.,&Parmentier,G.2007,Chinese Journal of Astronomy and As-trophysics,7,155Fall,S.M.,&Zhang,Q.2001,ApJ,561,751F¨o rster Schreiber,N.M.,Genzel,R.,Lutz,D.,&Sternberg,A.2003,ApJ, 599,193Freedman,W.L.,et al.1994,ApJ,427,628Girardi,L.,Bertelli,G.,Bressan,A.,Chiosi,C.,Groenewegen,M.A.T., Marigo,P.,Salasnich,B.,&Weiss,A.2002,A&A,391,195Greve,A.,Wills,K.A.,Neininger,N.&Pedlar,A.A&A,383,56Kroupa P.,2001,MNRAS322,231Ma´ız-Apell´a niz,J.2004,Astrophysics and Space Science Library,315,23110Mayya et al.Mayya,Y.D.,Bressan,A.,Carrasco,L.,&Hernandez,L.2006,ApJ,649, 172Mayya,Y.D.,Romano,R.,Rodr´ıguez-Merino,L.H.,Luna,A.,Carrasco,L. &Rosa-Gonz´a lez,D.2007,ApJ submittedMelo,V.P.,Mu˜n oz-Tu˜n´o n,C.,Ma´ız-Apell´a niz,J.,&Tenorio-Tagle,G.2005, ApJ,619,270Mutchler,M.,Bond,H.E.,Christian,C.A.,Frattare,L.M.,Hamilton,F., Januszewski,W.,Levay,Z.G.,Mountain,M.,Noll,K.S.,Royle,P.,Gal-lagher,J.S.,&Puxley,P.2007,PASP,119,1O’Connell,R.W.,Gallagher,J.S.,III,Hunter,D.A.,&Colley,W.N.1995, ApJL,446,L1Rieke,G.H.,Loken,K.,Rieke,M.J.,&Tamblyn,P.1993,ApJ,412,99 Smith,L.J.,Westmoquette,M.S.,Gallagher,J.S.,O’Connell,R.W., Rosario,D.J.,&de Grijs,R.2006,MNRAS,370,513Sofue,Y.1998,PASJ,50,227Spitzer,L.1987,Princeton,NJ,Princeton University Press,1987,p15。