Red Halos of Galaxies - Reservoirs of Baryonic Dark Matter?
E. Zackrisson, N. Bergvall, C. Flynn, G. Ostlin, G. Micheva, B. Caldwell
aa r X i v : . [ a s t r o - ph ] A ug Dark Galaxies and Lost BaryonsProceedings IAU Symposium No. 244, 2007XXX, eds. c (cid:13) Red halos of galaxies – Reservoirs ofbaryonic dark matter?
E. Zackrisson , , , N. Bergvall , C. Flynn ,G. ¨Ostlin , G. Micheva and B. Caldwell Tuorla Observatory, University of Turku, V¨ais¨al¨antie 20, FI-21500 Piikki¨o, Finland Stockholm Observatory, AlbaNova University Center, 106 91 Stockholm, Sweden Department of Astronomy and Space Physics, Box 515, 751 20 Uppsala, Sweden
Abstract.
Deep optical/near-IR surface photometry of galaxies outside the Local Group haverevealed faint and very red halos around objects as diverse as disk galaxies and starburstingdwarf galaxies. The colours of these structures are too extreme to be reconciled with stellarpopulations similar to those seen in the stellar halos of the Milky Way or M31, and alternativeexplanations like dust reddening, high metallicities or nebular emission are also disfavoured. Astellar population obeying an extremely bottom-heavy initial mass function (IMF), is on theother hand consistent with all available data. Because of its high mass-to-light ratio, such apopulation would effectively behave as baryonic dark matter and could account for some of thebaryons still missing in the low-redshift Universe. Here, we give an overview of current red halodetections, alternative explanations for the origin of the red colours and ongoing searches for redhalos around types of galaxies for which this phenomenon has not yet been reported. A numberof potential tests of the bottom-heavy IMF hypothesis are also discussed.
Keywords.
Galaxies: halos – galaxies: stellar content – dark matter – Galaxy: halo – stars:subdwarfs
1. Introduction
The quest to unravel the nature of the dark matter, estimated to account for ≈ ≈ /
3; Fukugita 2004; Fukugita& Peebles 2004) are still at large in the local Universe.The old idea of baryonic dark matter in the form of faint, low-mass stars has recentlygained new momentum through surface photometry detections of very red and exceed-ingly faint structures – so-called red halos – around different types of galaxies. Thehistory of this topic goes back to the mid-90s, when deep optical and near-IR images(e.g. Sackett, Morrison, Harding, et al. et al. et al. et al. (1994) also announced the detection of a red haloaround the cD galaxy at the centre of Abell 3284. Skepticism grew with the discoveryof what appeared to be the remnants of a disrupted dwarf galaxy close to NGC 5907(Shang, Brinks, Zheng, et al. et al. et al. et al. (2000) failed to detect anyred halo at wavelengths of 3–5 µ m, but since their upper limit of 15% on the contributionfrom hydrogen-burning stars to the overall dark matter of this galaxy is more or lessidentical to the cosmic baryon fraction (Ω baryons / Ω M ≈ .
15; Spergel, Bean, Dor´e, et al. baryonic component of the dark matter.A few years later, Bergvall & ¨Ostlin (2002) and Bergvall, Marquart, Persson et al. (2005)discovered similar faint and abnormally red structures in deep optical/near-IR images ofblue compact galaxies (BCGs). Even more recently, Zibetti, White & Brinkmann (2004)stacked the images of 1047 edge-on disk galaxies from the Sloan Digital Sky Survey(SDSS) and detected a halo population with a strong red excess and optical colours curi-ously similar to those previously derived for NGC 5907 – again very difficult to reconcilewith standard stellar populations. The halo detected around an edge-on disk galaxy atredshift z = 0 .
322 in the Hubble Ultra Deep Field shows similarly red colours (Zibetti &Ferguson 2004).
2. A bottom-heavy stellar initial mass function?
Zackrisson, Bergvall, ¨Ostlin et al. (2006) analyzed the colours of these new detectionsand found that the halos of both BCGs and edge-on disks could be explained by a stellarpopulation with a very bottom-heavy IMF ( dN/dM ∝ M − α with α ≈ .
50, where α = 2 .
35 represents the Salpeter slope). For an IMF slope as extreme as this, only starswith masses 0 . M ( M ⊙ ) M/L B & α ≈ > M ⊙ (Massey 2002; Gouliermis, Brandner & Henning 2006).The characteristic mass of the putative red halo stars (0 . M ( M ⊙ )
1) also coincideswith that of the claimed MACHO detections in the halos of the Milky Way and M31(Alcock, Allsman, Alves, et al. et al. et al.
3. Ongoing searches
To get to the heart of the red halo mystery, our group has begun a large number ofobservational searches for red halos around types of galaxies for which this phenomenonhas not yet been reported. This includes post-starburst galaxies (PI Zackrisson), ellipticalgalaxies (PI Bergvall), and dwarf galaxies in the Local Group (PI Zackrisson). We arealso obtaining multiband data for an extended sample of BCGs (PI ¨Ostlin). If someclasses of objects systematically turn out to have red halos, whereas others don’t, thiswill provide important clues to the formation of these structures. A non-detection of red ed halos of galaxies et al. (2004) detected a red halo around edge-on disk galaxies throughthe stacking of high surface brightness disks the SDSS, our team has recently carriedout the same procedure for ≈ low surface brightness disks (Caldwell & Bergvall2007; Bergvall et al. in preparation). This resulted in the detection of a red halo withcolours even more extreme than the ones previously discovered around edge-on disks(both the stacked SDSS disks and NGC 5907). The fact that a red excess (compared towhat can easily be explained by a stellar population with a Salpeter-like IMF) has nowbeen seen in a completely independent sample of stacked disks suggests that the Zibetti et al. detection is unlikely to be a statistical fluke (i.e. a “2 σ effect”). Whatever its origin,the red halo phenomenon seems to be here to stay.
4. Potential tests
The hypothesis that the red halo colours are generated by a stellar population with anabnormally high fraction of low-mass stars can in principle be tested using a number ofdifferent methods, which we outline below.( a ) Improved optical/near-IR data . Once red halo data in more filters (e.g.
BV RIJHK ) become available, the bottom-heavy IMF may no longer be able to explain thespectral energy distribution, and can then be rejected. With H α and H β emission-linedata (obtained through narrowband photometry), the contribution from nebular emissionat large projected distances from the centre of the target galaxies can also be assessed.This test is currently being implemented using very deep multiband data (including H α and H β ) obtained at the 2.5m Nordic Optical Telescope (PI Zackrisson) for UM456, aBCG for which previous, shallower data has already revealed the presence of a very redhalo. Near-IR surface photometry data obtained from space (e.g. with AKARI) wouldalso be very useful, as this could alleviate some of the difficulties associated with ground-based measurements of the red halo signal against the very bright near-IR sky.( b ) Mid-IR data . A large population of preferentially low-mass stars should be de-tectable at near/mid-IR wavelengths with the Spitzer Space Telescope. Spitzer data forthe halo regions of a handful of BCGs has already been obtained (PI Bergvall), and theanalysis is underway.( c ) Spectroscopy of red halos . While extremely challenging at the faint surfacebrightness levels at which the red halos are detected, spectroscopy could potentiallyreveal signatures in the spectral energy distribution of the red halos that are inconsistentwith a bottom-heavy IMF. Integral-field spectroscopy with VLT/VIMOS has recentlybeen attempted for a handful of BCG halos (PI Bergvall).( d ) Star counts in nearby galaxies . Since a bottom-heavy IMF would generatea much smaller fraction of red giants compared to a normal (i.e. Salpeter-like) IMF,the bottom-heavy IMF hypothesis could be falsified through direct star counts in haloregions for which surface photometry data have already revealed extremely red colours.Individual giants can be detected out to distances of ≈
10 Mpc with the HST and verydeep data is already available in the HST archive for a number of BCGs within thisdistance. The ground-based surface photometry required to perform this test has beencarried out at the 2.5m Nordic Optical Telescope (PI Zackrisson). A test of this typehas already been carried out by Zepf et al. (2000) for the red halo of NGC 5907, withthe intriguing result that the small number of giants detected indeed seemed to favour abottom-heavy IMF. Because of the controversial nature of the surface photometry data Zackrisson et al.for this object, it is, however, not entirely clear how robust this conclusion really is. Anabsence of red giants would moreover not rule out a non-stellar explanation for the redcolours observed.( e ) Star counts in Local Group galaxies
While the previous methods all have thepotential to falsify a bottom-heavy IMF as the origin of the observed red halo colours,none of them will be able to provide a direct confirmation of the presence of unusuallymany low-mass stars. To accomplish this, one has to turn to objects within the LocalGroup, where stars at masses < M ⊙ can be individually resolved. Unfortunately, nored halos have yet been discovered (or looked for) among the dwarf galaxies of the LocalGroup. The main reason for this is that deep surface photometry becomes exceedinglydifficult for galaxies with large angular sizes. If Local Group dwarf galaxies do exhibitred halos, these are likely to completely fill the field of view of current detectors (about10 ′ × ′ for typical CCDs and 30 ′ × ′ for wide field cameras), leaving few or no regionsin the resulting image from which a clean measurement of the sky flux can be made. To beable to reliably subtract the sky down to the very faint surface brightness levels where redhalos are expected to be seen, one would either have to employ very rapid sky chopping(even in the optical), or alternatively target only part of the halo region, hoping that anylarge-scale gradients in the sky flux can be corrected for. While admittedly challenging,we are currently attempting this at the ESO 2.2m/WFI (PI Zackrisson). If successful,follow-up star counts should be able to robustly confirm or reject the connection betweenlow-mass stars and the red excess. Deep HST images are available for many of the LocalGroup dwarfs, but with a few exceptions, these have been obtained too close to thecentres of these galaxies, where a red halo signal in unlikely to be detectable.
5. A red halo of low-mass stars around the Milky Way?
While the identification of a red halo around stacked high-surface brightness diskgalaxies in the SDSS (Zibetti et al. all disk galaxieshave a red structure of this type. It is nonetheless interesting to ask whether the MilkyWay itself could be surrounded by a hitherto undetected red halo of low-mass stars,with photometric properties similar to that detected around stacked edge-on disks. Thedetected baryonic components (thin disc, thick disk, bulge and known stellar halo) ofthe Milky Way contribute around 5–6 × M ⊙ (e.g. Sommer-Larsen & Dolgov 2001;Klypin, Zhao & Sommerville 2002) to its virial mass of ≈ × M ⊙ (Klypin et al. baryons / Ω M ≈ .
15 (Spergel et al. ≈
90% of the cosmicaverage for a galaxy-sized halo (Crain, Eke, Frenk et al. ≈ . × M ⊙ of baryonic material within its virial radius, leavingaround 60% of it still to be found.If some of the missing baryons of the Milky Way are locked up in the form of hydrogen-burning stars in a red halo, such a structure must somehow have evaded the faint starcounts aimed to constrain the luminosity function of halo subdwarfs (e.g. Gould, Flynn &Bahcall 1998; Gould 2003; Digby et al. uniform halo population with anIMF as extreme as that envisioned by Zackrisson et al. (2006) can safely be ruled out,a halo with a strong radial population gradient (with an abnormal fraction of low-massstars only at large Galactocentric distances) may be more difficult to disqualify. ed halos of galaxies
6. Alternative explanations
The idea of a halo population with an abnormally high fraction of low-mass starsis admittedly controversial, but so far, no alternative explanation has come close toexplaining the red halo colours observed. The more mundane mechanisms that havebeen considered include:( a ) Nebular emission . Zibetti et al. (2004) suggest that the red halo colours couldbe caused by diffuse nebular emission (i.e. emission-lines and continuum associated withthe ionized interstellar medium), and although such radiation may indeed be presentat large distances from sites of active star formation, current models (Zackrisson et al. b ) High metallicity . Rudy et al. (1997), Bergvall & ¨Ostlin (2002) and Zibetti et al. (2004) have all suggested that the red halo colours may (at least partly) be explainedby a stellar population with a high metallicity. However, as demonstrated by Zackrisson et al. (2006), a high-metallicity population with a Salpeter-like IMF fails to explain thecolours of the red halo detected around disk galaxies in stacked SDSS data. Due tothe degeneracy between bottom-heavy IMFs and high metallicities in
BV K filters, ahigh-metallicity population with a normal IMF can in principle explain the red halos ofBCGs, but this would require the metallicity to be very high (Solar or higher), whichis in conflict with the gaseous metallicities typically observed in BCGs ( ≈
10% Solar).While the stellar population outside the bright star-forming centres of BCGs (where thegaseous metallicites are measured) may indeed be more metal-rich, the high metallicitiesinferred would require BCGs to be far more massive than indicated by current estimates(Bergvall & ¨Ostlin 2002). This degeneracy between high stellar metallicities and bottom-heavy IMFs can be broken via I -band observations (Zackrisson et al. c ) Dust reddening . While dust reddening can, in principle, give rise to very redcolours, normal dust reddening vectors (i.e. Milky Way, LMC, SMC dust) run almostparallel to the age vectors for old, normal-IMF stellar populations). While dust may beable to mimic a high age, it cannot explain the extreme colours observed in red halos.Assuming that the central star-forming regions of BCGs are encompassed from all anglesby their red halos, the low extinction measured through the H α/ H β emission-line ratiosin BCGs (Bergvall & ¨Ostlin 2002) moreover imposes a very strong upper limit on theamount of dust reddening in the halo. At least for these objects, dust reddening does notappear to be a viable explanation for the red halo phenomenon.( d ) Dust emission . Emission from hot dust could in principle help explain the K -band excess seen in the red halos of BCGs, but at least in the case of Haro 11 – oneof the BCGs with the reddest halo observed so far – ISO observations have ruled outnear-IR dust emission as an explanation for the colours observed (Bergvall & ¨Ostlin2002). Extended red emission from photoluminescent dust (e.g. Witt & Vijh 2004) couldpossibly contribute to the red excess of halos observed only in the optical (i.e. the redhalo of stacked SDSS disks), but not likely to the red excess seen at longer wavelengths. Zackrisson et al.( e ) Spectral synthesis problems . The fact that a low-metallicity stellar populationobeying a Salpeter-IMF fails to explain the observed red halos has been confirmed byseveral independent spectral evolutionary models. Recent updates related to the evolutionof thermally pulsating asymptotic giant branch stars moreover seem to have but a modestimpact on the interpretation of the colours used (Maraston 2005; Bruzual 2007), anddo not appear to be able to challenge this conclusion. The models moreover have noproblem in explaining the colours of other red stellar populations like globular clustersor elliptical galaxies. There are, however, substantial uncertainties concerning the best-fitting age, metallicity and IMF slope derived from the red halo data in the frameworkof a bottom-heavy IMF, since current models tend to treat low-mass stars ( < . M ⊙ )in a very superficial manner or omit them completely. While justified in the case of aSalpeter-like IMF, such stars give a substantial contribution to the integrated light ofbottom-heavy IMF populations, and should therefore be treated as carefully as possible.To alleviate this problem, a model of spectral evolution more suited for the predictionof the photometric properties of stellar populations obeying extreme IMFs is currentlyunder development (Zackrisson, in preparation).( f ) Sky subtraction problems . The brightness of the night sky is about µ B ≈
23 mag arcsec − at the darkest sites on Earth, whereas red halos are detected at surfacebrightness levels of µ B ≈ − for BCGs and µ g ≈
28 mag arcsec − for thestacked SDSS disks. Hence, we are trying to measure brightnesses that may be as low as1% of the sky (and in the redder bands even less). The sky itself is also very red, implyingthat even a tiny undersubtraction of the sky could lead to a spurious detection of a veryred structure. One way of testing for such sky subtraction problems is to insert syntheticimages of galaxies (based on analytical surface brightness profiles, e.g. exponential orSersic profiles, with known parameter values) into the observed sky frames and checkwhether the reduction routines used are able to return a surface brightness profile thatis consistent with that used as input, out to the surface brightness levels where red halosare detected. Tests indicate that our sky subtraction software does indeed pass this trial.Bad things can and will happen once one attempts to approach surface brightness levelssignificantly below that of the extragalactic background light (Zackrisson & ¨Ostlin, inpreparation), but since many of our red halo detections have been made at isophotesbrighter than this, problems in properly correcting for the extragalactic background isunlikely to be the cause of the red halo phenomenon.( g ) Point spread function effects . One may also worry that the red halos may bethe result of some instrumental artefact. Since red halos have so far only been detected inthe outskirts of bright stellar systems, selectively scattered light from the central galaxiescould in principle be a viable explanation for their unusual colours. Michard (2002) arguethat, depending on the quality of the telescope mirror, the point spread function (PSF)may indeed develop extended red wings which would give rise to diffuse, red structures inthe vicinity of bright objects. While Zibetti et al. (2004) estimate that the PSF may givenon-negligible contribution to the halo signal observed in stacked SDSS data, they argue,based on measurements of SDSS PSF, that this cannot shift the colours substantially inthe redward direction. Measurements of the PSF in the images where the red halos ofBCGs have been detected have furthermore failed to reveal any similar red halos aroundstars, indicating that these detections are unlikely to result from PSF effects. Moreover,the red halos observed around BCGs sometime appear to be displaced relative to theBCG centre. To advocate PSF effects to explain such structures would require the PSFto be very asymmetric, and there is no evidence for that in the data. ed halos of galaxies
7. Red halo formation
If the red halo phenomenon is indeed caused by a stellar population overly abundantin low-mass stars, how would such a structure have formed? In the currently favouredscenario for stellar halo formation, the halo stars originally formed in less massive darkmatter halos. As these low-mass halos fell into the potential well of a more massive galaxy,they disrupted due to tidal effects and redistributed their stellar content in the form of adiffuse halo of stars surrounding the centre of the more massive galaxy (see simulationsby e.g. Bullock & Johnston 2005; Abadi, Navarro & Steinmetz 2006). Simulations of thiskind can of course only be expected to predict the gross features of such halos, since allthe star formation details remain hidden on scales far below current resolution limits,and have to be inserted in the form of semi-analytical assumptions. While a red haloof low-mass stars is certainly not predicted by current simulations of this type, it isnot obvious that such structures necessarily are inconsistent with this picture of stellarhalo formation. It may indeed be possible to explain the existence of red halos in thisframework, given suitable modifications of the star formation details that are inserted“by hand”.At the current time, we can do little but speculate on the mechanisms responsible forthe emergence of red halos. Here are a few scenarios that come to mind:( a ) Population III stars . Population III stars (i.e. stars with metallicity Z ≈
0) areexpected to be distributed throughout the dark halo of a Milky Way-sized galaxy, albeitwith a spatial distribution different from that of the cold dark matter (Scannapieco,Kawata, Brook, et al. b ) A bottom-heavy IMF in dark galaxies . Simulations based on ΛCDM generi-cally predict that each galaxy-mass cold dark matter halo should contain a huge numberof subhalos (typically accounting for ≈ ∼ et al. et al. et al. c ) Dynamical mass segregation.
In principle, halo colours indicative of an abnor-mally high fraction of low-mass stars need not necessarily imply a bottom-heavy IMF.We may instead be witnessing the effects of dynamical mass segregation. Given sufficienttime, equipartition of kinetic energy will tend to slow high-mass stars down and cause Zackrisson et al.them to sink towards the centre of a stellar system, whereas low-mass stars will speed upand statistically end up at larger distances from the centre. Assuming that this processhas had time to be efficient in each of the low-mass halos that eventually merged to formthe high-mass systems that we witness today, one may envision that low-mass stars inthe outer parts of these small halos were more easily stripped from these systems thanthe high-mass stars in the centre. This could give rise to a situation in which low-massstars preferentially end up in the stellar halo, whereas high-mass stars tend to sink to thecentre of the parent galaxy. While this could possibly explain the many low-mass starsthat seem to inhabit the red halos, this scenario requires an excess of high-mass starssomewhere else in the galaxies that exhibit the red halo phenomenon.( d ) In situ formation of low-mass stars . A final option is of course that low-massstars formed in situ in the halos of these systems, possibly in a cooling flow (Mathews& Brighenti 1999).
8. Summary
The red halos detected through deep optical/near-IR surface photometry of differenttypes can currently only be understood by advocating a stellar halo population harbour-ing an abnormally high fraction of low-mass stars, possibly as an effect of a bottom-heavy IMF. Due to its high mass-to-light ratio, such a population effectively behavesas baryonic dark matter and could help explain the baryons still missing from currentinventories relevant for the low-redshift Universe. Searches for red halos around types ofgalaxies have not yet been reported (post-starburst galaxies, elliptical galaxies and LocalGroup dwarfs) are currently underway, along with a number of observational tests aimedto constrain the nature of the red halos.
Acknowledgements
EZ acknowledges research grants from the Swedish Research Council, the Academy ofFinland and the Swedish Royal Academy of Sciences. CF acknowledges support from theAcademy of Finland.
References
Abadi, M.G., Navarro, J.F., Steinmetz, M. 2006,
MNRAS , 365, 747Alcock, C., Allsman, R.A., Alves, D.R, et al.
ApJ , 542, 281Bergvall, N. & ¨Ostlin, G. 2002,
A&A , 390, 891Bergvall, N., Marquart, T., Persson, C., Zackrisson, E. & ¨Ostlin, G. 2005, in: A. Renzini. &R. Bender (eds.),
Multiwavelength Mapping of Galaxy Formation and Evolution , Springer-Verlag, Berlin, p. 355Brandner, W. 2005, in: E. Corbelli, F. Palle & H. Zinnecker (eds.),
The Initial Mass Function50 years later , Astrophysics and Space Science Library Volume 327, Springer, Dordrecht,p. 101Bruzual, G. 2007, to appear in: A. Vallenari, R. Tantalo, L. Portinari, A. Moretti (eds.),
From Stars to Galaxies: Building the Pieces to Build Up the Universe , ASP Conf. Ser.,(astro-ph/0702091)Bullock, J.S. & Johnston, K.V. 2005,
ApJ , 635, 931Calchi Novati, S., Paulin-Henriksson, S., An, J., et al.
A&A , 443, 911Caldwell, B. & Bergvall, N 2007, in: F. Combes & J. Palous (eds.),
Galaxy Evolution Across theHubble Time , Cambridge University Press, p. 82Crain, R.A., Eke, V.R., Frenk, C.S., Jenkins, A., McCarthy, I.G., Navarro, J.F., Pearce, F.R.2007,
MNRAS , 377, 41Digby, A.P., Hambly, N.C., Cooke, J.A., Reid, I.N., Cannon, R.D. 2003,
MNRAS , 344, 583 ed halos of galaxies Fukugita, M. 2004, in: S.D. Ryder, D.J. Pisano, M.A. Walker, K.C. Freeman (eds.),
InternationalAstronomical Union Symposium no. 220 , Astronomical Society of the Pacific., p. 227Fukugita, M. & Peebles, P.J.E. 2004,
ApJ , 616, 643Gould, A., Flynn, C. & Bahcall, J.N. 1998,
ApJ , 503, 798Gould, A. 2003,
ApJ , 583, 765Gouliermis, D., Brandner, W. & Henning, T., 2006,
ApJ , 641, 838James, P.A. & Casali, M.M. 1998,
MNRAS , 301, 280Klypin, A., Kravtsov, A.V., Valenzuela, O., Prada, F. 1999,
ApJ , 522, 82Klypin, A., Zhao, H. & Somerville, R.S. 2002,
ApJ , 573, 597Lequeux, J., Fort, B., Dantel-Fort, M., Cuillandre, J.-C., Mellier, Y. 1996,
A&A (Letters), 312,L1Massey, P. 2002,
ApJS , 141, 81Maraston, C. 2005,
MNRAS , 362, 799Mathews, W.G. & Brighenti, F. 1999,
ApJ , 526, 114Michard, R. 2002,
A&A , 384, 763Molinari, E., Buzzoni, A., Chincarini, G., Pedrana, M.D. 1994,
A&A
ApJ (Letters), 524, L19Rudy, R.J., Woodward, C.E., Hodge, T., Fairfield, S.W., Harker, D. 1997,
Nature , 387, 159Sackett, P.D., Morrison, H.L. Harding, P., Boroson, T.A. 1994,
Nature , 370, 441Scannapieco, E., Kawata, D., Brook, C.B., Schneider, R., Ferrara, A., Gibson, B.K. 2006,
ApJ ,653, 285Shang, Z., Brinks, E., Zheng, Z., et al.
ApJ , 504, 23Sommer-Larsen, J. & Dolgov, A. 2001,
ApJ , 551, 608Spergel, D.N., Bean, R., Dor´e, O., et al.
ApJS , 170, 377Tisserand, P., Le Guillou, L., Afonso, C, et al.
A&A , 469, 387Tully, R.B., Somerville, R.S., Trentham, N., Verheijen, M.A.W. 2002,
ApJ , 569, 573Verde, L., Oh, S. P., Jimenez, R. 2002,
MNRAS , 336, 541Witt, A.N. & Vijh, U.P. 2004, in: A.N. Witt, G.C. Clayton & B.T. Draine (eds.),
Astrophysicsof Dust , ASP Conference Series, Vol. 309, p. 115Yost, S.A., Bock, J.J., Kawada, M., Lange, A.E., Matsumoto, T., Uemizu, K., Watabe, T.,Wada, T 2000,
ApJ
ApJ AJ , 119, 1701Zheng, Z., Shang, Z., Su, H., et al. AJ , 117, 2757Zibetti, S., White, S.D.M. & Brinkmann, J. 2004, MNRAS , 347, 556Zibetti, S. & Ferguson, A.M.N. 2004,