ALMA – Atacama Large Millimeter Array
• ISM evolution at high z
• Arp 220 -- merging starburst galaxy
• CII probing high z galaxies
early universe galaxy formation : hierarchical growth of dark matter concentrations accretion of baryonic gas in DM centers/halos galaxy evolution morphological & dynamical evolution interstellar gas è stars if no accretion, star formation stops enhanced SF -- starbursts due to merging
present day galaxies : spirals – young stars + gas ellipticals – old stars, no gas
M51 M87 thin disk ‘spherical’
why such diverse properties ? cosmic evolution black hole accretion (AGN) & star formation (SF)
20x increase from z = 0 to 2 ! due to more gas (initial supply or accretion) or higher efficiency gas è stars , AGN starbursts – merging ? cosmic evolution black hole accretion (AGN) & star formation (SF) 20x increase from z = 0 to 2 !
star formation
AGN accretion x 3300 evolution due to more gas (initial supply or accretion) or higher efficiency gas è stars , AGN starbursts – merging ? 1.5 < z < 2.5!
1000! M /M = 5x1010! variation w/ stellar mass * evolution! w/ time
! 1! ] -1 ! ] -1 yr ! 100! [Gyr SFR [M sSFR
10! 10!
1010 1011 1012! 0 0.5 1 1.5 2!
stellar mass (M!) ! redshift !
Rodighiero ‘11, Sargent etal ‘12 need to measure ISM contents measuring ISM CO emission CO Rotational Levels CO (J=1-0) in Orion KL Nebula
collisional excitation by H2 1976ApJ...210L..39K
TB (K) well-calibrated correlation of M w/ CO 1-0 luminosity H2 CO w/ MH2 from Mvirial for GMCs A1-0 = -8 -1 13CO x 5.1 7x10 sec
estimating H2 masses – for resolved clouds Mvir correlated with LCO (= area TCO Δv) 1987ApJS...63..821S
CO L 5 Mvir (10 M!) How can an optically thick CO line measure mass ?? later at high z, need : robust and fast measure of ISM
CO – ok , but ...
CO/H2 conversion factor excitation dependence (often measure high J CO) slow even w/ ALMA (hours per gal.)
alternative, measure dust IR continuum + dust / gas ratio
ALMA cycle 0 & 2 projects
similar approach by advocated by Eales Santini etal ’14 using Herschel data emitted SED -- increasing Mdust
12 L=10 L¤
Mdust= 108 è 9 Scoville, 2011 Canary Is. 6x10 M¤ winter school lectures
• peak shifts to longer λ for increased τ (or dust mass)
• flux on long λ tail scales linearly with Mdust R-J tail is optically thin, è 2 2 FRJ = κν Tdust ν Mdust / (4πd )
Tdust = 20-25 K in Gal. SF
global mass-weighted Tdust doesn’t vary much calibrate : Lν / MISM = < κν Td MISM / Mdust > local galaxies Milky Way (Planck) SMGs local galaxies w/ total 850μm fluxes & ISM masses
20 1x10 erg/s/Hz/M¤
w/ less than factor 2 dispersion
Planck: Milky Way è 20 0.8x10 erg/s/Hz/M¤
β = 1.8 +- 0.1
850μm from Dale ‘05, Clements ’09, Dunne & Eales ‘09) z = 2 – 4 SMGs with CO (1-0) obs. :
è same calibration as local gal. & MW metallicity886 variations ? DRAINE ET AL. Vol. 663
within the dwarf galaxy [illumination of the outer envelope will suffer the same attenuation as radiation escaping the galaxy, from which (L)B was obtained]. Only one galaxy in Table 6, the dwarf starburst NGC 1705, has U0:5 > 0:2. The a k 0:03 mgrainsthatnormallydominatethe160m emission are expected to have temperatures T 17U 1/6 K 12(U/0:1) 1/6 K, with peak p at k hc/6kT 180(0 :2/U )1/6 m.¼ Therefore, for the seven galaxies in Table 6 with U0:5 < 0:2, even the MIPS 160 mband(withhalf-responsepointsat140and 174 m) is relatively insensitive to whatever dust may be present at and beyond 0:5 except for a small fraction that may be close to local stellar sources. MIPS photometry for these galaxies is dominated by the dust in the luminous regions, and the SED fit- ting is therefore fitting the parameters for this dust. If there is dust present beyond the half-mass radius, submillimeter observations will be required to detect it in all of the galaxies in Table 6 except NGC 1705. In the case of SBS 0335 052, where Hunt et al. (2005) report an Fig. 17.—Galaxies with and without SCUBA fluxes (circles and diamonds, À 7 respectively), including 28 galaxies with only lower limits on the gas mass. The extraordinarily low dust-to-gas mass ratio Mdust/MH 1 ; 10À ,a solid line shows eq. (13); dashed lines show factor of 2 variations around the solid factor of 5000 below equation (13), we note that this galaxy¼ has not line. Note that the upper bounds are all galaxies where Mdust has been determined, been detected at wavelengths between 100 mand1cm.The but either H i or CO has not been observed. Filled diamonds show Mdust/MH for dust models are poorly constrained, and it is possible that they regions where IR emission is detected (see 7.2). [See the electronic edition of the may have significantly underestimated the mass of cool dust. In Journal for a color version of this figure.]x ‘ draineaddition, 07 much of the H i mass in SBS 0335 052 may be in an extended envelope, where the metallicity is unknownÀ and where In the case of NGC 1705, UV absorption line studies show the dust temperatures would be very low. Deep submillimeter ob- that the H i envelope has AO 7:43 0:22 (Heckman et al. servations would be of value to better determine the dust content 2001): the oxygen abundance in¼ the HÆi envelope is a factor of of SBS 0335 052. 6 below the oxygen abundance determined from the H ii regions À (J. Moustakas et al. 2007, in preparation). For this oxygen abun- 7.3. Dust-to-Gas Mass Ratio and Galaxy Type dance, Mdust/MH for NGC 1705 is consistent with equation (13). The dependence of the dust-to-gas mass ratio on galaxy type The dust content of the envelope is often also uncertain. The is explored in Figure 18. Spiral galaxies (Sab, Sb, Sbc, Sc, Scd, stellar densities in the outer regions of these galaxies are low, and Sd) routinely have Mdust/MH 0:007 to within a factor of 2. the starlight will be dominated by light coming from the central We have dust mass estimates for three elliptical galaxies regions. Table 6 lists (L)B, the total stellar luminosity in the B (NGC 0855, NGC 3265, and NGC 4125) but lack information band. For each of these galaxies we give the estimated half-mass on the gas content of NGC 0855 and NGC 4125. Detection of radius 0:5 for the H i,andwehaveestimatedthestarlightin- 21 cm emission gives a lower bound on the gas mass in NGC 3265. tensity scale factor U at the H i half-mass radius, If the molecular and ionized gas mass in NGC 3265 is small compared to the H i mass, then this elliptical galaxy has a dust- 1 L U B 19 to-gas ratio consistent with solar metallicity. 0:5 13 3 ðÞ2 ; ¼ 1:982 ; 10À ergs cmÀ 4cD0:5 ð Þ The three S0 galaxies in our samples have dust-to-gas ratios ðÞðÞ that are consistent with the range seen for normal spirals: NGC 13 3 where 1:982 ; 10À ergs cmÀ is (u)B for the local ISRF 3773 has Mdust/MH < 0:01, and NGC 1482 and NGC 5866 have (Mathis et al. 1983). Equation (19) neglects internal extinction Mdust/MH 0:005.
TABLE 6 Parameters for Selected Dwarf Galaxies
D (L )B IR f(H i) 0.5(H i) Galaxy Morphological Type (Mpc) MB (L ) (arcmin) < IR (arcmin) U0.5(H i)
DDO 053...... Im 3.56 13.44 2.1 ; 107 ... 0.33a 1 0.11 À b b NGC 6822...... Im 0.49 14.03 3.7 ; 107 17 0.33 33 0.034 À c Holmberg I...... Im 3.84 15.12 1.0 ; 108 ... 0.29 2 0.11 À d e NGC 1705...... Im 5.10 15.76 1.8 ; 108 0.33 0.15 0.69 4.0 À f a NGC 2915...... Im 2.70 16.41 3.3 ; 108 0.5 0.028 4.75 0.13 À c Holmberg II...... Im 3.39 16.69 4.2 ; 108 ... 0.20 40.15 À c IC 2574 ...... Sm 4.02 17.38 8.0 ; 108 ... 0.19 60.089 À g NGC 4236...... Sdm 4.45 17.94 1.3 ; 109 4.25 0.5 4.2 0.18 À a Becker et al. (1988). b Estimated from H i profile in Cannon et al. (2006b). c Walter et al. (2007). d Cannon et al. (2006a). e Meurer et al. (1998). f Estimated from H i data in Meurer (1997). g Estimated from Rots (1980) map, assuming FWHM 100 Gaussian beam. ¼ for ALMA Bands 3 - 7 predict :
3 σ in ~2 min for M = 1x1010 ISM 20x faster than CO ! ALMA Cycle 2 – observations --180 galaxies w/ 2 min !!! per gal. w/ Sheth, Aussel, Vanden Bout, Capak, Bongiorno, Casey, Murchikova, Koda, Pope, Toft, Ivison, Sanders, Manohar, Lee z = 601.15 @ z = 1 z = 60 2.2 @ z = 2 z = 60 4.8 @ z = 5
band 5 (350 GHz) 5 (350 GHz) 6 (230 GHz)
stack obs for each z in cells of M* and SFR 2 min detection rates -- 3 redshift ranges :
flux mass z = 2.2 images : ISM masses vs sSFR = SFR / M*
10 MISM (10 M¤) ISM masses vs sSFR = SFR / M* 10 MISM (10 M¤)
peak MISM at z ~ 2 , decreases ~5x from z = 2 to 1 decrease from z = 2 to 5 ISM mass fraction :
MISM / (MISM + Mstellar) ISM mass fraction :
MISM / (MISM + Mstellar) ISM masses increase above the main sequence !! è increase in SFRs above the MS due to larger ISM masses analytic fits : M ISM frac = ISM M + M ISM stellar " % −0.30 Mstellar 0.86 −1.24 = 0.006 $ 11 ' SFR (1+ z) # 10 Msun &
" % 0.9±0.07 MISM −1 SFR = 110 ± 7 $ 10 ' Msunyr # 10 Msun &
€ a single ‘linear’ SF law gas depletion times 2100 R. Genzel et al.
Fig. 3 (and also Fig. 2) we did not attempt to assign individual errors Here and elsewhere, we computed the dynamical time-scale from (unlike K98a), since in our opinion essentially all uncertainties are the ratio of the radius to the circular velocity vc.Forthez > 1 SFGs systematic in nature and apply to all data equally. This slope is in and SMGs we took R R1/2 and applied a pressure correction to = 2 2 1/2 very good agreement with the spatially resolved relation for nearby the inclination-corrected rotation velocity vrot,vc (vrot 2σ ) , spirals in Bigiel et al. (2008, green/orange/red-shaded region in the where σ is the local 1D-velocity dispersion in= the galaxy.+ This left-hand panel of Fig. 3). The new data do not indicate a signifi- relation is applicable to rotation-dominated, as well as pressure- 2 2 cant steepening of the slope at surface densities of >10 M pc− , dominated galaxies. The slope we find is close to that of K98a, neither at z 0noratz 1. Within the limited statistics⊙ of the who find a slope between 0.9 and 1. High-z SFGs have somewhat ∼ ≥ currently available data, we do not find a break in the slope near higher "star formation than low-z galaxies (by 0.71 0.21 dex) but the 2 2 ± 10 M pc− , as proposed by Krumholz et al. (2009). The slope of difference is probably only marginally significant. A fit with unity 1.33 found⊙ by Krumholz et al. (2009) in the high-density limit is slope yields a star formation efficiency per dynamical time of 0.019 marginally larger. A steeper slope in this regime (1.28 to 1.4) was ( 0.008). This is in agreement with 0.01, the value found by K98a suggested earlier by the K98a starburst sample, but that analysis when± corrected to a Chabrier IMF. included some mergers (see below) and the combined scatter of both data sets suggests a 1σ uncertainty of 0.15, which makes the 4.2 KS relation for luminous mergers difference in slope of 0.1–0.23 only marginally∼ significant. Low- and high-z SFGs overlap completely, again with the obvious Fig. 4 summarizes our analysis of the luminous mergers at both low exception of EGS12012083 and BX389. The data in Fig. 3 suggest and high z. The left-hand panel shows the case of applying the best that the KS relation in normal SFGs does not vary with redshift, in single common conversion factor determined from the observations Downloaded from agreement with the conclusions of Boucheetal.(2007)andDaddi´ (αmerger 1, Section 2.6), such that mergers and SFGs now have et al. (2010a,b). conversion∼ factors that differ by a factor of 3.2. The slope of the In the right-hand panel of Fig. 3, we analyse the data with the merger relation (1.1 0.2) is consistent with that of the SFGs ‘Elmegreen–Silk’ relation (see also K98a), which relates SFR sur- (1.17). Again low- and± high-z mergers lie plausibly on the same face density to the ratio of gas surface density and global galaxy relation. Independent of whether the merger slope is fit or forced to
dynamical time-scale. There is a reasonably good correlation as well be the same as that of the SFGs, the difference in SFR at a given http://mnras.oxfordjournals.org/ with a slope of slightly less than unity (0.84 0.09). The scatter in gas surface density between the two branches is 1.0 ( 0.2) dex ± this relation (0.44 dex) is larger than in the surface density relation, (see also Bothwell et al. 2010). ∼ ± which may in part be attributable to the larger total uncertainties As we have argued in Section 2.6, a Galactic conversion factor for in "molgas/τdyn, which we estimate to be 0.32 dex (74 per cent). all luminous low- and high-z mergers is almost certainly excluded very different than previous work from CO ± No. 1, 2010 SF LAWS FOR DISKS AND STARBURSTS AT LOW AND HIGH REDSHIFTS L119 4 4 y=1.17*x-3.48 y=1.17*x -2.44 (0.12) y=+1.27 (0.075)*x - 3.63 (0.21), stdev: 0.32 3 SFGs 3 by guest on May 11, 2015 ) ) -2 -2 2 2 kpc kpc -1 -1 1 α=1 merger 1 α=3.2 merger yr yr α=3.2 SFG α=3.2 SFG sun sun 0 0 (M (M
-1 -1 star form star star form star Σ Σ -2 -2 log ( log ( ( log -3 -3
typical uncertainty typical uncertainty -4 -4 0 1 2 3 4 5 0 1 2 3 4 5
Figure 1. Comparison of molecular gas masses and total IR bolometric luminosities: BzK galaxies (red filled circles; D10), z 0.5diskgalaxies(redfilledtriangles;-2 -2 F. Salmi et al. 2010, in preparation), z 1–2.3 normal galaxies (Tacconi et al. 2010;browncrosses),SMGs(blueemptysquares;Greveetal.log (Σ (M∼ pc ) ) 2005;Frayeretal.2008; log (Σ (M pc ) ) = mol gas sun mol gas sun DaddiDaddi et al. 2009a etal, 2009b 2010), QSOs (green triangles; see Riechers et al. 2006), local ULIRGs (black crosses; Solomon et al. 1997), and local spirals (black filled squares, Leroy et al. 2009;blackfilledtriangles,Wilsonetal.2009). The twoFigure nearby 4.starburstsMolecular M82 Kennicutt–Schmidt and the nucleus of surface NGC 253 density are also relation shown for (dataluminous fromz Weiß0and et al.z 20011–3;.5 mergers (z 0LIRGs/ULIRGs:magentasquares,z 1 Genzel etal 2010 ∼ ∼ ∼ ≥ Houghton et al. 1997;Kanedaetal.2009). The solid line (Equation (SMGs:1), slope red of squares). 1.31 in the The left left-hand panel) is panel a fit shows to local their spirals location and BzK in the galaxies KS plane and along the dotted with the line SFGs is (at all z,opengreycircles)fromFig.3iftheaprioribest the same relation shifted in normalization by 1.1 dex. The dashed line in the left panel is a possible double power-law fit to spirals and BzK galaxies. For guidance, 1 conversion factors for SFGs (α αG) and mergers (α αG/3.2) are chosen. The right-hand panel shows the same plot for the choice of a universal conversion two vertical lines indicate SFR 2and200M yr− in the right panel. = = = ⊙ factor of α αG for all galaxies in the data base. This was the choice in the K98a paper but leads to a significant overestimate of gas fractions in almost all both used different CO conversion= factors for SB and MS (A color version of this figure is available in the online journal.) major mergers. The fits assign equal weight to all data points and uncertainties in brackets are 3σ formal fit errors. The crosses in the lower right denote the typical total (statistical systematic) 1σ uncertainty. + These allow us to study more typical high-redshift galaxies Figure 1 is equivalent to Figure 13 in D10, after replacing with SFRs much larger than those of local spirals but less L′ with MH2.TherightpanelshowstheratioofC LIR to MH2 C CO ⃝ 2010 The Authors. Journal compilation ⃝ 2010 RAS, MNRAS 407, 2091–2108 extreme than those of distant SMGs. The sample of six CO- plotted versus LIR.Theimpliedgasconsumptiontimescales detected z 1.5normal(BzK-selected)galaxiesispresented (τgas MH2/SFR; right panel of Figure 1) are 0.3–0.8 Gyr for = = in D10. We also use CO detections of three near-IR selected the BzK galaxies,10 about 2–3 times that for spirals, and over disk galaxies at z 0.5. A detailed discussion of the z 0.5 1orderofmagnitudesmallerforlocal(U)LIRGsanddistant = = data set will be presented elsewhere (F. Salmi et al. 2010, in SMGs. In a simple picture, this finding can be interpreted in preparation). For comparison, we also show measurements for terms of two major SF modes: a long-lasting mode appropriate normal CO-detected galaxies at z 1–2.3 from Tacconi et al. for disks, that holds for both local spirals and distant BzK = (2010), although we do not use these in our analysis. These new galaxies, and a rapid starburst mode appropriate for ULIRGs, observations are placed in context with the literature data for local starbursts like M82 or the nucleus of NGC 253, and distant ULIRGs, SMGs, and local samples of disk galaxies. SMGs/QSOs. For the disk galaxies we formally fit In order to investigate the location of these populations of normal high-z galaxies in the gas mass versus SFR plane, either log LIR/L 1.31 log MH2/M 2.09, (1) for the integrated properties or for the surface densities, a crucial ⊙ = × ⊙ − ingredient is, again, the αCO conversion factor. Comparing with an error on the slope of 0.09 and a scatter of 0.22 dex. the dynamical and stellar mass estimates, D10 derive a high Combining ULIRGs and SMGs we find that they define a trend 1 2 1 8 αCO 3.6 0.8 M (K km s− pc )− for the BzK galaxies , with a similar slope, but with about 10 times higher LIR at fixed = ± ⊙ quite similar to that for local spirals (αCO 4.6). This is not MH2. unexpected, given the evidence that the z 1=.5near-IRselected Asimilarpictureappliestothesurfacedensities(Figure2). galaxies appear to be high-redshift analogs∼ of local disks with We here use the original K98 measurements for local spirals enhanced gas content (see, e.g., discussions in Daddi et al. 2008, and (U)LIRGs, but apply our choice of αCO and a Chabrier 2010;Dannerbaueretal.2009;Tacconietal.2010,andlaterin (2003)IMF.ForconsistencywiththeK98relation,wemeasure this Letter). In the following, we adopt this value of αCO 3.6 Σgas adding H i and H2 for spirals, and H2 for IR-luminous for the z 0.5–2.5 normal galaxies9 and the “consensus”= galaxies, in Figures 2 and 3.Theresultswouldnotchangeifwe = value for the other populations (αCO 4.6forlocalspirals, had used H2 only for all galaxies. Values for SMGs are taken = αCO 0.8forlocal(U)LIRGsanddistantSMGs/QSOs), and from Boucheetal.(´ 2007). For the BzK galaxies, we derive explore= the consequences for the relation between gas masses gas and SFR surface densities using the UV rest-frame (SFR) and IR luminosities/SFRs. sizes. These are consistent with the CO sizes (D10) but are
10 1 10 We apply a conversion SFR[M yr− ] 10− LIR/L , treating the 8 ⊙ ⊙ This conversion factor refers to the total gas mass, including H i,H2,and two quantities as equivalent. In the case that= a significant× active galactic helium, in their proportion within the half-light radius. nucleus (AGN) contribution affects LIR (e.g., for the QSOs), SFRs would be 9 Tacconi et al. (2010)assumeasimilarfactor. correspondingly lower. our work è single, linear SF law at z = 1 to 6 and on MS and above MS " % 0.9±0.07 MISM −1 SFR = 110 ± 7 $ 10 ' Msunyr # 10 Msun & M ISM 8 ⇒ τ ISM→ stars = ≈ 10 yrs (10x faster than z = 0) SFR huge accretion rates € replace entire ISM w/i 108 yrs why is SF more rapid at z > 1 ??
Note – do not fit for Td -- Lum.- vs mass-weigthed Arp 220 @ 77 Mpc 2μm 12 1 arcsec è 361 pc LIR = 2.5x10 L¤
West East
AV ~ 2000 mag towards nuclei !! L = 2.5x1012 Arp 220
Cycle 0 ALMA --
Band 7 (0.5’’) : HCN (4-3) , CS (7-6) , H26α Band 9 (0.25’’) : HCN (8-7) Cycle 1 è Band 7 w/ 0.2’’ w/ Sheth, Walter, Manohar, Zschaechner, Yun, Koda, Walter, Sanders, Murchikova, Thompson, Robertson, Genzel, Hernquist, Tacconi, Brown, Narayanan, Hayward, Barnes, Kartaltepe, Davies, van der Werf, Fomalont ApJ Jan 2015
also -- Wilson et al 2014 Band 9 0.2 – 0.36 “ res
è τ dust ~ 5.7, 1.7 at λ = 434 μm Tdust = 80 – 200 K
Arp 220 dust continuum
348 GHz ~ 850 μm 698 GHz ~ 450 μm 0.5’’ res. 0.25’’ res.
360 pc dust continuum è ISM mass Arp 220 and NGC 6240 13
Table 3 Gaussian Source Sizes from Table 2
Beam Source Deconvolved ——————————– ———————————————————— —————————- Source major minor PA major minor PA major minor TB 00 00 00 00 00 00 K
Band 7
Arp 220 W continuum 0.60 0.42 -32.0 0.66 0.01 0.50 0.01 -29 10.280.2733.9 Arp 220 E continuum 0.60 0.42 -32.0 0.71 ± 0.02 0.50 ± 0.01 -16 ± 10.370.2711.1 Arp 220 W continuum 0.52 0.39 -27.2 0.63 ± 0.01 0.46 ± 0.02 -28 ± 10.360.2434.9 Arp 220 E continuum 0.52 0.39 -27.2 0.65 ± 0.01 0.51 ± 0.01 -17 ± 30.380.329.6 NGC 6240 S continuum 0.55 0.46 65.6 0.74 ± 0.04 0.60 ± 0.03 -152 ± 13 0.50 0.39 0.4 NGC 6240 N continuum 0.55 0.46 65.6 1.22 ± 0.11 0.70 ± 0.06 -150± 81.090.530.1 NGC 6240 S continuum 0.53 0.44 64.7 0.72 ± 0.02 0.59 ± 0.02 -155 ± 20.490.400.3 NGC 6240 N continuum 0.53 0.44 64.7 1.19 ± 0.01 0.72 ± 0.02 -149 ± 11.070.570.03 ± ± ± Band 9
Arp 220 W continuum 0.32 0.28 -38.6 0.39 0.01 0.34 0.01 -42 10.230.19148.9 Arp 220 E continuum 0.32 0.28 -38.6 0.44 ± 0.01 0.37 ± 0.02 -139± 10.300.2447.2 NGC 6240 continuum 0.27 0.24 29.7 0.86 ± 0.02 0.39 ± 0.02 -173 ± 10.820.300.2 ± ± ± Lines
Arp220 W CS (7 - 6) 0.60 0.42 -32.0 0.78 0.02 0.60 0.02 -31 30.490.4310.1 Arp220 E CS (7 - 6) 0.60 0.42 -32.0 0.72 ± 0.02 0.54 ± 0.02 -35 ± 20.400.357.5 Arp220 W HCN (4 - 3) 0.52 0.39 -27.2 0.77 ± 0.01 0.57 ± 0.01 -24 ± 10.570.4139.3 Arp220 E HCN (4 - 3) 0.52 0.39 -27.2 0.78 ± 0.01 0.59 ± 0.01 20 ± 20.580.4521.5 NGC 6240 CS (7 - 6) 0.55 0.46 65.6± ....a ± ....a ....±a ....a ....a ....a NGC 6240 HCN (4 - 3) 0.53 0.44 64.7 1.26 0.02 0.74 0.02 -178 31.140.601.5 ± ± ±
Note. — Sizes and major axis PA estimates obtained from 2d gaussian fits listed in Table 2. The uncertainties in the parameters were the formal errors from the multi-component Gaussian fitting. a No convergent fit obtained.
Table 4 masses of ISM from dustISM emission Masses from Dust Continuum
a d b c c Source ⌫obs Flux Td RJ Mass diameter Radius < ⌃gas >
9 2 2 GHz mJy K 10 M 00 pc M pc cm Arp 220 total 347.6 490 100 0.917 5.9 5 24 Arp 220 East 347.6 161 100 0.917 1.9 . 0.38 . 69 & 1.3 10 & 6.0 10 Arp 220 West 347.6 342 100 0.917 4.2 0.36 65 3.1 ⇥ 105 1.5 ⇥ 1025 . . & ⇥ & ⇥ 4 23 NGC 6240 693.5 126 25 0.468 1.6 0.8 190 & 1.4 10 & 7.0 10 AV ~ 2000 mag towards nuclei !! ⇥ ⇥
Note.—ISMmassesderivedfromtheRJcontinuumfluxusingEquation4withdistancesof74and103 Mpc for Arp 220 and NGC 6240. a Adopted dust temperature used for mass calculationlines as... discussed in text. b Angular diameter estimate from deconvolved major axis of dust emission Gaussian fit (Table 3) using the Band 7 measurements for Arp 220 (since the Band 9 dust emission is optically thick) and Band 9 for NGC 6240 (since the SNR is better in Band 9). c Mean gas surface density and H2 column density assuming the gas mass is distributed uniformly over a disk with the radius given in column 8. d Rayleigh-Jeans correction factor from Equation 4. This 2-level formalism can be applied to the observed HCN (4 - 3), the peak line brightness temperatures are HCN (4 - 3) emission in Arp 220 to constrain the gas den- 20 and 40 K (Table 3). Assuming the line is optically sity in the nuclear sources. In the nuclei at radii less than thick⇠ (as in Galactic sources), these brightness tempera- 65 pc we adopt a gas kinetic temperature T = 100 tures imply excitation temperatures T 20 and 40 K. ⇠ k x ⇠ K, based on the dust blackbody-limit temperature and In the evaluation below, we adopt Tx 40 K and there- 4 3 ' the expectation that at densities above 10 cm the gas fore the gas density will be constrained to that giving should be thermally coupled to the dust. This high ki- Tx/Tk 0.4 in Equation 13. netic temperature is consistent with the brightness tem- For HCN,' hB/k =2.13 K and for the J = 4 -3 tran- 3 1 peratures observed for CO (Sakamoto et al. 2009). For sition, T0 = 17.0 K with A4 3 =2.1 10 sec . ⇥ Arp 220 E+W!
LSB! USB!
Rest frame frequency (GHz)! Arp 220! HCN (4-3)! < V >!
CS (7-6)! < V >!
è counter-rotating disks (as in Sakamoto etal ‘98) 1414 Scoville et al. Scoville et al. 1414 Scoville et al. Scoville et al.
Table 3 Table 3 Gaussian SourceTable Sizes 3 from Table 2 Gaussian SourceTable Sizes 3 from Table 2 14Gaussian14 Source Scoville Sizeset from al. Table 2 Gaussian Source Scoville Sizeset from al. Table 2 Beam SourceBeam Deconvolved Source Deconvolved ——————————–Beam ———————————————————— Source——————————–Beam —————————- ———————————————————— Deconvolved Source —————————- Deconvolved Source major minor PA major minor PA major minor T ——————————–Source ————————————————————Table 3 ——————————– major minor PA —————————-————————————————————Table major 3 B minor PA major —————————- minor TB K Source major00 minor 00Gaussian PA Source Source major 00 Sizes from Table minordouble 00 major00 2 minorGaussian 00Gaussian PA PA Sourcefits major 00 major 00 Sizes minor 00 from T TableB minor 00 2 PA major 00 minor 00 TKB 00 00 00 0000 00 00 00 00 K 00 00 00 K BeamBand Source 7 Beam Deconvolved Band Source 7 Deconvolved ——————————– ————————————————————Band——————————– 7 —————————- ————————————————————Band 7 —————————- ArpSource 220 W continuum 0.60major 0.42minor PA-32.0Source 0.66 major0.0022 0.50 minor major0.0016 minor -29.0 PA PA0.57 major 0.28 major minor 0.27 T 33.9 minor PA major minor TB Arp 220 W± continuum± 0.60 0.42± -32.0 0.66 0.0022 0.50B 0.0016 -29.0 0.57 0.28 0.27 33.9 Arp 220 EW continuum 0.6000 0.42 00 -32.0Arp 220 0.710.66 WE 000.00510.0022 continuum 0.50 00 0.600.00360.001600 0.42 00 -15.7-29.0 -32.0 1.020.57 0.660.71 0.370.28 00± 000.00220.0051 0.27 00 0.50 11.133.9K ± 000.00160.0036 -29.0-15.7 ± 0.571.02 0.280.37 00 0.27 00 33.911.1K Arp 220 W continuum 0.52 0.39 -27.2 0.63 ± 0.0041 0.46 ± 0.0029 -28.1 ± 0.97 0.36± 0.24 34.9± ± Arp 220 E continuum 0.60 0.42 -32.0Arp 220 0.71 EW± 0.0051 continuum 0.50 ± 0.520.600.0036 0.420.39 -15.7 ± -27.2-32.01.02 0.630.71 0.370.00510.0041 0.27 0.500.46 11.1 0.00360.0029 -15.7-28.1 1.020.97 0.370.36 0.270.24 11.134.9 Arp 220 EW continuum 0.52 0.39 -27.2Arp 220 0.650.63 WE± 0.00970.0041 continuum 0.510.46 ± 0.520.00760.0029 0.39 -17.4-28.1 ± -27.23.060.97 0.630.65 0.380.36± 0.00410.0097 0.320.24 0.460.51 34.9 9.6± 0.00290.0076 -28.1-17.4 ± 0.973.06 0.360.38 0.240.32 34.9 9.6 NGC 6240 S continuum 0.55 0.46 65.6 0.74 ± 0.0405 0.60Band± 0.0329 7 -152.3 ±13.07 0.50± 0.39Band 0.4± 7 ± Arp 220 E continuum 0.52 0.39 -27.2ArpNGC 220 6240 0.65 E S±band0.0097 continuum 7 0.51 continuum± 0.520.550.0076 0.390.46 -17.4 ± -27.2 65.63.06 0.650.74 0.380.00970.0405 0.32 0.510.60 9.6 0.00760.0329 -152.3 -17.4 13.073.06 0.380.50 0.320.39 9.60.4 NGC 6240 NS continuum 0.55 0.46NGC 65.6 6240 1.220.74 SN± 0.11660.0405 continuum 0.700.60 ± 0.550.06680.0329 0.46 -152.3 -150.4 ±13.07 65.68.08 1.220.74 1.090.50± 0.04050.1166 0.530.39 0.600.70 0.10.4± 0.03290.0668 -152.3 -150.4±±13.078.08 1.090.50 0.530.39 0.10.4 Arp 220 W continuum 0.60 0.42 -32.0Arp 220 0.66 W± 0.0022 continuum 0.50 ± 0.600.0016 0.42 -29.0 ± -32.00.57 0.66 0.28± 0.0022 0.27 0.50 33.9± 0.0016 -29.0±± 0.57 0.28 0.27 33.9 NGC 6240 SN continuum 0.530.55 0.440.46NGC 64.765.6 6240 0.721.22 NS± 0.00490.1166 continuum 0.590.70 ± 0.550.530.00400.0668 0.460.44 -154.7-150.4±± 65.664.71.708.08 1.220.72 0.491.09± 0.11660.0049 0.400.53 0.700.59 0.30.1± 0.06680.0040 -150.4-154.7 ± 8.081.70 1.090.49 0.530.40 0.10.3 Arp 220 E continuum 0.60 0.42 -32.0Arp 220 0.71 E± 0.0051 continuum 0.50 ± 0.600.0036 0.42 -15.7 ± -32.01.02 0.71 0.37± 0.0051 0.27 0.50 11.1± 0.0036 -15.7 ± 1.02 0.37 0.27 11.1 NGC 6240 NS continuum 0.53 0.44NGC 64.7 6240 1.190.72 SN± 0.01390.0049 continuum 0.720.59 ± 0.530.00840.0040 0.44 -149.3-154.7 ± 64.71.101.70 1.190.72 1.070.49± 0.00490.0139 0.570.40 0.590.72 0.00.3± 0.00400.0084 -154.7-149.3 ± 1.701.10 0.491.07 0.400.57 0.30.0 Arp 220 W continuum 0.52 0.39 -27.2Arp 220 0.63 W± 0.0041 continuum 0.46 ± 0.520.0029 0.39 -28.1 ± -27.20.97 0.63 0.36± 0.0041 0.24 0.46 34.9± 0.0029 -28.1 ± 0.97 0.36 0.24 34.9 NGC 6240 N continuum 0.53 0.44ArpNGC 64.7 220 6240 1.19 E N± 0.0139 continuum 0.72 ± 0.530.520.0084 0.440.39 -149.3 ± -27.2 64.71.10 0.651.19 1.07± 0.00970.0139 0.57 0.510.72 0.0± 0.00760.0084 -149.3 -17.4 ± 3.061.10 0.381.07 0.320.57 9.60.0 Arp 220 E continuum 0.52 0.39 -27.2 0.65 ± 0.0097 0.51Band± 0.0076 9 -17.4 ± 3.06 0.38± 0.32Band 9.6± 9 ± NGC 6240 S± continuum± 0.55 0.46± 65.6 0.74 ± 0.0405 0.60 ± 0.0329 -152.3 ±13.07 0.50 0.39 0.4 NGC 6240 S continuum 0.55 0.46 65.6 0.74 0.0405 0.60 0.0329 -152.3 13.07 0.50 0.39Band 0.4 9 NGC 6240 N continuum 0.55 0.46NGC 65.6 6240 1.22 N±band0.1166 continuum 9 0.70 continuumBand± 0.550.0668 9 0.46 -150.4 ± 65.68.08 1.22 1.09± 0.1166 0.53 0.70 0.1± 0.0668 -150.4± 8.08 1.09 0.53 0.1 Arp 220 W continuum 0.32 0.28 -38.6Arp 220 0.39 W0.0009 continuum 0.34 0.320.0008 0.28 -41.6 -38.60.79 0.39 0.23± 0.0009 0.19 0.34 148.9± 0.0008 -41.6 ± 0.79 0.23 0.19 148.9 NGC 6240 S continuum 0.53 0.44NGC 64.7 6240 0.72 S± 0.0049 continuum 0.59 ± 0.530.0040 0.44 -154.7 ± 64.71.70 0.72 0.49± 0.0049 0.40 0.59 0.3± 0.0040 -154.7 ± 1.70 0.49 0.40 0.3 Arp 220 EW continuum 0.32 0.28 -38.6Arp 220 0.440.39 WE0.00180.0009 continuum 0.370.34 0.320.00160.0008 0.28 -139.3 -41.6 -38.61.230.79 0.390.44 0.300.23± 0.00090.0018 0.240.19 0.340.37 148.9 47.2± 0.00080.0016 -139.3 -41.6 ± 0.791.23 0.230.30 0.190.24 148.9 47.2 NGC 6240 N continuum 0.53 0.44NGC 64.7 6240 1.19 N± 0.0139 continuum 0.72 ± 0.530.0084 0.44 -149.3 ± 64.71.10 1.19 1.07± 0.0139 0.57 0.72 0.0± 0.0084 -149.3 ± 1.10 1.07 0.57 0.0 NGCArp 220 6240 E continuum 0.270.32 0.240.28 -38.6ArpNGC 29.7 220 6240 0.860.44 E± 0.00540.0018 continuum 0.390.37 ± 0.270.320.00230.0016 0.240.28 -173.1-139.3 ± -38.6 29.70.341.23 0.440.86 0.820.30± 0.00180.0054 0.300.24 0.370.39 47.2 0.2± 0.00160.0023 -139.3-173.1 ± 1.230.34 0.300.82 0.240.30 47.2 0.2 ± ± ± ± ± ± NGC 6240 continuum 0.27 0.24NGC 29.7 6240 0.86 0.0054 continuum 0.39 0.270.0023 0.24 -173.1 29.70.34 0.86 0.820.0054 0.30 0.39Band 0.2 0.0023 9 -173.1 0.34 0.82 0.30 0.2 ± bandBand Lines7± lines 9 ± ± Lines± ± Lines Arp 220 W continuum 0.32 0.28 -38.6Arp 220 0.39 W0.0009 continuum 0.34 Lines 0.320.0008 0.28 -41.6 -38.60.79 0.39 0.230.0009 0.19 0.34 148.9 0.0008 -41.6 0.79 0.23 0.19 148.9 Arp220 W CS (7 - 6) 0.60 0.42 -32.0Arp220 W0.78 0.005 CS (7 - 6) 0.60 0.600.005 0.42 -31.4 -32.03.00 0.78 0.49± 0.005 0.43 0.60 10.1± 0.005 -31.4 ± 3.00 0.49 0.43 10.1 Arp 220 E continuum 0.32 0.28 -38.6Arp 220 0.44 E±±0.0018 continuum 0.37 ±± 0.320.0016 0.28 -139.3 ± -38.61.23 0.44 0.30±0.0018 0.24 0.37 47.2±0.0016 -139.3 ± 1.23 0.30 0.24 47.2 Arp220 E CS (7 - 6) 0.60 0.42 -32.0Arp220 WE0.72 0.004 CS (7 - 6) 0.54 0.600.003 0.42 -34.9 -32.02.00 0.720.78 0.40± 0.0050.004 0.35 0.600.54 7.5± 0.0050.003 -31.4-34.9 ± 3.002.00 0.490.40 0.430.35 10.1 7.5 NGCArp220 6240 W continuumCS (7 - 6) 0.270.60 0.240.42 -32.0NGC 29.7 6240 0.86 0.78±±0.0054 continuum0.005 0.39 0.60±± 0.270.00230.005 0.24 -173.1 -31.4 ± 29.70.343.00 0.86 0.820.490.0054 0.300.43 0.39 10.1 0.2 0.0023 -173.1 0.34 0.82 0.30 0.2 Arp220 W HCN (4 - 3) 0.52 0.39 -27.2Arp220 EW0.77 ± 0.004 CSHCN (7 (4 - 6) - 0.57 3)± 0.520.600.003 0.390.42 -23.5 ± -27.2-32.01.00 0.770.72 0.57±± 0.004 0.41 0.540.57 39.3±± 0.003 -34.9-23.5 ± 2.001.00 0.400.57 0.350.41 39.3 7.5 Arp220 E CS (7 - 6) 0.60 0.42 -32.0 0.72±± 0.004 0.54±± 0.003 -34.9 ± 2.00 0.40± 0.35 7.5± ± Arp220 EW HCN (4 - 3) 0.52 0.39 -27.2Arp220 WE0.78 0.77 ± 0.004 HCN (4 - 0.590.57 3)± 0.520.003 0.39 -23.5 20.0 ± -27.21.501.00 0.770.78 0.580.570.004 0.450.41 0.570.59 21.539.3 0.003 -23.5 20.0 1.001.50 0.570.58 0.410.45 39.321.5 ± a Lines± a ± a a± aa Linesa± a ± a a a a NGCArp220 6240 E CSHCN (7 (4 - 6) - 3) 0.550.52 0.460.39 -27.2Arp220NGC 65.6 6240 E 0.78 ± HCNCS0.004 .... (7 (4 - 6) - 0.59 3)± 0.520.550.003.... 0.390.46 20.0 ± -27.2 65.61.50.... 0.78 0.58.... 0.004 .... 0.45.... 0.59 21.5.... 0.003.... 20.0 1.50.... 0.58.... 0.45.... 21.5.... ± a ± a ± a a a a NGC 6240 HCNCS (7 (4 - 6) - 3) 0.530.55 0.440.46Arp220NGC 64.765.6 6240 W1.26 ± 0.009 CSHCN .... (7a (4 - 6) - 0.74 3)± 0.600.530.550.008....a 0.420.440.46 -178.3 ± -32.0 64.765.63.00....a 0.781.26 1.14....a 0.0050.009 .... 0.60....a 0.600.74.... 1.5a 0.0050.008.... -178.3 -31.4 3.00.... 0.491.14.... 0.430.60.... 10.1.... 1.5 Arp220 W CS (7 - 6) 0.60 0.42 -32.0NGC 6240 0.78 ± 0.005 HCN (4 - 0.60 3)± 0.530.005 0.44 -31.4 ± 64.73.00 1.26 0.49± 0.009 0.43 0.74 10.1± 0.008 -178.3 ± 3.00 1.14 0.60 1.5 Arp220NGC 6240 E CSHCN (7 (4 - 6) - 3) 0.600.53 0.420.44 -32.0Arp220 64.7 E0.72 1.26 ± 0.004 CS0.009Dust (7 - 6) 0.540.74 : major± 0.600.0030.008 0.42axis -178.3 -34.9 radius± -32.02.003.00 ~ 0.72 0.401.14 0.12± 0.004’’ 0.350.60 è 0.54 7.51.540± 0.003pc -34.9 ± 2.00 0.40 0.35 7.5 Arp220 W± HCN (4 - 3)± 0.52 0.39± -27.2 0.77 ± 0.004 0.57 ± 0.003 -23.5 ± 1.00 0.57 0.41 39.3 Arp220 W HCN (4 - 3) 0.52 0.39 -27.2 0.77 0.004 0.57 0.003 -23.5 1.00 0.57± 0.41 39.3± ± Arp220Note.—SizesandmajoraxisPAestimatesobtainedfrom2dgaussianfitslistedinTable2.Theuncertaintiesintheparameterswerethe E HCN (4 - 3) 0.52 0.39 -27.2Arp220Note.—SizesandmajoraxisPAestimatesobtainedfrom2dgaussianfitslistedinTable2.Theuncertaintiesintheparameterswerethe E0.78 ± 0.004 HCN (4 - 0.59 3)± 0.52T0.003B ~ 148 0.39 20.0 &± -27.2471.50 K 0.78(450 0.58 0.004μ 0.45m) 0.59 21.5 0.003 20.0 1.50 0.58 0.45 21.5 ± a ± a ± a a± aa a± a ± a a a a formalNGCNote errors 6240.—SizesandmajoraxisPAestimatesobtainedfrom2dgaussianfitslistedinTable2.Theuncertaintiesintheparameterswerethe from CS the (7 multi-component - 6) 0.55 Gaussian 0.46 fitting.formalNGC 65.6Note errors 6240.—SizesandmajoraxisPAestimatesobtainedfrom2dgaussianfitslistedinTable2.Theuncertaintiesintheparameterswerethe from CS the .... (7 multi-component - 6) 0.55.... Gaussian 0.46 fitting. 65.6...... a a formalNGCNo convergent errors 6240 from fit HCN the obtained. multi-component (4 - 3) 0.53 Gaussian 0.44 fitting.formalNGC 64.7No convergent errors 6240 1.26 from fit0.009 HCN theHCN obtained. multi-component (4 - 0.74 3): major 0.530.008 Gaussian 0.44axis -178.3 fitting.radius 64.73.00 ~ 1.26 1.14 0.250.009’’ 0.60 è 0.74 1.5 900.008 pc -178.3 3.00 1.14 0.60 1.5 a ± ± ± ± ± ± a No convergent fit obtained. No convergent fit obtained. TB ~ 40 K Note.—SizesandmajoraxisPAestimatesobtainedfrom2dgaussianfitslistedinTable2.TheuncertaintiesintheparametersweretheNote.—SizesandmajoraxisPAestimatesobtainedfrom2dgaussianfitslistedinTable2.TheuncertaintiesintheparametersweretheTable 4 Table 4 formal errors from the multi-component GaussianISM fitting.formal Masses errors from from Dust the multi-component Continuum GaussianISM fitting. Masses fromTable Dust 4 Continuum a a No convergentTable fit 4 obtained. No convergent fit obtained. ISM Masses from Dust Continuum ISM Masses from Dust Continuum a d b a c d c b c c Source ⌫obs Flux Td RJ MassSource diameter ⌫obsRadiusFlux< T⌃d gas > RJ
J=1 CO Rotational Levels CO (J=1-0) in Orion KL Nebula
J=0 1976ApJ...210L..39K A=7x10-8 sec-1 è n ~ 3000 cm-3 crit T (K) B but τ >> 1 èline photon trapping
A è A/τ è n ~ 30 – 300 cm-3 crit CO Tx è TKinetic , TB ~ TK
A1-0 = 7x10-8 sec-1 13CO x 5.1 How can an optically thick CO measure mass ?? 2 LCO = area × TCOΔV (K km/s pc ) = πR 2T ΔV -- for virial eq. , ΔV = GM / R k 1/ 2 & 3πG) = ( + TkMGMC ' 4ρ * & T ) k i.e. LCO ∝ ( 1/ 2 + MGMC ' ρ *
if T & ρ ~ constant è MGMC = constant x LCO € add mass to GMC, size and linewidth increase è increases LCO 2 constant ~ 4 - 5 M¤ / K km/s pc
Arp 220 and NGC 6240 15
conversion factor in ULIRGs estimates from the CO (1 - 0) using the standard Galactic Table 6 CO-to-H2 conversion factor are 2 - 3 times higher than Arp 220 – CO and Dust Masses the estimates obtained from the dust continuum (see Ta- ble 6). The lower CO estimates are probably more valid S2 1 VS1 0 VISM(CO) ISM(dust) for the nuclei since the flux ratio of the high excitation 109M 109M gas is very likely to be thermalized and therefore have the limiting 4:1 ratio. East 120 30 – 46 3.9 – 6.0 2.0 West 187 47 – 72 6.0 – 9.2 4.2 For self-gravitating molecular clouds or galactic nu- clei where the gas is the dominant mass component, one expects the CO-to-H2 conversion factor to scale as Note.—CO(2-1)fluxesforthenucleiarefromSakamotousing standard Galactic α et al. (1999). Similar 2-1 fluxes are given by DownesCO & pnH2 /Tk (Dickman et al. 1986; Bryant & Scoville 1996), 1 Solomon (1998): 130 and 220 Jy km s (assuming their table relative to the density and temperature of the Galactic COentries & dust-based for the East masses and West only nuclei disagree are reversed).by factor 2 Flux ! units GMCs for which the standard conversion factor was de- 1 1/2 are consistent Jy km s with.Alowerlimitonthe1-0fluxisobtainedby < n > / T scaling p 2 k rived. The Galactic GMC ratio is 300/10 1.7 and assuming the flux scales as ⌫ and the ratio is therefore 4:1. ⇠ ⇠ The upper limit is obtained by adopting the global CO (2-1) the ratio for the Arp 220 nuclei is p105/100 3.2. no basis for 4 – 8 change in α often used for ‘ULIRGs’ ⇠ ⇠ /(1-0)fluxratioofArp220(1071/410=2.6Scovilleetal.CO Thus a factor 2 change in ↵CO(1 0) is to be expected 1997). ISM(CO) is the ISM mass estimate obtained from – this is consistent⇠ with the discrepancy between the CO- the CO (1-0) flux assuming the standard Galactic CO-to-H2 conversion factor and ISM(dust) is the ISM masses estimated and dust-based ISM mass estimates given in Table 6. from dust RJ continuum flux as listed in Table 4. Thus we conclude that the e↵ective CO-to-H2 conver- tion. In some instances, the excited rotational levels can sion factor in the Arp 220 nuclei is approximately a factor be populated via absorption of near/mid infrared dust 2 reduced from the standard Galactic value. Most of this continuum photons in the molecular vibrational bands, reduction can be attributed to the expected changes in followed by spontaneous decay to the excited rotational the mean gas density and gas kinetic temperature. The levels of the ground vibrational state (Carroll & Gold- CO-to-H2 conversion factor can also be reduced if the smith 1981). For CS and HCN, this mode starts for dust line width is increased due to there being a significant black body temperatures above 120 and 400 respec- stellar mass contribution in the self-gravitating region ⇠ tively for CS and HCN but considerably higher Td are containing the molecular gas (Bryant & Scoville 1996; required for the high J states observed here (see Carroll Downes & Solomon 1998). In the Arp 220 nuclei, we & Goldsmith 1981). find that this mechanism is not necessitated since the apparent changes in the conversion factor can be largely 7. DISCUSSION attributed to the expected scaling for higher Tk and den- Here, we briefly discuss the constraints on variations sity. Lastly, we point out that the masses estimated here in the CO conversion factor used to translate CO (1 - 0) from the dust emission and from the CO (as discussed emission line flux to mass of H2 and the estimation of above) are consistent with the dynamical mass estimated star formation rates from the IR fine structure lines such from the kinematic modeling (as shown in Figure 9-left). as CII. Thus the derived conversion factor, reduced by a factor 2 from the standard one, is certainly not implausible.
7.1. CO-to-H2 conversion factor 7.2. SFRs from CII and the ’IR Line Deficits’ Using the masses derived from the dust emission in The 158µm CII line is the most luminous far-infrared 4 and Table 4, we can assess the e↵ective CO (1 - 0) § line from dusty star forming regions and a number of conversion factor for the Arp 220 nuclei (assuming the investigations have attempted to use the CII line fluxes masses obtained from the dust emission are valid). To to probe SFRs (cf. Malhotra et al. 2001; Stacey et al. circumvent the problem that angular resolution of exist- 2010). However, it is now well established that in many ing CO (1 -0) is insu cient to separate the two nuclei, ULIRGs the CII line luminosity is suppressed by factors and to separate their emission from the more extended of 10-100 relative to the far infrared luminosity (Luhman CO emission, we make use of the CO (2-1) observations et al. 2003; Graci´a-Carpio et al. 2011; D´ıaz-Santos et al. of Sakamoto et al. (1999) at 0.500 resolution. To es- 2014). The latter is a robust indicator of dust obscured timate the e↵ective (1-0) flux one can scale the (2-1) SF in galaxies not having strong AGN. flux by a factor 1/4, assuming the two lines are thermal- There are a number of explanations which have been ized and have constant brightness temperature. Alter- proposed to account for the suppressed line flux: 1) a natively, one could scale by the global (2-1)/(1-0) flux su ciently high density in the emission line regions that ratio = 2.6 (Scoville et al. 1997). The 1-0 flux obtained the fine structure levels are collisionally de-excited (i.e. using this latter ratio should be viewed as an upper limit the gas density is above the fine structure transition crit- since the global flux ratio includes low excitation CO ical density), 2) some of the UV luminosity which would from larger galactic radii. These estimates are listed in normally photoionize the HII and PDR gas is absorbed Table 6. by dust, thus enhancing LIR relative to LCII or 3) emis- For the standard Galactic CO-to-H2 conversion factor 20 2 1 1 sion line flux is suppressed by absorption due to overlying ↵CO(1 0) =3 10 H2 cm (K km s ) ,theH2 dust. ⇥ 4 2 3 mass is MH2 =1.18 10 S1 0 VdMpc (see Appendix For the CII line, the critical densities are ne 50 cm ⇥ 4 3 ' in Sanders et al. 1991). The resulting CO-based mass for ionized gas at 10 K and n 3000 cm for neutral estimates are given in column (4) of Table 6 after multi- gas at 100 K. The mean gas densities' estimated above for 5 3 plying by a factor 1.36 to account for He. The ISM mass the Arp 220 nuclei (n 5 10 cm ) are 100 times H2 ' ⇥ high excitation lines , e.g. HCN , CS , HCO+ 2 μ ~ 3 (compared w/ 0.1 for CO) è A ~ 30 ACO suggests ncrit ~ 1000 ncrit (CO) but for τ >> 1 , Aeffective è A/ τ , τ ~ A Nmol Aeffective α / Nmol --- indep. of A !! è such lines trace high column density, not volume density
CII line at 158 μm CI ionized long of Ly limit è CII in HI
H2 PDR HII lots of C è bright CII but critical density is low è collision suppression of emission at moderate densities opacity can be significant !
but since CII is bright, it is an excellent dynamics tracer
C. De Breuck et al.: ALMA resolves turbulent, rotating [CII] emission in a young starburst galaxy at z = 4.8
C. De Breuck et al.: ALMA resolves turbulent, rotating [CII] emission in a young starburst galaxy at z = 4.8
Fig. 4. [CII]velocityfield.Theleft panel presents the observed data, the central panel the best fit rotating disk model (see Sect. 3.2) and the right panel the residuals. The plus and cross mark the locations of the optical counterpart (Fig. 1)andthe[CII]peakflux(seeTable1), respectively. The observed motions of the [CII]emissionareconsistentwitharotatingdiskmodel.
Fig. 4. [CII]velocityfield.Theleft panel presents the observed data, the central panel the best fit rotating disk model (see Sect. 3.2) and the right panel the residuals. The plus and cross mark the locations of the optical counterpart (Fig. 1)andthe[CII]peakflux(seeTable1), respectively. The observedCII motions in ofa therotating [CII]emissionareconsistentwitharotatingdiskmodel. SB at z = 4.8 ALESS 73.1. De Breuck etal 2014 V/σV ~ 3 : 1
Fig. 5. Left:position–velocitydiagramalongthemajoraxisofthediskmodelofFig.10 4.Contoursshowourbestfitrotatingdiskmodelat1%, Fig. 5. Left:position–velocitydiagramalongthemajoraxisofthediskmodelofFig.M = 3x10 M (i= 30-40 deg) !! 4.Contoursshowourbestfitrotatingdiskmodelat1%, 5%,dyn 20%, 30%, 40%, 50%, 60%, 70%, 80%,¤ 90%, and 95% of the peak flux. Centre:rotationcurveextractedalongthemajoraxiswithour 5%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 95% of the peak flux. Centre:rotationcurveextractedalongthemajoraxiswithour11 best-fit model overplotted best-fit as6.6 model a dashed mjy overplotted line. continuum as aRight dashed:thevariationinvelocitydispersionasafunctionofradiusinthedisk.Weestimateatypical line. Right:thevariationinvelocitydispersionasafunctionofradiusinthedisk.Weestimateatypical at 330 Ghz è 2x10 M¤ Vrot/σint 3.1inthedisk,showingthatitisrelativelyturbulent.TherisetowardsthecentreintheobserveddispersionisanartifactresultingfromV /σ 3.1inthedisk,showingthatitisrelativelyturbulent.Therisetowardsthecentreintheobserveddispersionisanartifactresultingfrom ∼ rot int ∼ the limited spatial resolutionMthe limited of the spatial= data.1.6x10 resolution The of synthesised the10 data. M The synthesised beamCopin+ size beam size is shown is shown 2010 as as a a horizontal horizontal w/ black α bar black = in the 0.8 topbar left in corner the oftop the leftcentral corner and right of the central and right panels. panelsH2. ¤ we also modelled the disk with two alternative models. First, To derive the (model-independent) intrinsic velocity disper- we used the KINematicwe also Molecular modelled the Simulation disk with two alternative (KinMS) models. rou- First, sionTo of derive the the disk (model-independent) at each pixel intrinsic (corrected velocity disper- for the contribution of tines of Davis et al. (2013we used)TheK the KINematicINMS Molecular routine Simulation coupled (KinMS) to the rou- thesion ofvelocity the disk at gradient each pixel (corrected across for the the synthesised contribution of beam), we follow Bayesian Monte Carlotines Markov of Davis chainet al. (2013 (MCMC))TheKINMS fitter routineKinMS_fit coupled to the Swinbankthe velocity gradient et al. across(2012a the synthesised). At each beam), pixel we follow in the velocity disper- (Davis et al., in prep.) matches the brightness distribution of each sion map, we measure the luminosity weighted velocity gra- pixel in the simulatedBayesian and observed Monte Carlo datacubes, Markov chain (MCMC) rather fitter thanKinMS_fit fit- dientSwinbank across et al. (2012a the ).FWHM At each pixelof the in the beam velocity at disper- that pixel and subtract ting Gaussians like the(Davis fitting et al., code in prep.) described matches the brightness above. distributionSecond, of we each thission map, from we the measure velocity the luminosity dispersion. weighted In velocity Fig. gra-5 right, we show both used a simple arctan modelpixel in the (e.g. simulatedSwinbank and observed et al. datacubes, 2012a rather), where than fit- thedient observed across the FWHM andof intrinsic the beam at one-dimensional that pixel and subtract velocity dispersion the observed emissionting is Gaussians fitted assuming like the fitting its code rotation described above. curve Second, uses we profilethis from wethe velocity derived, dispersion. extracted In Fig. 5 alongright, we the show major both kinematic axis of 1 the form v(r) = 2π− vasymused aarctan simple arctan (r/rt model), where (e.g. Swinbankvasym is et the al. 2012a asymp-), where thethe observed galaxy. and This intrinsic shows one-dimensional that the velocity intrinsic dispersion velocity dispersion of totic rotational velocity and r is the effective radius at which the disk is σ = 40 10 km s 1 (Fig. 5 right). The ratio the observedt emission is fitted assuming its rotation curve uses profile we derived,int extracted along± the major kinematic− axis of the rotation curve turns over. Both the1 alternative models, which of rotational-to-dispersion-support vrot/σint = 3.1 1.0imply- the form v(r) = 2π− vasym arctan (r/rt), where vasym is the asymp- the galaxy. This shows that the intrinsic velocity dispersion of ± have significantly different flux distributions from the model de- ing that this is a highly turbulent1 rotating disk. Such values are scribed above, obtaintotic similar rotational results. velocity This and rt is provides the effective confidence radius at which afactorofthe disk is σint three= 40 lower10 km s than− (Fig. local5 right). disk The galaxies ratio observed in CO ∼ ± that our assumption ofthe rotation a rotating curve turns disk over. is Both a the good alternative (though models, not which (e.g.of rotational-to-dispersion-supportDownes & Solomonvrot 1998/σint =),3. but1 1 comparable.0imply- to other high- necessarily unique) representationhave significantly di offferent the flux observed distributions [C fromII]velocity the model de- redshifting that this disks is a highly with turbulent similar rotating resolution disk. Such± valuesdata arefrom the Hα line (e.g. field. However, as wescribed barely above, spatially obtain similar resolve results. the This flux provides distribu- confidence Cresciafactorof etthree al. lower 2009 than; Genzel local disk etgalaxies al. 2011 observed; Swinbank in CO et al. 2012a). tion within the disk, wethat cannot our assumption distinguish of a rotating which disk is flux a good distribu- (though not Carniani(e.g. Downes∼ et & al. Solomon(2013 1998)reportsa[C), but comparableII to] otherv/σ high-1.5inboththeSMG tion is more appropriate. We will therefore quote the full range and quasar in the BRI1202 system, while∼ in a quadruple system of uncertainties from allnecessarily three unique) models representation in any parameters of the observed [C derivedII]velocity observedredshift disks in with CO, similarIvison resolution et data al. from(2013 the H)reportα line (e.g. v/σ 6inthetwo from these models (notablyfield. However, the dynamical as we barely mass,spatially see resolve Sect. the flux 3.3.1). distribu- brightestCresci et al. 2009 systems,; Genzel and et al. 2011v/σ; Swinbank< 1inthefaintestsystems.Wedo et al. 2012a). ∼ tion within the disk, we cannot distinguish which flux distribu- Carniani et al. (2013)reportsa[CII] v/σ 1.5inboththeSMG tion is more appropriate. We will therefore quote the full range and quasar in the BRI1202 system, while∼ in a quadruple system A59, page 5 of 10 of uncertainties from all three models in any parameters derived observed in CO, Ivison et al. (2013)reportv/σ 6inthetwo from these models (notably the dynamical mass, see Sect. 3.3.1). brightest systems, and v/σ < 1inthefaintestsystems.Wedo∼
A59, page 5 of 10 [CII] Emission from Low Redshift Analogs of Epoch of Reionization Galaxies
Ranga Ram Chary (IPAC/Caltech), Hyunjin Shim (KNU, S. Korea)!
Abstract( Observa-ons(&(Results( ! We present Herschel/PACS spectroscopy of a CII and FIR A15 -- Chary & Shim population of ~20 local (z<0.2), star-forming galaxies Herschel-Pacs• We present results CII fromfor 20 Herschel/PACS low z gal – CII/FIRspectroscopy mergers of 20 of & these Z local galaxies with evidence for high ionization identified in the Sloan Digital Sky Survey whose specific parameter.! star-formation rates and extreme emission line properties detect! CII in 10/20 4 non-mergers (¤) 6 mergers () make them very similar to z>5 Lyman-break galaxies. We • We detect [CII] in half the sample, with 6 clear mergers (solid circles) and 4 non-mergers (!)! detect [CII] emission in about half the sample, with [CII]/ LIR ratios which lie between 10-4 and 10-3. We attempt to Quasar host galaxies w/ low CII correlate the strength of [CII] with the extensively studied Wang etal 2014 multi-wavelength properties of these galaxies. In The Astrophysical Journal,773:44(10pp),2013August10 Wang et al. particular, we assess if the [CII]/LIR ratios correlate with detections in this work, J1148+5251 (Maiolino et al. 2005), morphology or metallicity and find some evidence that CFHQS J0210 0456 at z 6.43 (Willott et al. 2013), and ULAS J1120+0640− at z = 7.08 (Venemans et al. 2012). clear mergers may have lower ratios but do not find We compare them to samples= of [C ii]-detected local normal star forming galaxies, ULIRGs, submillimeter galaxies, and evidence for a metallicity dependence. We predict based FIR-luminous quasars at high redshift (Malhotra et al. 2001; Luhman et al. 2003;Staceyetal.2010; Maiolino et al. 2009; on these measurements that ALMA spectroscopy of z>5 /SFR Ivison et al. 2010;DeBreucketal.2011;Swinbanketal.2012; galaxies should reveal median [CII]/LIR ratios of CII Wagg et al. 2012; Pety et al. 2004;Galleranietal.2012;Carilli 3.5×10-3. et al. 2013;Valtchanovetal.2011;Riechersetal.2013;Marsden =L et al. 2005). The five z 6quasarspresentedinthiswork,as ∼ 4 well as J1148+5251, show luminosity ratios of 2.9–5.1 10− , CII ×
ε which are comparable to the typical values found in local ULIRGs and 1 " z " 5[Cii]-detected quasars, and a few to 10 times lower than that of the disk star forming galaxies and submillimeter galaxies. We also notice that the other two z>6 quasars with moderate FIR luminosities (1011–1012 L )show Proper-es(of(z>5(Galaxies! ⊙ higher L[C ii]/LFIR than most FIR luminous objects (Venemans • More than 70% of z>5 galaxies in extragalactic deep è mergers have lower CII / FIR et al. 2012; Willott et al. 2013). We estimate the CO (1–0) • [CII] correlates well with Hα and [OIII], consistent with a direct correlation with the linestrength luminosities of forthe the ionizing five objects radiation from the PdBI CO (6–5) fields show evidence for strong nebular line emission, detections (last column of Table 1), assuming a CO excitation field.! L /L L Figure 4. [C ii] FIR vs. FIR.Theblackstarsindicatethequasarsat ladder similar to J1148+5251 (Riechers et al. 2009). The particularly Hα, but [OIII] as well (Chary et al. 2005, z ! 5.8, including the five new [C ii]-detected z 6 quasars in this work, J1148+5251 at z 6.42 (Maiolino et al. 2005;Leipskietal.∼ 2013), CFHQS calculated [C ii]-to-CO (1–0) line luminosity ratios are about Shim et al. 2011, Stark et al. 2013).! J0210 0456 at z = 6.43 (Willott et al. 2013), and the z 7.08 quasar ULAS 2400–4700, which is slightly higher than the values found in − = = • Mergers appear to have lower [CII] emissionJ112001.48+064124.3 compared (Venemans to et al.non-mergers.2012). We also plot the ! luminosity local ULIRGs (Luhman et al. 2003), and the highest value is ratios from samples of local normal star forming galaxies (Malhotra et al. close to the median luminosity ratios found in starburst galaxies 2001), ULIRGs (Luhman et al. 2003), z>1 star forming galaxies and mixed (i.e., L /L 4400; Stacey et al. 1991, 2010;Swinbank • The origin of this nebular emission appears to be due systems (Stacey et al. 2010;Marsdenetal.2005), z>2 submillimeter bright [C ii] CO 40-42 • Applying the same ratios for high-z galaxies,galaxies (SMG; z>5 Maiolino galaxies et al. 2009;Ivisonetal. should2010 have;DeBreucketal. [CII]2011 luminosities; et al. 2012 ).in the∼ range of 10 erg/s ! to a population of hot young stars which are Swinbank et al. 2012;Waggetal.2012;Valtchanovetal.2011;Riechersetal. The [C ii], FIR, and CO luminosity ratios of the five 2013), and 1
• Based on the nebular line emission, the ionizing radiation field is thought to be stronger in HAEs; that is supported by the fact that the FUV strength required to explain the [CII] emission is x100-1000 more intense than the Milky Way.! • Thus top-heavy IMFs responsible for hard radiation fields may be ubiquitous at z>6 as predicted in Chary 2008 ". ! • I would have liked to show ALMA data at an ALMA conference but our 3 hr ALMA integration at 1.4mm of high-z galaxies showed no detections # (Berger et al. 2014).!
!
References( Acknowledgments( © Copyright Colin Purrington for the poster template. Berger et al. 2014, ApJ, (arXiv: 1408.2520) Shim & Chary, 2013, ApJ, 765, 26 Shim et al., 2011, ApJ, 738, 69 Chary, 2008, ApJ, 680, 32
CII line deficit : CII emission efficiency : ε = L / SFR CII CII
LCII ∝ nnCII σv × volume for n < ncrit
∝ nCII × volume for n > ncrit
nCII ∝ MCII / volume
SFR ∝ nCIIn × volume
εCII ∝ constant − − independent of n, SFR for n < ncrit −1 εCII ∝ n for n > ncrit
n > n n < ncrit crit LCII/SFR €
SFR Capak etal ’15
ALMA w/ 20 min integ. detects 9 normal (LBG) galaxies in CII at z = 5-6
HZ1 HZ2 HZ3 HZ4 have low IR emission
HZ5 HZ6 HZ7 HZ8
HZ5a
HZ8W
HZ9 HZ10 N
1” E
Figure 1 | The [CII] line detections (red contours) and weak ~158μm FIR continuum detections (blue contours) are shown with the rest frame UV images as the background. The images are 5"x5" and the contours are 2, 6 and 10 σ with [CII] line profiles for each source shown in Figure 3. The background images are from HST-ACS in the F814W16 band where the morphologies will be affected by Ly-α, except for HZ10, which is Subaru z' band. All objects are detected in [CII] showing that a large amount of gas is present in these systems, but only 4 are detected in continuum. summary : conditions starbursts (Arp 220 -- ULIRGs) : 9 radii ~ 25 – 50 pc !! , M ~ 2-4x10 M¤ , mostly gas densities ~ 105 cm-3 (from dust, HCN, CO and grav.) ~ all the ISM mass of Milky Way into a 50 pc cloud dust continuum can be used to track evolution of ISM in gal. ALMA is revolutionary !! 2 new images – at 0.025 arcsec res.