Three-dimensional Line Transfer Studies of Compact Molecular ISM at - PowerPoint PPT Presentation

“Millimeter and Submillimeter Astronomy at High Angular Resolution”(June.8.2009) Three-dimensional Line Transfer Studies of Compact Molecular ISM at the Centers of Active Galaxies -Models meet Observation- Masako YAMADA(ASIAA) K. Wada, K. Tomisaka, (Y. Kurono)(NAOJ) Yamada, Wada & Tomisaka 2007 Yamada & Tomisaka 2009 (submitted) 2009 年 6 月 8 日月曜日 1

MY, Wada & Tomisaka(2007) I.Mol. Gas in Active Galaxies MY & Tomisaka, submitted NGC6764 (DSS:V band) ✦ compact molecular gas at the active center (R<1kpc) ✦ unified model predicts an obscuring torus [AGN] ✦ many molecular lines (CO,HCN,HCO + ,CS...) have been detected in numbers of galaxies ✦ Three(+) reasons for study molecular ISM 1. Formation & Evolution of Galaxies c p 0 0 1 ~ ✦ molecular gas as star formation site Large scale structure → (???) → present star formation (GMC, mol.core, YSO) physical conditions of the “compact” gases probed by mm/submm lines Yoshida et al.(2006) 2. Chemical & Thermal evolution of Molecular Gas in Extreme Environments 【 Astrochemistry viewpoint 】 ✦ Do mol. gas in active galaxies have peculiar character compared with normal galaxies? 3. As the Energy Budget of Active Galaxies ✦ AGN v.s. starburst ? which dominates the energy source?? 2009 年 6 月 8 日月曜日 2

Current Observations of Nuclear Mol. gas 12 CO(2-1) SMA ← CO obs. of NGC 1097 (Seyfert 1) 1 kpc 1kpc Kohno et al. 2003 VLT MELIPAL + P.-Y. Hsieh et al., 2008 PASJ, 55, L1 VIMOS ISM is in general very inhomogeneous & multi-phase in nature ✦ Current mm./submm. obs found “compact cores” at the centers of active galaxies ✦ typical size of a “compact core”~ 500pc - 1kpc : is it really a single entity? NO!! ✦ Christopher et al. 2005 non-local rad. coupling would take place (e.g. GC) ✦ ex.) Global 3D line- Galactic Center transfer simulation Composite (red: radio; green: mid-infrared; blue: X-ray) 2009 年 6 月 8 日月曜日 3

Physics of ISM : lines as a toolbox obs.: data cube(x, y, ν ) ISM : T kin (x, y, z) n(x, y, z) v=(v x , v y , v z ) τ ν , T ex y mol (x, y, z).. interpreting data sets of I ν in terms ✦ of T kin , n, y(=n mol /n H ), v is not straightforward line RT can form a toolbox to ✦ decipher tangled “riddles” printed in observed line data cube My dear Watson, circumstance evidence is a very tricky thing... and there is nothing more deceptive than an “obvious fact”. 2009 年 6 月 8 日月曜日 4

64pc II. Calculations (density) Hydrodynamic simulations: ✦ 256x256x128grids( → 64x64x32 grids) ✦ evolution of rotating gas in gravitational ✦ potential of SMBH and halo radiative cooling and SNe heating ✦ feedback are included highly clumpy & turbulent torus 20K<T<1000K, n H < 2x10 6 cm -3 ✦ Wada&Tomisaka, 2005 , Δ v ～ 50km/s torus is globally quasi-steady(Wada&Norman) ✦ ✦ Radiative Transfer: [ray tracing with long characteristics method] non-LTE level population up to J=10 for each grid ✦ assume uniform chemical abundance distribution ✦ 【 to examine chemistry & clumpiness separately 】 0.6km/s<v therm <4.1km/s ⇔ Δ v turb ~50km/s 、 ✦ → velocity structure is taken into account by absorption coeff. profile with micro-turbulence − ( v − v 0 ) 2 � � 1 1 φ ( ν ) d ν = √ π exp dv ∆ v 2 ∆ v turb turb V turb =20km/s Hogerheijde&van der Tak(2000) 2009 年 6 月 8 日月曜日 5

III. Results : Intensity Distribution HCN(1-0), θ =0deg(face-on). HCO + (1-0), θ =0deg. 45deg. y=2x10 -9 , v turb =20km/s ✦ Intensity distributions look quite alike of similar strength HCN/HCO + rotational lines : B(rotational constant) ✦ & μ (electric dipole moment) are almost identical for both molecules 90deg. ✦ HCN(1-0) and HCO + (1-0) lines display clumpy distribution, reflecting the inhomogeneous structure of AGN mol. torus R clump <O(10pc) : Δθ ～ 0.1”@D=20Mpc ✦ ALMA can resolve them 2009 年 6 月 8 日月曜日 6

Current Observations of Nuclear Mol. gas 12 CO(2-1) SMA ← CO obs. of NGC 1097 (Seyfert 1) 1 kpc 1kpc Kohno et al. 2003 VLT MELIPAL + P.-Y. Hsieh et al., 2004 ↓ simulation (HCN) PASJ, 55, L1 VIMOS ISM is in general very inhomogeneous ✦ R clump <O(10pc) : Δθ ～ 0.1”@D=20Mpc in ✦ our simulations -- currently substructures in a compact ✦ nuclei might be smeared out in a obs. beam... .. but we can predict and prepare ✦ what we could expect in ALMA era 2009 年 6 月 8 日月曜日 7

Line Profiles (I) HCO + (1-0) HCN(1-0) Tb[K] Current obs.-> average turbulent bulk veolocity ✦ ALMA-> can find clumps inside ✦ Vr[km/s] obs: NMA(Kohno et al.) Gaussian ave.profile 2009 年 6 月 8 日月曜日 8

Line Profiles (II) H 12 CN H 13 CN (a) (b) J=1-0 J=4-3 1.5 8 6 1.0 T b [K] T b [K] 4 0.5 2 0.0 0 -2 -1 0 1 2 -2 -1 0 1 2 V r /100 [km sec -1 ] V r /100 [km sec -1 ] thick lines and thin isotopologue -> a good indicator of clumpiness ✦ ALMA-> can distinguish line profile distributions of HCN & H13CN ✦ statistical studies of mol. clouds even in the centers of distant galaxies as well as ✦ dynamics will be available with ALMA 2009 年 6 月 8 日月曜日 9

non-LTE effects: Optical Depth & Intensity ✦ Distributions of N H , integrated intensity, and optical thickness τ 0 T b - N H good correlation between N H & Intensity, but not with τ 0 ✦ ✦ nonLTE n J (T kin , n H ) in clumpy torus generates a dispersion of α ν → τ 0 -N H relation widen as the dispersion of α ν (including negative α ν due to pop. inversion) LVG analysis cannot reasonably reproduce (T b , N H , τ 0 ) ✦ τ 0 - N H T b - τ 0 HCN(1-0), y=2x10 -9 2009 年 6 月 8 日月曜日 10

Excitation Temperature Distribution HCN thin limit Tex y=10 -8 HCN y=10 -9 y=10 -10 HCO + LTE T ex <0 low n T=10K, 30K, 100K, 200K, 1,000K(light color lines for low T) ✦ increment of y(abundance) raises the emission rate (j ij =h ν ij A ij ・ n i , n i ∝ y*n H2 ) average torus temperature <T kin >=174K >> T 10 ～ 4.5K ✦ emission from torus stronger than CMB leads to high T ex & weak overshoot ✦ ✦ photon trapping( β ) : lowers effective n cirt (T ex [thin] is shift to lower n H2 ) in our simulation, < τ 0 > ≈ O(1) and β ≦ 1 ( → minor effect compared with the former) ✦ 2009 年 6 月 8 日月曜日 11

non-LTE effects: Optical Depth & Intensity ✦ highly clumpy torus → N H is composed of color: intensity, contour: τ 0 =1 HCN(1-0), y=2*10 -9 1. tenuous gas encompasses the large scale height 2. dense clump + small amount of tenuous ambient Both of 1, 2 are OK ✦ intensity become strong if dense ( Λ ∝ n 2 ) and/or pop. inversion (@ n ～ n crit ) (2):sub-thermal (1):pop.inversion (3):single clump density ∆ τ 2009 年 6 月 8 日月曜日 12

IV.Results of RT Simulations [HCN/HCO + ] rad. transfer results observations HCN/HCO + “pure AGN”(XDR?) “starburst(PDR?)” HCN/CO Imanishi et al. 2004, Kohno et al. 2005 R HCN/HCO+ >2 is observed in a number of galaxies, ✦ but results of rad. transfer suggest “R HCN/HCO+ =1 HCO + ceiling” HCN In order to obtain R HCN/HCO+ ~2: ✦ ✦ HCN should be much abundant than HCO + ↑↓ ✦ XDR/PDR models (y HCN <y HCO+ in XDR) Yamada, Wada, & Tomisaka, 2007 2009 年 6 月 8 日月曜日 13

V. High-J Lines in inhomogeneous ISM ✦ Multi-level analysis → way to evaluate precise (T kin , n) “high density tracer” is necessary to reveal density structure ✦ T kin : high-J transition obs. in submm. band is now AVAILABLE ✦ mol. ν 10 ν J,J-1 HCN 88.6GHz 354.5GHz (J=4-3) CO 115.3GHz 345.8GHz(J=3-2) ✦ Increasing # of obs. of high-J lines have found variation in R 43/10 & possible chemical abundance variation Papadoupolus, 2007 What do high-J lines tell us about inhomogeneous ISM? 2009 年 6 月 8 日月曜日 14

V. HCN R 43/10 face-on edge-on ✦ average ratio takes a value around 1 ⇔ peak-to-peak ratio >1 ✦ difference between (a)&(b) → clumpiness inside ✦ as y increases, R 43/10 decreases below 1 ✦ one-zone analysis suggests R 43/10 ->1 as tau increases ?? Multi-phase nature should be taken into account 2009 年 6 月 8 日月曜日 15

Multi-zone Excitation Analysis : R 43/10 ✦ Model multi-phase torus with a simple isothermal two-phase ISM 2 phases = (dense clumps + tenuous ambient), & optically thin over a whole region ✦ If we assume optically thin, average line ratio becomes : ✦ � vol ( n H 2 yf 4 A 43 h ν 43 ) � vol ( n H 2 f 4 A 43 ) vol ( n H 2 yf 1 A 10 h ν 10 ) = ν 43 R ′ 43 / 10 ≃ × � � vol ( n H 2 f 1 A 10 ) ν 10 =(A) pxpxpxpxpxpx ξ 43 ( n H 2 f 4 A 43 ) d + (1 − ξ 43 )( n H 2 f 4 A 43 ) t ( A ) = ξ 10 ( n H 2 f 1 A 10 ) d + (1 − ξ 10 )( n H 2 f 1 A 10 ) t � � − h ν 41 ξ 4 4 exp ( n H 2 f 1 A 10 ) d + (1 − ξ )( n 2 H 2 G ( T kin , J = 4) f 0 ) t k B T kin ≃ ξ ( n H 2 f 1 A 10 ) d + (1 − ξ )( n 2 H 2 G ( T kin , J = 1) f 0 ) t � G ( T kin , J ) ≡ γ J, 0 t. J d. in dense clumps : ✦ n J ~thermalized in tenuous ambient : n 0 ~1 (balance between coll. excitation and spontaneous decay) 2009 年 6 月 8 日月曜日 16