Dust processing of radiation in galaxies & Frank Heymann Ground based MIR observations Dust model Starburst Spirals AGN + vectorized Monte Carlo CenA (ESO)
Ground based MIR imaging Nuclear activity in nearby galaxies Surface brightness AGN viz. starbursts 10m telescopes
Ground based MIR imaging Nuclear activity in nearby galaxies Surface brightness AGN viz. starbursts - AGN - Starburst - Spirals Siebenmorgen Ralf et al. 08 Dwarfs: Madden, Lisenfeld
ELT 40m MIR surface brightness up to D ~ 500Mpc + 25 times deeper
MIR spectroscopy PAH No PAH Silicate emission Siebenmorgen Ralf et al. (2004, 2005) more on pc scale by Almudena Prieto
ISM dust Fitzpatrick (99)
ISM dust extinction Large Si + ac : 60Å < a <0.2-0.3µm ~ a -3.5 Small Graphite : 5Å < a < 80 Å ~a -3.5 PAH : 30 C + 200 C
ISM dust extinction Buat et al. 2011, at 1<z<2 : - bump 35% of MW - UV flatter SMC 1<z<2 LMC MW
ISM dust solar neighborhood Abundances [X/H in ppm]: 31Si + 150aC + 50gr + 30PAH extinction Si + ac : 60Å < a <0.2-0.3µm ~ a -3.5 Graphite : 5Å < a < 80 Å ~a -3.5 PAH : 30 C + 200 C
PAH band ratios: ionisation dehydrogenation
E o PAH destruction Unimolecular dissociation Arrhenius form: t dis ~ exp(e o /kt) / ν 0 «t cool ~ 1s T [K] t cool T min = E o /k ln(ν 0 ) ~2000K; E o ~ 5eV; ν 0 = 10 13 Hz ΔE = 3N c kt min ~ 0.1 N c. E b => N c < 2 ΔE /[ev] (PAH unstable) t abs ~ 1h time Tielens (2005) ; Micelotta et al. (2010) ; Siebenmorgen & Krűgel (2010)
PAH destruction E o Arrhenius form: t dis ~ exp(e o /kt) / ν 0 «t cool ~ 1s T min = E o /k ln(ν 0 ) ~2000K; E o ~ 5eV; ν 0 = 10 13 Hz ΔE = 3N c kt min ~ 0.1 N c. E b => N c < 2 ΔE /[ev] (PAH unstable) 1) single hard photon : independent of distance 2) many soft photons : ~inner torus Omont (1986) ; Tielens (2005) ; Siebenmorgen & Krűgel (2010)
Dust Radiative Transfer (RT) + Ray tracing techniques + Monte Carlo techniques 'exact RT' = dust self-absorption, multiple scattering, etc. Each team requires to implement approximations which are (sometimes) difficult to trace back. Models fit SED of Arp220 but differ in: M dust, R dust and
Dust radiative transfer models Mapping: Groves (talk), Dopita Efstathiou (talk) & Rowan-Robinson (2003) Takagi et al. (2003) Grasil: Silva, Granato 1D, no dust self-absorption only forward scattering 1D, (2D), uncoupled J MC +J cir, exact RT in SB, cirus as screen 1D, exact RT of homogenous distributed stars in dust sphere, no clumps 2d, uncoupled superposition of J * +J MC +J cir Cirus: no dust self absorption, approx. scattering Popescu, Tuffs, Kylavis, Xilours 2d, uncoupled bulge+2disk+clumps (template) no dust self absorption, 1 st scattering +approx. Siebenmorgen et al. 1+1D, exact RT, coupled J * +J MC +J cir
Krügel & Siebenmorgen (1994) ; Tutokov & Krügel (1978) (2007)
Starbursts Nucleus of NGC1808 Siebenmorgen et al. (2001)
PAH cross section Siebenmorgen et al. (2001), Draine & Li (2007); Tielens (2008)
SED library for SB luminosity size mass www.eso.org/~rsiebenm Siebenmorgen et al. (2007)
SB models of the nucleus: - UV: add photo-evolution models - submm: add cold dust - Outer region of M82 talk by Kaneda et al. (2011)
SMG at high z Efstathiou & Siebenmorgen (2009)
Spirals Controversy on energy balance M17 L FIR ~ 3 L * NGC891 M31
Model of bulge + old stellar population Xilouris et al. (1998)
NGC891 2 Disk components NGC 891 Ralf Popescu Siebenmorgen et al. (2000,2011) : SED + MIR radial, vertical profiles
2 Disk components Popescu et al. (2011,2000)
2 Disk components Bulge Popescu et al. (2000,2011)
Tau(B)=3.5 NGC891 2 Disk components Attenuation-inclination relation (Driver et al. 2007)
Tau(B)=3.5 NGC891 2 Disk components Hot Spots? Bianchi (2008)
Tau(B)=3.5 NGC891 1 Disk L FIR = 3 L * Baes et al. (2011) Talk: Andreas Efstathiou on 1disk model Ilse De Looze on Sombrero Large fluffy grains?
Krügel&Siebenmorgen (1994) Ralf talk Siebenmorgen by Bendo et al. (2010)
AGN + vectorized Monte Carlo CenA (ESO)
AGN + vectorized Monte Carlo face-on edge-on - density structure 3D models: Heymann (PhD) Schartmann et al. (2008) MIDI: Tristram (2007) Statistical model: Nenkova et al.(2002, 2008)
Arbitrary dust distribution Pseudo adaptive mesh Monte Carlo 1. geometry
Monte Carlo Source emits photon packages of equal energy 1. geometry 2. source
Monte Carlo = - ln(ζ) absorption / scattering / no interaction 1. geometry 2. source 3. inter-action 4. dust temperature
Monte Carlo Photons escape model cloud 1. geometry 2. source 3. inter-action 4. temperature 5. detection
MC parallelization Multiple photons at a time: Challenges Cell locked when hit by photon Parallel random number generator (Mersene Twister) Computer games Graphical Processing Units (CUDA) Heymann (2010, PhD)
MC methods Method Parallelization Advantage Time Lucy Bjorkman & Wood YES (but floating) Partly (not independent) Benchmark sphere τ~1000) Optical thin >1h No iteration 5min our YES GPU <1min
2D benchmark face-on edge-on 1 = 10 100 ~10% for 0
AGN + vectorized Monte Carlo - Application Siebenmorgen et al. (2005)
Model X-ray heating (E&RR 93) N clumps ~ 200 5000 r i ~ dust sublimation r out ~ 20 r i dust temperatures [K]
Shadows caused by clump structure dark side: cold bright side: hot
Emission + scattering light AGN tours
MIR observations from ground Conclusions Dust model of the diffuse ISM for extinction and emission applied to external galaxies PAH: - ionisation viz hydrogenation - destruction by hard photons Starburst: - simplifications in RT - SED library for dust in SB nuclei Spirals: AGN: - controversy on energy balance - 2 disks viz. fluffy grains - vectorized MC on GPU - clumpy AGN torus model 3C249.1
Mean SED of 3CR sources Data: Haas et al (2008), Leipski et al. (2010) Θ
2D benchmark Pascucci et al. (2004) Disk: T * = 5800K L * = L sun ρ(r) : hydro static equilibrium (Chiang & Goldreich 1997)
Number of clouds SED of AGN Quasar RG
PAH in a mono-energetic heating bath if U f U i hν < ½ ΔU f : A fi = K ν F ν / hν & Krűgel (2010)
1D MC versus benchmark 1 A V =10 100 1000 mag Sphere T * = 2500K ρ(r) = const. ~5% for 0
Comparison of 2 ray tracing codes Dust sphere: A V = 1000mag, heated by star Benchmark = Dusty code (Iveciz et al. 1999): unphysical at faint flux levels
PAH replenishment Vertical mixing in torus? l /v = t exp > N C t abs & Krűgel (2010)
AGN and hard photons (EUV, soft X-ray)
Monte Carlo + PAH
Monte Carlo + PAH store PAH absorption events of each cell compute PAH emission neglect PAH self absorption
PAH band ratios: Observations
PAH band ratios: Hardness of exciting radiation
Voshchinikov (2004) ISM dust polarisation Linear polarisation spheroids Circular polarisation Prolate Oblate
ISM dust polarisation spherical grains