Andromeda in all colours

Transcription

1 Andromeda in all colours Sébastien Viaene and the HELGA team: J. Fritz, M. Baes, G.J. Bendo, J.A.D.L. Blommaert, M. Boquien, A. Boselli, L. Ciesla, L. Cortese, I. De Looze, W.K. Gear, G. Gentile, T.M. Hughes, T. Jarrett, O.L. Karczewski, M.W.L. Smith, L. Spinoglio, A. Tamm, E. Tempel, D. Thilker, J. Verstappen

2 Why we like M31: Large galaxy (>10 10 M ) Nearby (< 1 Mpc) Properties of ETG Properties of LTG Signs of a troubled past

3 Outline A vast dataset SED fitting and dust scaling relations Radiative Transfer Simulations Towards a complete model of M31

4 Herschel Exploitation of Local Galaxy Andromeda HI + optical Thilker et al. (2004)

5 Herschel Exploitation of Local Galaxy Andromeda PACS 160 µm Fritz et al. (2012)

6 HELGA: Dataset Modelling the panchromatic SED of M µm 160 µm 250 µm 350 µm 500 µm

7 HELGA: Herschel maps Cold dust emission

8 HELGA: Herschel maps Cold dust emission

9 HELGA: FIR maps MIPS 70 µm MIPS 24 µm WISE 22 µm K. Gordon, T. Jarrett

10 HELGA: FIR maps Warm dust emission

11 HELGA: FIR maps Warm dust emission

12 HELGA: MIR maps WISE 12 µm IRAC 8 µm IRAC 5.8 µm WISE 3.3 µm T. Jarret, P. Barmby

13 HELGA: MIR maps Hot dust / PAH + stellar emission

14 HELGA: MIR maps Hot dust / PAH + stellar emission

15 HELGA: Optical/UV Composite gri (SDSS) E. Tempel GALEX NUV D. Thilker

16 HELGA: Optical/UV Stellar emission

17 HELGA: Optical/UV Stellar emission (unattenuated)

18 HELGA: Optical/UV Stellar emission (attenuated)

19 HELGA: Zooming in Pixel-by-pixel SED fitting Masking foreground stars Convolution to SPIRE 500 μm beam Same pixel grid Working resolution: 36 -> 140 pc

20 Resulting Images NUV 500 µm

21 Resulting Images NUV 500 µm

22 Resulting Images NUV 500 µm independent pixels!

23 MAGPHYS: SED fitting Multi-wavelength Analysis of Galaxy PHYSical properties E. da Cunha et al Bayesian SED fits theoretical SEDs Construct Probability Density Functions (PDFs)

24 M31: Main Regions log( L /L ) log( L /L ) Bulge pixel 2 = 4.15 Ring pixel 2 = (µm) Inner Disk pixel 2 = 2.21 Outer Disk pixel 2 = (µm)

25 L dust Offset from centre (degrees) L M dust LINEAR L Tot C L Tot PAH L TC ISM M? ssfr M31: Parameter maps K SFR M L M

26 T ISM C M? ssfr M31: Parameter maps K yr 1 SFR V V ISM TW BC f µ K M M yr 1

27 M31: Local vs Global

28 M31: Local vs Global

29 HRS: Dust scaling relations 1 log(mdust/m?) log(µ? )[M /kpc 2 ] NUV-r Cortese et al. 2012

30 HRS: Dust scaling relations log(mdust/m?) Outer disk Ring M31 Inner disk Bulge Outer disk Ring Inner disk M31 Bulge log(µ? )[M /kpc 2 ] NUV-r

31 HRS: Dust scaling relations 1 log(mdust/m?) log(µ? )[M /kpc 2 ] NUV-r

32 HRS: Dust scaling relations 1 log(mdust/m?) log(µ? )[M /kpc 2 ] NUV-r

33 M31: Dust scaling relations 1 log(mdust/m?) log(mdust/m?) log(µ? )[M / kpc 2 ] NUV-r

34 M31: Dust scaling relations 1 log(mdust/m?) log(mdust/m?) log(µ? )[M / kpc 2 ] NUV-r

35 M31: Dust heating sources

36 M31: Dust Scaling Relations In Summary Panchromatic, sub-kpc SED modelling is now possible, BUT requires: Special data treatment (masking, convolution,...) Extended parameter space Resolved maps pc - of stellar and dust properties Sub-kpc regions follow galaxy-galaxy dust scaling relations; local nature of the underlying processes Viaene et al. A&A, 567, A71

37 The bulge of M31 NUV Hα 8 μm 22 μm 250 μm HI

38 The bulge of M31 NUV Hα 8 μm Can radiative transfer 22 μm simulations 250 μm bring an answer? HI

39 Why 3D RT modeling? Panchromatic (UV-mm) 3D RT models Self-consistent study of dust attenuation + dust emission 3D asymmetric geometry of stars & dust Non-local character of dust heating What can we learn from 3D RT models? 3D spatial distribution, clumpiness of stars & dust Dust heating by old/young stars at λ IR Non-locality of dust heating Grain composition/size distribution/properties De Looze+2012a

40 The dust radiative transfer equation Simulateously solved 6D problem in space: propagation of photons across the domain in direction: scattering couples the intensity in all directions in wavelength: absorption/re-emission changes wavelength

41 SKIRT Stellar Kinematics Including Radiative Transfer 3D continuum Monte Carlo C++ code in OOP-fashion in QT framework (Camps & Baes 2015) Parallel for shared and distributed memory systems Peeling-off technique, continuous absorption forced scattering, smart detectors,... A variety of geometries (stellar, dust), different dust properties, grid structures,... Smooth and clumpy models Mono-, oligo- and panchromatic mode LTE and NLTE

42 Forward Radiative Transfer Source Sink Source

43 Inverse Radiative Transfer

44 Inverse RT: FitSKIRT Optimization using genetic algorithms C++, OOP, QT framework Stellar emission (~ U to K-band) Multiple geometries (mainly edge-on spirals) Shared/Distributed memory parallelization Handles MC noise/ degeneracies

45 Monochromatic fitting

46 Oligochromatic fitting

47 Advanced dust grids Octree grids Saftly+2013

48 Advanced dust grids Octree grids kd-tree grids Saftly+2013

49 Advanced dust grids Octree grids kd-tree grids Voronoi grids Saftly+2013 Camps+2013

50 Small scale structure ISM filaments Circumbinary dust Star forming region AGN torus

51 Simulating simulated galaxies

52 Simulating simulated galaxies Saftly et al. 2014

53 Almost face-on (i 20 ) D 8.4 Mpc Huge dataset (FUV micron) Complex geometry Enables the construction of a highresolution multi-wavelength RT model

54 RT model reconstruction 3 main RT model components: Bulge: old stars (no dust) Thick disk: old stars Thin disk: dust (h z,d = 1/2*h z, ) + (non)ionizing young stars

55 3D geometry Old stars: 2D geometry (x-y plane) 3D geometry (add z): exp(- z /h z ), h z = 450 pc IRAC 3.6 micron Disc Bulge

56 3D geometry Young stars: 2D geometry (x-y plane) 3D geometry (add z): exp(- z /h z ), h z = 100 pc Myr < 10 Myr non-ionizing: FUV ionizing: Hα+24 micron

57 3D geometry Dust: 2D geometry (x-y plane) A FUV map: FUV-H to derive calibration coefficients 3D geometry (add z): exp(- z /h z ), h z = 225 pc

58 Model results Global SED!!

59 !! Spatially resolved maps Model results

60

61 Model results!! Colour maps Model&has&too&warm&colours in&center!

62 Model results Model deviations up to 50% spatial variation in ages/size of star clusters e.g. younger stars in outer regions due to interaction with NGC 5195 Relative grain abundance variations (PAH, VSG) e.g. PAH destruction in hard RF, grain shattering due to shocks) More diffuse dust component with large h z,d Vlahakis+2013 R![kpc] R![kpc] Calzetti+2005

63 Dust heating analysis

64 Dust heating analysis Global TIR emission: 63% by young stars 37% by old stars

65 Dust heating analysis Galaxy with M M F TIR,young = 100% for SFR 2.0 M /yr F TIR,young = 50% for SFR = 0.5 M /yr Caution to link TIR to SF in non-starburst galaxies!

66

67

68

69 M31: RT model? Working plan: Bulge-disk decomposition

70 Tempel et al M31: RT model?

71 M31: RT model? Working plan: Bulge-disk decomposition Prepare data products and first run Iteratively refine model Panchromatic SED fit -> main dust and SF properties.

72 M31: RT model? Working plan: Bulge-disk decomposition Prepare data products and first run Iteratively refine model Panchromatic SED fit -> main dust and SF properties.

73 M31: RT model? Working plan: Bulge-disk decomposition Prepare data products and first run Iteratively refine model Panchromatic SED fit -> main dust and SF properties. Full model & dust heating analysis???? Profit!