School of Physics & Astronomy Simulating the spectra of Quasars: A simple disk-wind model for BALQSOs Nick Higginbottom (Southampton University) Christian Knigge (Southampton University) Knox Long (STScI) Stuart Sim (Queens University - Belfast) James Matthews (Southampton University) Naples 21 st May 2013
School of Physics & Astronomy Overview The problem – BALQSOs, outflows and QSO unification A benchmark disk-wind model Physical state and synthetic spectra for benchmark model X-ray results and sensitivity Summary and the future Nick Higginbottom Modelling the spectra of (BAL)QSOs
School of Physics & Astronomy Elvis (2000) Gibson+ (2009) BALQSOs Blue shifted absorption features imply outflows at velocities of ≥ 0.1c (Weymann+ 1981) Continuum / emission features similar to other QSO types Common underlying structure? Disk-winds? BALQSO clearest indicator Reichard+ (2003) Nick Higginbottom Modelling the spectra of (BAL)QSOs
School of Physics & Astronomy BALQSOs – Evolution vs Orientation About 20% of the population of QSOs exhibit BAL properties (e.g. Knigge+ [2008] and Hewett & Foltz [2003]) Evolution: All QSOs spend 20% of the time as BALs Orientation BAL Outflows cover 20% of viewing angles Our Aims: Turn qualitative models into quantitative predictions using Monte Carlo radiative transfer code - PYTHON (Long & Knigge [2003], Higginbottom et al. [2013 in prep]) Produce something that looks like a BALQSO from some directions Can such models look like other types of QSO from other directions? Can the X-ray properties of such models be made to agree with observations? Nick Higginbottom Modelling the spectra of (BAL)QSOs
School of Physics & Astronomy P YTHON – a 3D ionization and radiative transfer code. Arbitrary 3D wind geometry Kinematic models Hydrodynamic simulations Monte Carlo radiative transfer Fast ionization calculations Modified Saha approximation Thermal / Radiative equilibrium heating/cooling: free free, line, Compton photoionzation, recombination Nick Higginbottom Modelling the spectra of (BAL)QSOs Validation of PYTHON (dots) vs CLOUDY (lines )
School of Physics & Astronomy A benchmark wind model Geometry based on Shlosman and Vitello (1993) Nick Higginbottom Modelling the spectra of (BAL)QSOs INPUT SPECTRUM M BH = 10 9 M M acc =5M yr -1 L x =10 43 ergs s -1 WIND PARAMETERS R min =300R G R max =600R G θ min =70° θ max =82° M wind =5M yr -1 V ∞ =V escape
School of Physics & Astronomy Nick Higginbottom Modelling the spectra of (BAL)QSOs Properties of benchmark model
School of Physics & Astronomy Nick Higginbottom Modelling the spectra of (BAL)QSOs Properties of benchmark model
School of Physics & Astronomy Nick Higginbottom Modelling the spectra of (BAL)QSOs Properties of benchmark model
School of Physics & Astronomy Nick Higginbottom Modelling the spectra of (BAL)QSOs Properties of benchmark model
School of Physics & Astronomy Properties of benchmark model
School of Physics & Astronomy Predicted Spectra – 40° Nick Higginbottom Modelling the spectra of (BAL)QSOs (Weak) thermal / scattering emission lines Slight continuum enhancement due to electron scattering Continuum without wind Disk-wind spectrum
School of Physics & Astronomy Predicted Spectra – 75° Nick Higginbottom Modelling the spectra of (BAL)QSOs Sightline into wind cone Strong BAL features Continuum without wind Disk-wind spectrum Attenuated continuum
School of Physics & Astronomy Predicted Spectra – 85° Nick Higginbottom Modelling the spectra of (BAL)QSOs Sightline through base of wind Emission lines appear brighter due to attenuated continuum Continuum without wind Disk-wind spectrum Attenuated continuum
School of Physics & Astronomy X-ray properties of benchmark model Benchmark model Figure from Saez Nick Higginbottom Modelling the spectra of (BAL)QSOs
School of Physics & Astronomy Increasing L X 1.Increasing X-ray luminosity destroys the BAL Mass loss rate through wind X-ray luminosity Nick Higginbottom Modelling the spectra of (BAL)QSOs Subplot scales -3x10 9 velocity (cms -1) +3x10 9 flux
School of Physics & Astronomy 1.Increasing X-ray luminosity destroys the BAL 2.Increasing the mass loss rate gets it back! Mass loss rate through wind X-ray luminosity Nick Higginbottom Modelling the spectra of (BAL)QSOs Subplot scales -3x10 9 velocity (cms -1) +3x10 9 flux Increasing L X and M wind
School of Physics & Astronomy 1.Increasing X-ray luminosity destroys the BAL 2.Increasing the mass loss rate gets it back! 3.For L x = 2x10 44 ergs s -1 need M wind =20M yr -1 4.M wind =4M acc OK?? Mass loss rate through wind X-ray luminosity Nick Higginbottom Modelling the spectra of (BAL)QSOs Subplot scales -3x10 9 velocity (cms -1) +3x10 9 flux Increasing L X and M wind...
School of Physics & Astronomy X-ray properties of high L x M wind model Figure from Saez Nick Higginbottom Modelling the spectra of (BAL)QSOs Benchmark modelObservable values for L 2-10keV =2e44ergs s -1, 25 solar mass per year mass loss
School of Physics & Astronomy Summary and plans for the future We have produced a simple disk-wind BAL model Higginbottom et al. (2013 in prep) Correct ionization state Strong BAL features X-ray weak Weak line emission The next steps Explore parameter space Uniqueness? X-rays? Emission? Investigate hydro-models - Higginbottom, Proga et al. (2013 in prep) Nick Higginbottom Modelling the spectra of (BAL)QSOs Data from Proga & Kallman 04
School of Physics & Astronomy Thanks! Nick Higginbottom Modelling the spectra of (BAL)QSOs
School of Physics & Astronomy Making CIV in the wind – lessons learnt Photoionization absorption in the root of the wind attenuates UV
School of Physics & Astronomy Making CIV in the wind – lessons learnt Easier to make CIV with a larger BH:
School of Physics & Astronomy Varying X-rays Fiducial model is X-ray weak Input α OX varies from -2.4 (pole on) to -1.8 (edge on) Emergent α OX much lower due to X-ray absorption in the wind for BAL sightlines