Stochas(c Stellar Feedback in Low-Mass Galaxies Chris Power, ICRAR/UWA with Lilian Garra,-Smithson, Mark Wilkinson, Graham Wynn (U Leicester), Mar<n Bourne (KICC Cambridge) The Role of Gas in Galaxy Dynamics, Malta, October 2017 1
What do we mean by stochas(c feedback? Condi<ons for massive stars capable of producing poten<ally • disrup<ve feedback - UV photo-ioniza-on, strong winds, supernovae, X-ray binaries - not determinis(c. Effects most likely to be pronounced in low-mass galaxies, reflec<ng; • Inefficient star forma<on & triggering by perturba(ons . • Low masses of star forming regions & random sampling of ini(al • mass func(on regulates numbers of massive stars. Inefficient stellar feedback – sites of star forma<on biased • towards central regions, sensi(ve to local condi(ons (e.g. gas density, cooling, radia<on field, etc..) Forma<on paths of binaries into energe<c (high) mass X-ray • binaries subject to complex evolu(on 2
What is the impact of stochas(c feedback? 1. Low masses of star forming regions & random sampling of ini(al mass func(on regulates numbers of massive stars. 2. I nefficient stellar feedback – sites of star forma<on biased towards central regions, sensi(ve to local condi(ons (e.g. gas density, cooling, radia<on field, etc..) 3. Forma<on paths of binaries into energe<c (high) mass X-ray binaries subject to complex evolu(on How does this stochas(city impact the abundance and star forma(on histories in low-mass galaxies? 3
M31 • Occupancy of low-mass dark ma,er sub- haloes with satellite galaxies apparently stochas(c (e.g. Boylan-Kolchin et al. 2011) • Power et al 2014 - result of stochas<c star forma<on and feedback?
Stochas(c Feedback in Low-Mass Halos I Essen(al Idea NGC1569 • Binding energy of gas comparable to a single supernova in halo with 10 8 -10 9 M ¤ Star forma<on proceeds in random • order in molecular clouds. • Pre-MS <mescale varies as M -2.5 – high mass stars form more rapidly than lower mass counterparts. • How does this influence stellar popula<on that forms…? Power et al. 2014 Credit: ESA/NASA 5
Stochas(c Feedback in Low-Mass Halos II Power et al. 2014 Model Redshi> when cooling first supported. • Monte Carlo Merger Trees for 10 7.7 M ¤ distribu<on of progenitors of present-day low-mass halos. • M vir =10 7.7 -10 10 M ¤ at z=0 10 10 M ¤ • When were they first massive enough to support cooling , i.e. T vir ~10 3 -10 4 K? • Es<mate binding energy of gas in halo progenitor, and energy liberated by supernovae, assuming stochas-c high mass star forma-on. 6 0 10 20 30 40
Stochas(c Feedback in Low-Mass Halos III Late forming, low-mass Power et al. 2014 haloes at present day Note spread in z cool Note spread in f bind Log 10 f bind =E SN /E bind Reioniza(on Complete P(Z cool ) 0 10 20 30 40 0 10 20 30 Z cool Z cool More massive subhalos at z=0 support cooling earlier – feedback can have greater effect, but more <me to re-accrete gas and form stars – 7 effects of stochas(city more pronounced for lower mass subhalos .
Locally Regulated Feedback Efficiency I Assump(on : Stellar-driven oujlows (winds, supernovae) sufficiently energe<c to expel ambient gas from the galaxy. Caveat: How oujlow couples to ambient gas is as important as energy and momentum carried by the oujlow. Test: Use hydrodynamical simula<ons to explore coupling of stellar wind driven oujlow from a nuclear star cluster couples to gas. From Bourne & Power 2016 Feedback from mul-ple sources superposes, seeds dense clumps in the ouAlow, 8 frac-on of mass expelled lower than one might naively expect.
Locally Regulated Feedback Efficiency II From Bourne & Power 2016 Radial Inflow Single Source Mul(ple Sources Problem: MulEple sources less efficient at clearing out gas from nuclear regions – seeds dense clumps robust to feedback – 9 available to grow stars, grow central black hole?
Locally Regulated Feedback Efficiency III From Bourne & Power 2016 Radial Inflow Efficient cooling Inefficient cooling Problem: Cooling efficiency will influence rapidity with which gas clumpiness is seeded – and therefore efficiency with which gas is expelled from the poten<al. 10
High Mass X-Ray Binaries (XRBs) Sana et al. 2012 • Binary frac<on for massive stars close to unity; some uncertain frac<on survives to form high mass X-ray binaries (e.g. Sana et al. 2012). • Compact object accretes from main sequence companion via wind capture or Roche Lobe Overflow. • Accre<on liberates energe<c radia<on with luminosity XRBs poten(ally important stochasEc source of feedback? 11
HMXB Photo-Ionizing Feedback I From Power et al. 2009 Explore ionizing output – UV • and X-ray - from a coeval stellar popula<on as a func<on of IMF (e.g. Kroupa, Chabrier) • HMXB survival frac<on • HMXB spectra (hard to sop) • Can extend photo-ionizing • life(me of star cluster up to >80 Myrs…. … but photo-ionizing feedback • is non-local in nature. 12
HMXB Photo-Ionizing Feedback II From Power et al. 2009 From Power et al. 2013 Ionizing power depends on hardness of spectrum (Power et al. 2009) – power-law spectra usually assumed, but more realis<c empirical spectra have quite different 13 shape ( Power et al. 2013 ). Energy available for ioniza(on indicates importance…!
HMXB Kine(c Feedback I Cygnus X-1 - stellar mass black • hole fed by wind accre<on. Gallo et al. 2005 - amount of • energy in form of kine<c feedback (i.e. jet) as high as photo-ionizing feedback (i.e. X-ray luminosity). Expect outbursts to occur over • an extended period of <me. Jet inflates a lobe that is over-pressured rela<ve to its surroundings, driving a 14 thermalising shock Gallo et al. 2005
Lilian Garrah-Smithson Final year PhD student HMXB Kine(c Feedback II Hydrodynamic simula<ons to explore impact of kine(c feedback from HMXBs in From Garrah-Smithson, Wynn, & Power, In Prep addi<on to supernovae, in isolated star-forming clouds Gradual injec-on of energy acts to open up chimneys, releases thermal pressure, alters star forma-on rate. 15
HMXB Kine(c Feedback III Inherently stochas<c • Single ULX outburst • powerful enough to unbind gas in a dwarf galaxy Loca(on, loca(on, • loca(on…. • Can sweep galaxy free of gas… … or triggers star forma<on. • Power & Bourne, In Prep 16
Summary Sound physical reasons to expect that effects of stochas(c feedback • should be especially important in low-mass galaxies . Not captured in current galaxy forma<on models – a physics and numerical • resolu<on problem. Necessary to reconcile theory and observa<on? How energy couples to ambient gas can drama<cally impact • efficiency of gas explusion and influence subsequent star forma<on – very sensi<ve to local condi<ons Photo-ioniza<on? Gas frac<on? Instabili<es? • High mass X-ray binaries – whose forma<on is effec<vely stochas(c - • can release significant amounts of energy, providing non-local feedback and powerful kine<c oujlows. Drive turbulence in gas rich galaxies at high redship? Suppress collapse of gas • onto low-mass halos? 17
Recommend
More recommend
