How do Galaxies Get Their Gas?

1. The EVLA Vision: Galaxies Through Cosmic Time Dec. 2008 How do Galaxies Get Their Gas? Dušan Kereš Institute for Theory and Computation - Harvard-Smithsonian CFA Collaborators: Neal Katz, David Weinberg, Romeel Davé, Mark Fardal, Lars Hernquist, T. J. Cox, Phil Hopkins

2. Active star formation occurs during most of the Hubble time Hopkins & Beacom 2006 What drives the star formation? Why is SFR density so high at early times and decreases at low-z?

3. What drives this star formation evolution? • Stars form from dense/galactic gas • Galaxies have very high gas fractions at high redshift and extended gas reservoirs -> can provide high SFRs. • Depletion of this gas can then slow down the star formation • Could explain some of the observed trends • Let’s check if there is enough “galactic” gas at high redshift

4. ** z=0 The state of dense gas • Damped Ly Absorbers (DLA) are probing column densities typical of the gaseous galactic disks • Amount of gas in DLA at high-z is less than the mass locked in stars at z=0. • A vast majority (~80%) of stars form after z=2 (e.g. Marchesini et al. ‘08) • During the time when most of the stars formed dense HI phase stays constant • Dense gas phase needs to be constantly re-supplied. Prochaska&Wolfe ‘08

5. Is there an evidence for gas supply in our neighborhood?

6. Milky Way • MW’s SFR 1-3M_sun/yr (Kennicutt ‘01) • MW’s gas reservoir is ~5e9M_sun . Without gas supply this is spent in several Gyr. • Star formation rate in the Solar neighborhood relatively constant. Not much change in gas density despite large gas depletion (Binney et al. 2000) • Supply of gas is needed

7. HVCs around MW • Most direct measure of galactic infall • Many clouds with inwards velocities, net infall from known clouds > 0.25M/yr • Hard to explain by the galactic fountain • Low metallicities • Similar net infall rates from clouds around other galaxies Van Woerden et al. 2004

8. Conclusion: Accretion of gas from the intergalactic medium is ongoing at all epochs.

9. The Theory

10. Standard Model • E.g. White & Rees 1978 • Gas falling into a dark matter halo, shocks to the virial temperature Tvir at the Rvir, and continuously forms quasi-hydrostatic equilibrium halo. Tvir=106(Vcirc / 167 km/s)2 K. • Hot, virialized gas cools, starting from the central parts, it loses its pressure support and settles into centrifugally supported disk –> the (spiral) galaxy. • Mergers of disks can later produce spheroids. • The base for simplified prescriptions used in Semi-Analytic models – SAMs (e. g. White & Frenk 1991). Tvir

11. NASA/WMAP Structure formation Gas heating Gas cooling Dynamics-merging Star-formation FEEDBACK Complex and non-linear: Simulations needed Earliest Observation How do galaxies form and evolve? Today NOAO

12. SPH simulations of the galaxy formation • We adopt CDM cosmology. • Gadget-2 and 3 (Springel 2005) SPH code: entropy and energy conservation; proper treatment of the “cooling flows” • Cooling (no metals), UV background, a star formation prescription • A simple star formation prescription • Star formation happens in the two-phase sub-resolution model: SN pressurize the gas, but does not drive outflows • Star formation timescale is selected to match the normalization of the Kennicutt law. • Typically no galactic winds • Largest volume 50/h Mpc on a side, with gas particle mass ~1.e8 M⊙, but most of the findings confirmed with resolution study down to 1.e5 M⊙ • Lagrangian simulations -> we have ability to follow fluid (particles) in time and space.

13. Global Accretion • We define galaxies and follow their growth • Galaxies grow through mergers and accretion of gas from the IGM • Smooth gas accretion dominates global gas supply • Mergers globally important after z=1 • Star formation follows smooth gas accretion, owing to short star formation timescales • It is within factor of ~2 from the typical Madau plot (e.g. Hopkins & Beacom 2006) Kereš et al. 2008

14. Tmax = ? Shocked IGM Temperature history of accretion • We utilize the Lagrangian nature of our code • We follow each accreted gas particle in time and determine its maximum temperature - Tmax before the accretion event. • In the standard model one expects Tmax ~ Tvir Galactic gas

15. 250000K Evolution of gas properties of accreted particles • - Empirical division of 2.e5K, but results are robust for 1.5-3e5K • - The gas that was not heated to high temperatures -> COLD MODE ACCRETION • - > Accretion from cold filaments • - The gas that follows the standard model -> HOT MODE ACCRETION • - > Accretion from cooling of the hot halo gas Kereš et al. 2005 Katz, Keres et al. 2002

16. How important are these two accretion modes? • Cold mode dominates the gas accretion at all times. • Hot mode starting to be globally important only at late times

17. The nature of cold and hot modes

18. Low mass halos are not virialized • Halo gas (excluding galactic gas), within Rvir • Low mass halos are not virialized • Transition: Mh=1011.4 M⊙ • Approximate description by Birnboim&Dekel 2003 • Post shock cooling times are shorter than compression time at Rvir when Mh~1x1011 M⊙ • More gradual transition in our case because cold filaments enhance the density profile Kereš et al 2008

19. Low mass halos Low mass halos • High-z well defined filamentary accretion • Low-z -> Tvir is lower, filaments are warmer and larger than the halo size

20. High-z accretion • Strong mass dependence of accretion • Cold mode infalls on a roughly free fall timescale. It follows the growth of DM halos in the lower mass halos and still is within a factor of ~3 in massive halos • High accretion rates result in high SFRs of high redshift galaxies • Even in massive halos cold mode filaments supply galaxies with gas • Cooling from the hot atmosphere not important. • Satellites accrete with the same rates as central galaxies Kereš et al. 2008

21. < 1kpc resolution, m_p ~1.e6 M_sun Kereš et al 2008

22. Zoom to 0.8Rvir • z~>2 cold mode can supply the central galaxy and satellite galaxies efficiency to supply galaxies directly declines with time • z~1 cold mode less efficient in massive halos; lower density contrast -> easier to disrupt • It can supply satellites directly but supply to the center is limited and often goes trough cold clumps

23. Low-z accretion Red - hot mode Blue - cold mode • Drastic accretion change over time, factor of ~30 from z=4 to 0. • At low mass this roughly follows the drop in the dark matter halo accretion • Some hot mode accretion around the transition mass. • Halos more massive than ~1012M⊙ (an order of magnitude above transitions) stop cooling hot gas • At z~0, a fraction of massive halos, around MW mass is able to cool the hot gas Kereš et al. 2008

24. Recent & future simulations

25. New simulations 5kpc/h physical Z~2.6 110/h pc physical res. M_p~1.e5M_sun M_h~1.3e11M_sun Yellow Tvir Blue ~ 1e4K 300/h kpc ~ 2Rvir 1.2/h Mpc cmv.

26. z~2.6 25/h kpc (physical)

27. Halo mass at z=0, 7x1011M_sun Gas particle mass ~105M_sun

28. Zoom onto a ~Rvir region, with the lowest overdensity moving from 200 to 1500

29. Can we detect the smoothly accreting gas with EVLA?(very preliminary …)

30. z=2.5 z=2.0 z=1.5 400kpc region, 1pixel=1kpc, approx 5” at 40Mpc Log n_HI 23 22 21 20 19 18 17

31. High-z • Filaments have n_HI ~ 1.e18-1.e20, and around 1.e11Msun halos widths of several tens of kpc. • Massive halos, around MW size are good targets to constrain the accretion models • This continues to about z=1 • Lower mass galaxies are embedded in thick filaments • Need to work on kinematic signatures

32. z=1.0 z=0.5 z=0 400kpc region, 1pixel=1kpc, approx 5” at 40Mpc

33. EVLA range z=0-0.5 • Hard to detect filaments • Their physical size and typical temperature larger at low redshift: 20 to 100 kpc scale. This results in low columns. • Galaxies in MW size halos are surrounded by the cold clouds with mixed origin • No more clear distinction of the cold and hot modes. Many clouds form from cooling instabilities in warm ~1.e5K gas in the outskirts of halos. Similar process occurs when filaments approach the virial radius. • Clouds are easily detectable once inside ~50kpc. A large number of clouds should be visible around external galaxies with n_HI >1.e18 cm-2 • Clouds form in much larger numbers if there are galactic winds operating. • Observations could place some constraints on the level of such feedback, cloud formation radii and cloud survival. • Clouds should be common around MW size galaxies that are central in halos. Less common in lower mass halos. • More work is needed for more massive objects.

34. Where should an idealized HI telescope look? • Higher redshift -> more coherent structure • Higher mass galaxies -> higher density contrast • Low mass galaxies, relatively more HI gas compared to the galaxy size • no clouds far from the galaxy, harder to detect extended halo • z=1 is already an interesting regime.

35. Summary • Low mass halosdo not undergo classic virialization • In more massive halos gas is heated to Tvir • The transition into the hot halos is gradual because of the filaments that enhance the density structure in a halo • At high redshift (z > 1.5) smooth gas accretion is completely dominated by the cold mode accretion, even in massive halos • Cooling from the hot atmosphere is inefficient at all times • At low redshift hot mode is important in a fraction of objects but the most massive halos accrete very little from the hot atmosphere • Accretion of cold clouds with mixed hot and cold mode origin dominates the late time accretion in MW size halos, but more work is needed to understand the physics of this process. • EVLA could probe the last epoch where coherent structures feed galaxies and should be great for studies of cold (HVC like) clouds.