A Preponderance of Metals in the Circumgalactic Medium Jessica Werk IMPS Postdoctoral Fellow at UC Santa Cruz Santa Cruz Galaxy Workshop, August 16th 2012

This Work Made Possible By: whom wikipedia says is one of the world’s most powerful mutant scientific minds. J. Xavier Prochaska Jason Tumlinson Chris Thom

The COS-Halos Team Spitzer proposes a hot, extraplanar Sembach Spitzer proposes a hot, extraplanar “Galactic Corona” 1956 Spitzer’s RAND Memo Proposes “LST” 1946 Quasars Discovered Schmidt & Oke 1963 David Weinberg Romeel Davé Amanda Ford Neil Katz Kenneth Sembach Ben Oppenheimer Molly Peeples thom, tumlinson, tripp, meiring, o’meara, werk, prochaska, dave, oppenheimer, katz, weinberg, peeples, ford, sembach Joe Meiring John O’Meara The COS-Halos Team Todd Tripp

The Circumgalactic Medium (CGM) Diffuse gas, including metals and dust, often extending to 300 kpc, and largely bound to the dark matter halo The CGM is loosely defined as gas surrounding galaxies within their own halos of dark matter (out to 100- 300 kiloparsec),the structure of the CGM and its relation to galaxy properties are still uncertain. In this conference, it’s been refered to as both the “new frontier” (Joss Bland Hawthorne) and the WHIM corona. (Filipo Fraternali) It lies at the nexus of accretion and outflows, The CGM may also reflect the theoretically-predicted transition from filamentary streams of cold gas that feed low mass galaxies to hot, quasi-static envelopes that surround high mass galaxies. Both outflow and accretion through the CGM may be intimately connected to the observed dichotomy between blue, star-forming, disk-dominated galaxies and red, passively evolving, elliptical galaxies with little or no star formation.

Thus, we need a new, well-designed approach to study this important medium that is ubiquitous around galaxies. A complication: Circumgalactic galaxy halo gas is too diffuse to be studied in emission, and a random sightline through the IGM is intercepted by <1 massive galaxy halo.

Absorption Line Experiments Method A: Find absorber in spectrum, go hunting for a galaxy at the proper redshift Method B: Know redshifts of nearby galaxies in projection, go hunting for absorption in the spectrum at those redshifts z ~0.2 z ~0.8 z ~ ?

Statistically Sampling the CGM of L* Galaxies Background light source (QSO) “COS-Halos” The CGM PI = Jim Green, U. of Colorado Installed by John Grunsfeld & Drew Feustel on SM4, May 16, 2009 Optimized for UV spectroscopy, R = 2000 and R = 18000 gratings, low-background photon-counting detectors. FUV Channel: Effective area 10 - 20x that of STIS over 1150-1800 Å. 39 QSO sightlines in 134 HST orbits (17 “red and dead”, 34 star-forming galaxies)

Sightline Map 150 kpc 100 kpc 50 kpc Make sure to mention superiority and originality of COS-Halos – it’s a first of its kind. obtained a uniform dataset of multiphase ions for a homogeneous sample of 51 ~L* galaxies lying in front of bright SDSS+GALEX QSOs. Unlike most QSO-gals studies, COS-Halos mined SDSS for a particular sampling of galaxy properties (log M* = 10 - 11, R < 150 kpc for a full range of color/type), rather than observing blindly a set of bright QSOs and obtaining galaxy coincidences post facto. The COS-Halos team is building a statistical map of the physical state and metallicity of gaseous galaxy halos using the Cosmic Origins Spectrograph (COS) on the Hubble Space Telescope (Tumlinson et al. 2011). Specifically, we are surveying the halo gas of 51 Sloan Digital Sky Survey (SDSS) galaxies (z = 0.15–0.35) well inside their virial radii (with impact parameters ρ < 150 kpc) COS-Halos has spectral coverage that allows us to detect absorption from a range of multiphase ions: OVI (Tumlinson et al. 2011), SiIV, SiIII, SiII, CIII, CII, NIII, NII (this work). We combine Keck LRIS and Magellan MagE spectra with the COS-data to derive galaxy properties such as redshifts and star formation rates (Werk et al. 2012) We additionally obtain HIRES spectra of the COS-Halos QSOs to cover absorption from MgI and MgII (This work).

The CGM of z ~ 0, L* Galaxies: The HI Gas A cool (104 K) medium with high covering fraction of NHI > 1015 cm-2 exists around nearly every L* galaxy, even ellipticals, to 150 kpc. Thus, there is no obvious suppression of “cold accretion” around massive elliptical galaxies.

The CGM of z ~ 0, L* Galaxies: The Low Ionization State Metals Low-Ion metals (Mg II) are present throughout the CGM, and have 50% covering fraction, to 70 kpc There is no obvious distinction between blue and red galaxies These low-ions seem to trace high NHI ( > 1016 cm-2)

The CGM of z ~ 0, L* Galaxies: The Intermediate Ionization State Metals Intermediate ionization state metals (SiIII, CIII) are very common throughout the CGM, and have 70% covering fraction to 150 kpc (90% for CIII). There is no obvious distinction between blue and red galaxies. There is a likely trend of decreasing column with impact parameter.

The CGM of z ~ 0, L* Galaxies: The Higher Ionization State Metals OVI is more common around star-forming galaxies than around massive, red, ellipticals. (SF = 90%; Red = 40%) There is a likely trend of decreasing column with impact parameter.

What is the origin of the CGM? Accreting IGM gas? Cold flows? Cooling Coronal Gas? Extended HI disks? Supernovae Winds? AGN feedback? Galactic Fountains? Tidal debris? Ram-stripped material? All of the Above!

The Baryonic Content of the CGM of z~0, L The Baryonic Content of the CGM of z~0, L* galaxies Warm and Cool Phases

The Galaxy Missing Baryon Problem Anderson & Bregman 2010: Field spirals are observed to be missing their baryons, if you consider the content of the hot X-ray halo and the stellar content alone. Klypin, Zhao, and Somerville 2002: Dynamical models unequivocally require that 50% of the gas within the virial radius of a galaxy must not be within the disk or the bulge. But, perhaps they didn’t fall in to begin with? Davé et al. 2009 say that most are in the WHIM, outside of galactic halos. Galaxies are severely Baryon-depleted relative to the cosmological paradigm. Most simulators say they are embedded in the hot halo, but there are good observational constraints that say otherwise.

Theoretical Prediction “The simulations of individual galaxies predict that the ‘missing baryons’ are found in the CGM of galaxies.” – Stinson et al. 2011 “We further note that the HI surface densities at < 50 kpc are sufficiently large that the CGM will yield significant absorption from lower ionization states of heavy metals (e.g. MgII, SiII, SiIII). A proper estimate of the column densities for these ions, however, will require a full treatment of radiative transfer (i.e. to account for self-shielding by optically thick HI gas).” At odds with the Bregman observation, somewhat...

. . . then apply ionization correction fOVI. . . The CGM of z~0, L* Galaxies: The Highly Ionized Phase Mass Budget (Tumlinson+11) MOVI = πR2 NOVI 16mH M⊙ . . . then apply ionization correction fOVI. . . MOxygen = 1.2 x 107 (0.2/fOVI) M⊙ Mgas = 2 x 109 (Z⊙/Z) (0.2/fOVI) M⊙ HM01 Background+CIE O VI is a fragile ionization state, which never exceeds a fraction fOVI = 0.2 of the total oxygen for the physical conditions of halo gas and is frequently much less abundantwhere we have taken a typical ⟨NOVI ⟩ = 10 14.5 cm−2 and R = 150 kpc, and the hit rate cor- rection fhit computed separately in three 50 kpc annuli (Figs. 1 and 2). This mass of oxygen is strictly a lower limit because we have scaled to the maximum fOVI = 0.2 (Fig. 4). Thus: The large oxygen mass (a lower limit!) in the ionized halos of galaxies implies at least 1 Gyr worth of star formation and oxygen yield, efficiently transported out to at least 150 kpc. R = 150 kpc

(Almost) As Much Oxygen in the CGM as in the ISM! >10% >70% The minimum CGM oxygen mass is thus 10-70% of the ISM oxygen (Figs. 4 and S4). The covering fractions and column densities we find for star-forming galaxies are insensitive to M ∗ , while the ISM metal masses decline steeply with M∗ according to the mass-metallicity relation. The ISM oxygen comes from the mass-metallicity relation. For the densities typically expected at radii R ∼ 100 kpc, fOVI exceeds 0.1 only over a narrow temperature range 105.4−5.6 K, and it only exceeds 0.02 over 105.2−5.7 K (Fig. 4). Either a large fraction of CGM gas lies in this finely tuned temperature range — a condition that is difficult to maintain because gas cooling rates peak at T ≈ 10 5.5 K — or the CGM oxygen and gas masses are much larger than the minimum values we have quoted above. Lower density, photoionized gas can achieve high fOVI ∼ 0.1 over a wider temperature range, but at these low densities it is hard to produce a 1014.5 cm−2 column density within the confines of a galactic halo, especially if the metallicity is low (Fig. S5). Thus fOVI = 0.02 and Z = 0.1Z ⊙ are plausible conditions for the O VI-traced gas, but it is unlikely that both conditions hold simul- taneously. Tumlinson+11 plot by Molly Peeples Does the decline in CGM to ISM oxygen mass ratio imply more efficient metal escape from low-mass galaxies? Or just some mass dependence to the ionization conditions?

The Simulations at z ~ 2.8, Courtesy of Sijing Shen This is one of the "Eris2" zoom-in cosmological simulations, it uses a blastwave supernova feedback model which drives large galactic outflows and enrich the CGM. The simulation is currently at z = 2.8, it has a high SFR ~ 18 Msun/yr and metallicity 12 + log(O/H) = 8.45.  

Lower Limit: Cool, Ionized CGM ( 104 K < T < 105 K) MSiIII = CfπR2 NSiIII 28mH M⊙ . . . then apply ionization correction fSiIII. . . MSilicon = 5.5 x 105 (0.7/fSiIII) M⊙ Mgas > 8 x 108 (Z⊙/Z) (0.7/fSiIII) M⊙ Here is the first thing to consider:  At whatU does Si++ maximize, which means requiresthe smallest correction to total Si?The attached plot shows it is near log U = -3and increases a bit as N_HI goes above 10^17.And, it reaches nearly 70% or so.

Photoionization Modeling Single Phase Vanilla EUVB (HM 2011) model. Have run including SF. Tends to lower the metallicity for lower HI column density, which drives mass estimates upward.

The Ionization State and Metallicity of the CGM On all plots, symbol sizes are inversely proportional to the derived range of values (i.e. larger points are known with higher accuracy), and blue squares indicate absorbers paired with star-forming galaxies while red diamonds indicate absorbers paired with elliptical galaxies at the same redshifts. Werk+12b

The Mass Surface Density of Hydrogen : Hydrogen Mass Surface Density as a function of Impact parameter for the galaxy-absorber pairs in our sample for which we are able to derive U and Z solutions (25 / 50). The curve shown is for an assumed surface density profile that decreases with distance as 1/ (r0 + r)2. We also derive a total mass for the mean NH value assuming the CGM fills to 150 kpc (the limit of the COS-Halos survey). Werk+12b

Next Steps Extend the analysis to a diverse set of galaxies and environments: dwarfs (PI Tumlinson, HST Cycle 18); Red galaxies, groups, clusters; Map ρ = 200 – 500 kpc (Tripp+12) Improve the HI measurements since they are critical for mass and metallicity determinations Improve imaging over what is provided by SDSS at z ~ 0.2 Integrate kinematic information Direct Imaging of the CGM?

The CGM is an Important Reservoir of Baryons And can account for much of the missing mass! A cool, metal-enriched CGM traced by HI and lower ionization states of metal lines (MgII, SiII, SiIII, CII, CIII) is pervasive out to 300 kpc A second, highly-ionized phase traced by OVI is ubiquitous around star-forming galaxies, but absent around passively evolving elliptical galaxies Both high and low ionization phases in the halos of star-forming galaxies contain > 107 M⊙ of oxygen These metal masses are comparable to the entire ISM of the Galaxy Combined CGM (Warm + Cool Phases) can account for 50+% of a galaxies baryons, and may resolve the galaxy missing baryon problem The less ionized gas contains a similar mass in metals. These observations, and their alignment with the predictions from new self-consistent models, provides strong support for a vigorous baryon cycle in which outlows and subsequent cooling of halo gas plays a key role in forming disc galaxies.