Updated: 14 September 2011

My Research in a Large Nutshell *

by Kathy Cooksey

The cosmic enrichment cycle generically refers to the movement of metals from stars to the gas in and between galaxies and, possibly, back again (perhaps many times). The signature of this cycle is etched into the processed gas and reflected in its metallicity, elemental abundances, density, and/or spatial distribution. The study of the low-redshift (z < 1) IGM is the study of the last eight-billion years of cosmic chemical evolution. Quasar absorption-line (QAL) spectroscopy is a powerful tool for probing the gas in the large-scale structure of the Universe, otherwise known as the cosmic web. I have used QAL spectroscopy to address fundamental questions such as: where are the baryons today? How has the IGM evolved since z = 1? What is the relationship between galaxies and intergalactic absorbers?

Big Bang nucleosynthesis models, the cosmic microwave background, and the z ≈ 3 Lyα forest (e.g., O'Meara et al. 2006, Spergel et al. 2007) set strong constraints on the baryonic content of the Universe. However, the amount of baryons in stars, galaxies, and clusters at z = 0 is not enough to balance the baryon budget (Fukugita & Peebles 2004). The "warm-hot intergalactic medium" (WHIM) has been identified, in cosmological hydrodynamic simulations (e.g., Davé et al. 2001) as a likely reservoir of up to 40% of the z = 0 baryons. The WHIM is canonically defined to be diffuse (overdensities of 10 to 30), hot (5 < log T < 7), and collisionally ionized. All three properties make the WHIM difficult to observe since the best tracers are X-ray transitions and current technology is not sufficient.

However, the OVI doublet has been proposed as a good tracer. Oxygen is the most abundant metal; the OVI column density peaks at log T = 5.5 under the assumption of collisional-ionization equilibrium (CIE); and the doublet is readily observable in today's UV spectra from z = 0 to 0.5. There have been several large OVI surveys using spectra from the Far Ultraviolet Spectroscopic Explorer (FUSE) and the HST Space Telescope Imaging Spectrograph (STIS), including Danforth & Shull (2005, 2006), Thom & Chen (2008a,b) and Tripp et al. (2008).

In Cooksey et al. (2008), we presented a detailed analysis of the FUSE and STIS archival spectra of the sightline towards the low-redshift quasar PKS1302−102. Though the UV dataset was included in the aforementioned OVI surveys, we carefully modeled the absorption systems with any metals (not just OVI) with CLOUDY photoionization and CIE models. We also supplemented the QAL results with a survey of galaxies in the quasar field and characterized the absorber-galaxy connection. The metal absorption-line systems arose in a variety of galactic environments, where the separations and the galaxy properties (e.g., luminosity, spectral type) varied. However, there was indication that sub-L* galaxies were correlated with intergalactic metals, and this result has been confirmed and strengthened in our larger study of the absorber-galaxy connection (Prochaska et al., in preparation). In this work, we find that the properties of 0.02 < z < 0.2 metal-line absorbers can be reproduced well if sub-L* galaxies (explicitly, 0.1 L* < L < 1 L*) have an enriched gaseous halo of order 300 kpc.

From the PKS1302−102 study, we also concluded that none of the sightline's four OVI absorbers definitively traced the WHIM. Since the doublets were seen in association with the low-ionization state CIII absorption, the gas must be a multi-phase medium and not the canonical WHIM. This result agreed with other studies (e.g., Prochaska et al. 2004, Tripp et al. 2008) and with some current simulations (Oppenheimer & Davé 2009). There is a problem if the OVI absorbers often trace a multi-phase medium and the latter is already included in the z = 0 baryon budget through low-redshift Lyα forest observations, which say the forest holds about 40% of the z = 0 baryons. Then estimates that assume all OVI absorbers trace the WHIM and the Lyα forest traces the photoionized gas will double-count a significant fraction of the baryons in the local Universe.

I conducted the majority of the PKS1302−102 study, from the reduction of the UV spectra to the CLOUDY modeling, under close supervision of my thesis advisor X. Prochaska and following the template of Prochaska et al. (2004, 2006). I measured and applied the astrometric solution for the PKS1302−102 galaxy survey, as well as for the 18 other, similar surveys (described in detail in Prochaska et al. 2011). We knew at the time that I would be repeating much of the reduction and analysis techniques on the full set of HST archival UV spectra for my dissertation. Therefore, I developed the reduction and analysis software to also work for the survey for z < 1 CIV and SiIV doublets in the spectra of 49 low-redshift quasars (Cooksey et al. 2010, 2011, respectively). Our catalogs of CIV and/or SiIV systems are the largest low-redshift samples to date. Considering they incorporate 14 years of HST UV spectroscopy (pre-SM4), they will remain the largest for years, until the HST Cosmic Origins Spectrograph (COS; Morse et al. 1998), the repaired STIS, and, even, the World Space Observatory-UV (WSO-UV; Shustov et al. 2009) can observe a significant number of new quasars.

CIV and SiIV doublets have proven to be important tracers of the IGM and its evolution from z = 6 to 1.5 (e.g., Songaila 2001, 2005, Boksenberg et al. 2003, Scannapieco et al. 2006, Becker et al. 2009). These transitions are well-studied at high redshifts because: they are strong transitions of common metals; they are observable outside the Lyα forest, where they become easier to identify; they redshift into optical passbands at z = 1.5; and they are resonant doublets, which give them distinctive characteristics and enable the survey to be largely automated. However, CIV and SiIV absorbers at z < 1 trace the IGM and its evolution for last eight-billion years. Thus observations of these doublets at low redshift constrain the net effect of the cosmic enrichment cycle (i.e., cosmic star formation history and feedback processes).

We proved that the CIV and SiIV mass densities have increased slowly and steadily since z = 5, and new observations by D'Odorico et al. (2010) have increased the CIV statistics around z = 1.5 and supported our result. Previously, the high-redshift studies had shown that the ions' mass densities remained roughly constant for z = 5 to 1.5. It has been shown, in simulations, that the increasing metallicity of the Universe balancing the increasing ionization state of the IGM would result in a flat CIV mass density (Oppenheimer & Davé 2006). In Oppenheimer & Davé (2008), the CIV mass density increased from z = 1 to 0 due to the feedback from asymptotic giant branch (AGB) stars.

Interestingly, the CIV and SiIV line densities have remained roughly constant since z = 5. The increase in mass without the increase in numbers led us to propose that the doublets traced more circum-galactic than intergalactic gas at z < 1. Early results from our study of CIV and SiIV absorption in the (cosmological hydrodynamic) OverWhelmingly Large Simulations (OWLS; Schaye et al. 2010) confirmed that the simulated doublets were in the CGM and reproduced our observations reasonably well (see Cooksey 2009 and section 5 for more details).

In Cooksey et al. (2011), we found that the ionic ratio SiIV to CIV, as measured in our samples and in a 2 < z < 4.5 catalog (Boksenberg et al. 2003), did not evolve over the time probed. The ionic ratio is a function of the spatial distribution, abundance, and ionization state of silicon and carbon in the IGM. Our simple CLOUDY models showed that the UV background (UVB), which largely determines the ionization state, does not have to be soft (e.g., dominated by galaxies) to reproduce the observations. This may conflict with our interpretation that the z < 1 doublets trace gas close to galaxies, and I want to pursue this further with the OWLS in the future.

As was the case for Cooksey et al. (2008), I conducted the majority of the research for the CIV and SiIV surveys in close collaboration with Prochaska. C. Thom performed the Monte Carlo simulations to estimate the contamination rate of Lyman forest lines in the CIV survey, and I used his algorithms to do the same for the SiIV survey. For the OWLS study, I received significant help from S. Bertone, who is a simulator and works with OWLS, though I brought the expertise in spectroscopy and the observations.

A common theme is that the z < 1 metal-line systems detected in QAL spectra largely arise in gas associated with galaxies. In the PKS1302−102 sightline, the metal-line systems all had one or more L > 0.1 L* galaxies within a few hundred kilometers per second and kiloparsecs. It can be inferred that circum-galactic CIV and SiIV absorbers dominate the z < 1 observations from the fact that the line densities do not increase significantly from z = 5 to 0 while the mass densities do. In the OWLS, the majority of the CIV and SiIV absorbers are found to arise in gas with overdensities typical of galactic halos.


* This research narrative was originally composed for a faculty application with Stockholm University (January 2011), which is why there are these off-putting statements of what I did explicitly.