Adam Smercina
University of Toledo

REU Program (Summer 2014)
University of Wisconsin - Madison
Madison, WI 53706

adam.smercina@rockets.utoledo.edu

Lowell Observatory - Flagstaff, AZ

Home

Spectra


Just After THE FALL:
A Sample of E+B Post-Starburst Galaxies at z ~ 0.5


Adam Smercina & Christy Tremonti.




1. Introduction

Post-starburst galaxies represent a unique phase of galaxy evolution. Discovered in high-redshift surveys, post-starbursts are characterized by deep balmer absorption features and nearly uniform stellar populations, which indicate a rapid truncation of star formation sometime in the not-too-distant past (Dressler & Gunn 1983; Zabludoff et al. 1996). Many different mechanisms have been proposed as pivotal in the quenching process. The two most supported of these theoretical mechanisms are stellar and Active Galactic Nucleus-driven (AGN) feedback (Springel et al. 2005). Evidence of both phenomena has been observed and, in certain cases, both have been shown to be capable of quenching the star formation of their host-systems on short timescales (Tremonti et al. 2007; Alatolo et al. 2014). It is likely that a combination of these processes results in the rapid star formation quenching which ultimately results in a post-starburst phase. The majority of starburst events are thought to be merger-induced, ranging from moderately gaseous to truly gas-rich, often referred to as the "wetness" of the merger (Barnes & Hernquist 1991). The former appear to be the progenitors of local post starbursts, while the latter abound at higher redshift and are presumed to result in the more massive, high-redshift post-starbursts (Sell et al. 2014). The quenching mechanisms which dominate the wettest mergers are extremely efficient, capable of truncaing the starubrst in < 100 Myr.

To date, merger evolution research has concentrated on the end extremes of the evolutionary process: the starburst event and the A-star dominated post-starburst phase (E+A). The post-starburst evolutionary model, when applied to intermediate stages, predicts a B-star dominate epoch. B-star dominated post-starburst systems would be much younger than the classic E+A galaxy and would likely exhibit strong merger signatures, both morphological and spectral, while still hosting the definitive balmer absorption features and prominent 4000Å break present in E+A's. Such objects could provide an unprecedented window into the evolutionary transition from starburst to quiescent system. We have derived a sample of 23 B-star post-starbursts (E+B's). In this study we endeavor to probe the star formation histories of these as-yet unstudied E+B's and provide a compelling case for further observations.

2. Sample

This sample of E+B's was derived from the SDSS-III BOSS DR12 catalogue, whih consists of ~1.5 million objects. The selection process was divided into three parts: (1) The application of broad photometric, spectroscopic, and redshift constraints, (2) The fitting of a grid of 2003 Bruzual & Charlot simple stellar population models (SSP; BC03) (Bruzual & Charlot 2003) to the sub-sample obtained in the previous step and placing output constraints, and (3) The hand-inspection of each object contained in the small sample of resulting objects.

2.1 Broad Criteria

The broad selection process consisted of criteria derived directly from the DR12 data. A first cut was made by limiting the redshift search range to 0.35<z<1.1. The low-end constraint was made based on science interests, while the high-end was put in place with the spectral range of BOSS in mind. Next, the g-band magnitude was limited to 20.5 in an effort to exclude objects too dim for our study. On the spectroscopic end, the [O II] 3727Å equivalent width (EW) was limited to < 15Å in order to exclude excessively star-forming objects. Lastly, the 'ZWARNING_NOQSO' a data tag used to eliminate redshift errors due to quasars. This tag was constarined to 0, which eliminated all but the most elusive quasar host galaxies from the sample. The resulting sub-sample contained ~10,000 galaxies.

2.2 BC03 Fitting

We constructed an age-separated grid of SSPS's from the BC03 library, with ages of: 1 Myr, 4 Myr, 25 Myr, 50 Myr, 101 Myr, 286 Myr, 904 Myr, 2 Gyr, and 5.5 Gyr. Each spectrum was fit with a linear combination of the BC03 grid inputs, re-sampled using the velocity dispersion of each object. Reddening is treated as a free parameter, using the Charlot & Fall Law (2000). The fitting procedure returns both the light fraction used from each model and TV, which is defined as the total V-band extinction, A(V) × 1.086. We then grouped the light fractions into 4 categories: (1) the 1 Myr template; very young, (2) the 4 and 25 Myr templates; young, (3) the 50, 101, and 286 Myr templates; medium, and (4) the 904 Myr, 2 Gyr and 5.5 Gyr templates; old. The light fractions were constrained by limiting the young fraction S/N to > 2.2 and setting the medium and old fractions to > 0 and < 0.65, respectively. A main contributor to sample contamination was the presence of BL Lac objects, mostly red galaxies with underlying variable AGN that boost the blue continuum. In light of BL Lacs' strange spectral shape, we surmised that the BC fitting would attribute nearly all of their flux to only the very youngest and oldest age templates. To eliminate them from the sample, we restricted the 1 Myr + 2 Gyr + 5.5 Gyr light contribution to < 0.75. Using the BC03 fit for each object, a routine was used to fit emission lines to the continuum-subtracted spectra. The line widths of the fit were constrained to < 400 km/s, in an effort to eliminate broad-line AGN. The final sample was constructed by hand-inspection of the resulting 77 objects output by the previous steps.

3. Results

The results of this project are forthcoming, but the preliminary results are as follows:

3.1 The Star Formation History

Forthcoming.

3.2 WISE Photometry

A search of the WISE database revealed that 22 objects had been observed in the WISE All-Sky survey. Of those objects observed, only 9 possessed a S/N > 1.5 in all 4 WISE filters. The spectra and SDSS photometry shows a decrease, or leveling off, of the spectral energy distribution (SED) after the z band. However, the WISE photometry show a sharp increase of the SED's as they cross into the infrared. The height of the W4 (22 μm) band most likely signifies the presence of significant polycyclic aromatic hydrocarbon (PAH) emission. These small, complex dust grains trace many fascinating properties of the state of the interstellar medium (ISM).




References

Alatolo, K., et al. 2014, ApJ, 780, 186

Barnes, J., Hernquist, L. 1991, ApJ, 370, L65

Bruzual, G., Charlot, S. 2003, MNRAS, 344, 1000

Dressler, A., Gunn, J. 1983, ApJ, 270, 7

Sell, P., et al. 2014, MNRAS, 441, 3417

Springel, V., et al. 2005, MNRAS, 361, 776

Tremonti, C., et al. 2007, ApJ, 663, L77

Zabludoff, A., et al. 1996, ApJ, 466, 104