Building the red sequence through gas-rich major mergers
aa r X i v : . [ a s t r o - ph . C O ] O c t Stellar populations, planning for the next decadeProceedings IAU Symposium No. 262, 2009A.C. Editor, B.D. Editor & C.E. Editor, eds. c (cid:13) Building the red sequence through gas-richma jor mergers
Vivienne Wild , C. Jakob Walcher , Peter H. Johansson Institut d’Astrophysique de Paris, CNRS, Universit´e Pierre & Marie Curie,UMR 7095, 98bis bd Arago, 75014 Paris, France. email: [email protected] European Space Agency, Keplerlaan 1, 2200AG Noordwijk, The Netherlands Universit¨ats-Sternwarte M¨unchen, Scheinerstr. 1, D-81679 M¨unchen, Germany
Abstract.
Understanding the details of how the red sequence is built is a key question in galaxyevolution. What are the relative roles of gas-rich vs. dry mergers, major vs. minor mergers orgalaxy mergers vs. gas accretion? In a recent paper (Wild et al. 2009), we compare hydrodynamicsimulations with observations to show how gas-rich major mergers result in galaxies with strongpost-starburst spectral features, a population of galaxies easily identified in the real Universeusing optical spectra. Using spectra from the VVDS deep survey with < z > = 0 .
7, and aprincipal component analysis technique to provide indices with high enough SNR, we find that40% of the mass flux onto the red-sequence could enter through a strong post-starburst phase,and thus through gas-rich major mergers. The deeper samples provided by next generationgalaxy redshift surveys will allow us to observe the primary physical processes responsible forthe shut-down in starformation and build-up of the red sequence.
Keywords. galaxies: stellar content, starburst, interactions; surveys
1. Introduction
Recent observations have revealed that since a redshift of around unity the total massof stars living in red sequence galaxies has increased by a factor of two (e.g. Bell et al.2004). At the same time, the stellar mass density of the blue sequence has remainedalmost constant. The interpretation is that some blue galaxies migrate onto the redsequence after the quenching of their star formation, whilst the remainder continue toform new stars (e.g. Faber et al. 2007).Arnouts et al. (2007) measured the net mass flux which has taken place from the bluesequence to the red sequence. This amounts to 9 . × − M ⊙ /yr/Mpc , or about 1 . × M ⊙ /yr in the VIMOS VLT DEEP Survey (VVDS) volume. Recent star formationhistory can be used as a tool to identify galaxies as they enter the red sequence. Fortypical observable times of post-starburst features in VVDS optical galaxy spectra of ∼ . − . / M ⊙ )= 10 .
2. Post-starbursts in major merger simulations
Alongside toy models of star formation histories, modern advanced simulations afforda new means of interpretation of observational results. Smoothed particle hydrodynamic(SPH) simulations of galaxy mergers are now carried out routinely by several groups,11920 Wild et al.
Figure 1.
From left to right snapshots of a hydrodynamic simulation of a major galaxy merger,at 0.5 Gyr (initial starforming disks), 1.64 Gyr (peak of starburst), 2 Gyr (peak of post-starburstfeatures) and 3 Gyr (remnant) after the start of the simulation. From top to bottom: the x − y positions of gas particles and black holes (purple points); trace of combined star-formation rateof both galaxies; the optical continuum spectrum in 4000˚A break region (note that emissionlines, have not been plotted to improve clarity); the trace of PCA derived indices that describethe strength of the 4000˚A break and the Balmer absorption lines; the trace of two indices usedtraditionally to infer the recent star formation history of galaxies. uilding the red sequence δ A absorption line iscompared with our new method which combines all 5 Balmer lines visible in the 4000˚Abreak region, together with the changing shape of the continuum using a Principal Com-ponent Analysis based on the method described in Wild et al. (2007). As the mergercommences, the sudden presence of hot O and B stars cause a rapid increase in theblue-UV continuum causing a decrease in the measured 4000˚A break strength. Within ∼ < yrs) and the burst mass fraction must be large ( > −
3. VVDS post-starburst galaxies
The VVDS is a deep spectroscopic redshift survey, targeting objects with apparentmagnitudes in the range of 17 . < I AB <
24. The survey is unique for high-redshiftgalaxy surveys in having applied no further colour cuts to minimize contamination fromstars, yielding a particularly simple selection function. The spectra have a resolution (R)of 227 and a useful observed frame wavelength range, for our purposes, of 5500-8500˚A.In this work we select 1246 galaxies with 0 . < z < . I AB <
23 and per-pixel S/N > ∼ yrs) in comparison to the age ofthe galaxy (thus distinguishing them from more ordinary “starforming” galaxies). Theorange crosses indicate the galaxies with strong Balmer absorption lines, indicative ofa recent sharp shut-down in starformation. The encircled points indicate those galaxieswith no ongoing star formation, as measured from a fit to their multi-wavelength spectralenergy distribution (Walcher et al. 2008). For comparison to other E+A samples, thesegalaxies also have no visible [OII] λ Figure 2.
Left: the distribution of PC1 vs. PC2 for all galaxies in our VVDS sample. PC1is equivalent to the well-known index D n < − /yr are circled. Right: the mass distribution of galaxies, before(left) and after (right) correction for survey incompleteness effects. Histograms are normalisedto unity. The orange dot-filled histograms are all PSB galaxies, the line-filled histograms are thesubset with no ongoing residual star formation. The vertical dotted line indicates our PSB masscompleteness limit. correction for survey incompleteness effects. The post-starburst sample is complete abovea mass limit of 5 . × M ⊙ and the non-starforming post-starbursts have masses in thesame range as red-sequence galaxies at this redshift.
4. The mass flux onto the red sequence
By summing the total mass of post-starburst galaxies and using our knowledge fromsimulations to estimate how long they will be visible for, we can measure the massflux from the blue to the red-sequence that passes through a post-starburst phase. Inthis calculation, we include only the five galaxies which have completely stopped star-formation and therefore we argue are certain to be heading for the red-sequence.˙ ρ B → R , PSB = M B → R , PSB