Presentation on theme: "KASI Galaxy Evolution Journal Club Comparing the Relation between Star Formation and Galaxy Mass in Different Environments - B. Vulcani et al. 2010, ApJ,"— Presentation transcript:
KASI Galaxy Evolution Journal Club Comparing the Relation between Star Formation and Galaxy Mass in Different Environments - B. Vulcani et al. 2010, ApJ, 710, L1 - October 21, 2010 Joon Hyeop Lee
Staf Formation (SF) Activity varies systematically with galaxy mass and redshift. Ultra Compact Dwarf Damen et al. (2009) A&A, 453, 869
Strong and tight correlation between current star formation rate (SFR) and galaxy stellar mass for field SF galaxies at all redshifts out to z = 2. z ~ 2 galaxies: Pannella et al. (2009) ApJ, 698, L116
Gradual decline in the SFR of most galaxies since z ~ 1–2. Borch et al. (2006) A&A, 453, 869
Noeske et al. (2007) ApJ, 660, L47 Lower-mass galaxies have higher SSFR: massive galaxies formed most of their stars earlyier and on shorter timescales (downsizing).
How about the relationship between SFR and environemnt? Why it is expected that the SFR-M and SSFR-M relations should depend on environment: - While fast-acting environmental effects are unlikely to influence the SFR-mass relationship of SF galaxies, any physical mechanism slowly affecting the amount of gas available for SF should result in a slowly declining SFR, therefore a different SFR-M relation with environment. - Strangulation: the loss of halo gas reservoir in hierarchical merging process - The interruption of cold gas streams in dense environments that would leave unfueled galaxies to slowly consume their disk gas
How about the relationship between SFR and environemnt? In contrast, - Several works reported a lack of environmental dependence of the distribution on current SF activity. - Others failed to identify large population of galaxies in transition from the red sequence to the blue cloud in dense environments. Rines et al. (2005) AJ, 130, 1482 Cassata et al. (2007) ApJS, 172, 270
How about the relationship between SFR and environemnt? However, the relations between SFR and SSFR with mass have not yet been studied in groups and clusters. → The most direct way to clearly discrimnate between mass and environmental trends! SFR Stellar Mass RedshiftEnvironment ?
Ultra Compact Dwarf Data Set - EDisCS (ESO Distant Cluster Survey) : 604 spectroscopically confired members of 16 clusters (σ > 400 km s -1 ) + 10 groups (150 < σ < 400 km s -1 ) Quantities: stellar mass, SFR, SSFR - Stellar mass: photo-z fitting + M/L vs (B-V) relation (Bell & de Jong 2001) - SFR (Spizter 24μm flux) (Kewley et al. 2004), without dust correction for SFR [OII] for galaxies without 24μm detection, SFR tot = SFR [OII] corrected for dust - SSFR
Ultra Compact Dwarf Galaxies without 24μm detection - Divided into red and blue, in order to separate SF-powered (blue) and AGN-powered (red) → Blue emission-line (BEL) galaxies and Spitzer-detected galaxies have mostly late-type morphologies, in agreement with the assumption that they are star-forming. AGN contamination problem - Currently do not have a way to quantify AGN contamination. - 32% of red emission-line (REL) galaxies have late-type morphologies: AGN? Dusty SF? → 2 cases analysis: 1. [OII] emission in all red galaxies is dominated by an AGN and those REL galaxies are excluded 2. REL galaxies are included.
Comparison with the Field Comparison to the results of Noeske et al. (2007) for the field at 0.4
SFR-M relation in different environments (24μm + BEL galaxies): - change with redshift in both clusters (black) and field (green) - cluster SFRs are on average systematically lower than field SFR at the same mass: due to ~ 10% excess of galaxies with ‘reduced’ SFR for their mass.
Statistical Test: Monte Carlo simulations extracting randomly from the field sample a subsample with the same mass distribution as the clusters + Kolmogorov-Smirnov (K-S) test - z<0.6: cannot reject the null hypothesis of similar SFR distributions in clusters and the field. - probability < 90% in 54% of the simulations - probability < 95% in 29% of the simulations - z>0.6: rejects the hypothesis of similar SFR distributions - probability > 95% in 87% of the simulations SFR-M relation is flat above the mass limit - Spearman test: significant positive correlation only for the field at z>0.6 (99.9%), and no correlation in all other cases.
SFR-M relation (24μm + BEL + REL galaxies): - The difference between the field and clusters is more obvious and becomes progressively more pronounced towrd more massive galaxies. - ~ 25% ‘reduced’ SFRs for their mass compared to the field - K-S test: rejects the similarity with a probability of 100% (z 95% (z>0.6) in all the simulations.
The results highlight a change in the SFR-mass relation with environment. Quantifying the change: - mean SFR in the mass-matched simulations - The mean SFR of 24μm + BEL galaxies in clusters is 1.35±0.15 times lower than that in field. - The mean SFR of 24μm + BEL + REL galaxies in clusters is 1.63±0.20 times lower than that in field.
SSFR-M relation (24μm + BEL galaxies): - Clusters and field galaxies follow a similar decreasing trend of SSFR with mass (Spearman anticorrelation probability always > 99.9%).
SSFR-M relation (24μm + BEL + REL galaxies): - Cluster galaxies tend to have a lower SSFR than field galaxies. - For 24μm + BEL galaxies, the difference is not significant, but for 24μm + BEL + REL galaxies, K-S test confirms the difference with probabilities always > 95%. - Average SSFR ratio: 1.20±0.14 for 24μm + BEL, 1.31±0.17 for all galaxies.
- In all environments, the mass growth rate at a given mass decreases with time and it is lower for higher-mass galaxies. - A fraction of the SF cluster galaxies are building up their stellar mass at a lower rate than field galaxies: 10% and 30% of 24μm + BEL and all galaxies, respectively, lie below the field 25 percentile. - The cluster trends are steeper than the field trends (best-fit slopes differ by > 1σ), suggesting that SF in more massive galaxies differes more strongly with environment than SF in lower-mass galaxies.
Clusters versus Groups - clusters (σ > 400 km s -1 ) vs. groups (150 < σ < 400 km s -1 ) - only for high-z bin due to the sample size - Cluster environment deviates from the field trend, while group galaxies seem to follow the SFR-M relation of the field.
K-S test: - Clusters have a different SFR distribution from both groups and field with > 98% probability. - Cannot reject the hypothesis that the groups and field have a similar distribution. The group data are not sufficient to draw firm conclusions, but, if confirmed, this result suggests that the group environment is not influential for the link between SF activity and mass → Strangulation could not be relevant, at least in groups, and only cluster-specific processes could be important. Having removed the groups, - The mean SFR in clusters at z>0.6 is 1.93±0.02 (without REL) to 2.13±0.02 (with REL) times lower than in the field.
Conclusions 1. The first attempt to establish whether the SFR-M relation depends on environment: The relation in clusters is significantly different from the field in most cases. - Discriminating SF from AGN red galaxies will enable to quantify with higher precision the environmental effects. 2. Uncertainties: -The observed differences can be a lower limit to the real gap: projection effect, just recently accreted galaxies -Possible severe underestimation of the dust correction to the SFR [OII] : Many [OII] detected sources should have been detected at 24μm. 3.Any comparison of SFR and masses in different environments is meaningful only when all data probe down to low SFR levels.
Conclusions 4. There are significant differences between the SF activity of star- forming galaxies of the same mass in different environments. -Clusters show a lower SF activity then the field: pre-existing large population of early-type galaxies passively-evolving since high-z + currently star-forming galaxies host an average lower SFR than their field counterparts of similar mass. 5.Straightforward interpretation: “There are environmental effects suppressing SF in clusters”. -Fast-acting mechanisms would leave the SFR-M relation unchanged, while longer-timescale processes would affect it. -Strangulation would be ruled out if it is equally effective in groups and clusters and if the observed result in this paper is confirmed. -Ram pressure stripping (acts on a short timescale) may leave residual gas and low SFRs.
Conclusions 6.Alternative to environmental mechanisms: Other galaxy intrinsic properties besides mass (e.g. morphological distribution) influence the SF history and vary systematically with environment? - In clusters, a population of galaxies in transition from being blue SF to red passively-evolving, while such a population is much less noticeable in the field and groups.