Semi-analytic modelling in the era of wide, deep and multiwavelenght surveys Gabriella De Lucia Max-Planck Institut für Astrophysik Bologna – May 25, 2006

Outline: Semi-analytic models -- substructures & galaxies -- hybrid models – methods, limits & aims Applications -- the formation history of elliptical galaxies -- the hierarchical formation of BCGs (with J. Blaizot) -- some other prelimilary results…

galaxy formation Cooling (metallicity,structure, conductivity) Star formation (threshold, efficiency, initial mass function) Dust (formation, distribution, heating and cooling) Galaxy interactions (morphological transformations, induced star formation) Winds (IGM heating, IGM enrichment) Stellar evolution (spectro-photometric evolution, yields,feedback) AGN (BH growth, feedback)

The SAM – the classical models : Lacey & Cole (1993) Extended Press & Schechter “Planting” the merger trees These might provide a not adequate description of the merger rates (Benson et al., 2005)

Mathis H. et al., 2002  CDM  CDM The SAM – the hybrid models : Halo galaxy Central galaxy Satellite galaxy Springel et al. (2001) De Lucia et al., 2004

About 800 million subhaloes 64 time slices stored Group catalogues and information about their embedded substructures used to construct merger trees About 20 million galaxies Individual trees are stored separately so that the SAM can be run for each tree sequentially The Millennium Simulation (Springel et al. 2005)

The SAM – the physics : feedback star formation recycling Hot GasCold Gas cooling Ejected Gas Stars re-incorporation DLG, Kauffmann & White, AGN heating Croton et al. 2006

Luminosity function Colour-magnitude Tully-Fisher Metallicity- mass Gas fraction- luminosity DLG, Kauffmann & White, 2004

David, Forman, Jones 1995 Observational signature of different feedback models #1: ejection fast ejection slow wind fastwind slow DLG, Kauffmann & White, 2004

A slow ejection feedback scheme: David, Forman, Jones 1995 (M/L) V  190 – 230 (M/L)  (M/L) B  240 – 290 (M/L)  IMLR  – M  /L  DLG, Kauffmann & White, 2004

Ellipicals: an old controversy Visvanathan & Sandage (1977) Gladders et al. (1998) Bender 1992 Toomre & Toomre 1972 Kauffmann & Charlot (1998) Barnes 1992

Thomas et al Ellipicals: the [  /Fe] problem Thomas et al. did not use a self- consistent approach Previous SAM work based on extended PS theory Most previous SAM work not in a  CDM Universe Observations are still plagued by systematic uncertainties (template bias) Log [  /km s -1 ] [  /Fe] V observation s superwindsuperwind /Kennicutt conduction Nagashima et al. (2005) SNIa SNII

Three channels to make bulges: In a ‘minor’ merger the stellar mass of the merged galaxy is transferred to the bulge of the central galaxies + burst of a fraction of the combined cold gas A ‘major’ merger completely destroys the disc of the central galaxy + burst of a fraction of the remaining gas Disk instability (Mo, Mao & White 1998) 1,031,049 (810,486 ) Es with Mstar > 4x10 9 (1x10 10 ) Msun 16% Es / 66% Sp / 18 % S0 (MV < -18) (13%, 67%, 20% Loveday et al., 1996) AGN model for suppression of the cooling-flows (Croton et al., 2006) De Lucia et al., 2006 Ellipticals in a hierarchical model :

clusters field Elliptical galaxies also have a shorter formation timescale! This is “anti-hierarchical”!!! The star formation histories: De Lucia et al., 2006 redshift cluster ellipticals field ellipticals Lookback time (Gyr) M star = M sun M star = 10 9 M sun M  M  M  cluster field Lookback time (Gyr)

50 % stars formed 80 % Fraction of galaxies Ellipticals: formation & assembly: redshift 50 % stars in a single object 80 % redshift De Lucia et al., 2006

Comparison with previous work: Kauffmann & Charlot (1998 Kauffmann & Charlot (1998) most massive ellipticals younger than their less massive counter-parts!

AGN feedback Comparison with previous work: Cooling cutoff V crit = 350 km/s No AGN feedback No cooling cutoff reversal due to a combination of the change in the physical model and in the cosmology Kauffmann & Charlot: -- critical matter density -- cutoff 500 km/s

The Brighest Cluster Galaxies : NGC z ~ 0 CL z ~ 0.5 The most luminous and most massive objects in the Universe at the present epoch

The formation of BCGs: De Lucia & Blaizot, in prep.

The formation of BCGs: De Lucia & Blaizot, in prep. Time of last major merger 0.5 “Assembly vs formation” history 50 BCGs z = 0

The formation of BCGs: De Lucia & Blaizot, in prep. Mass in the main branch Fraction of mass gained through accretion >10 10 M sun <10 10 M sun Most of the stars of this BCG were not formed in the MB, but were instead accreted over time

The formation of BCGs: De Lucia & Blaizot, in prep. Mass in the main branch starburst quiescent the stars formed in separate entities in a quiescent mode. stars formed during Starbursts contribute only for about 10%.

The formation of BCGs: De Lucia & Blaizot, in prep. “Progenitors” more massive than M sun

The formation of BCGs: De Lucia & Blaizot, in prep. Burke, Collins & Mann 2000

B - V V Ellipticals: the evolution of the CM A well defined CM up to z~2 A clear bimodal distribution up to z~2 De Lucia et al., in preparation

The evolution of the stellar MF Drory et al Fontana et al ~ good agreement up to z~2 for massive gxs.

B - V Log [M stars] The origin of the colour-magnitude: B - V Log [M stars] Gallazzi et al., 2006 De Lucia et al., in preparation

Mock 2dF 75 x 4 deg 2 B AB < 19.5 Blaizot et al. 2005: MoMaF Tools to observe ideas :

D4000 HH Mock SDSS 18 < r < < r < < r < < r < 20 with J é r é my Blaizot Tools to observe ideas :

The colour-magnitude bimodality: u - r Quantitatively the CM bimodality is not well reproduced oTail of blue bright objects o Transition region not well populated o Excess of faint red satellites

The D4000 distribution: Same problems are visible in the distribution of D4000

Conclusions: Semi-analytic models are a technique for studying galaxy formation - they are not meant to be definitive! The observed down-sizing is not in contradiction with the hierarchical paradigm More massive systems assemble later than their less massive counter-parts (this does not seem to be in contradiction with observational data but more detailed comparisons nedeed here) The ever more detailed picture of our Universe also requires a more complex modelling and the development of new tools for a more straightforward comparison with observational data