Patricia B. Tissera- Galaxy Formation and Supernova Feedback

Patricia B. Tissera

GALAXY FORMATION AND SUPERNOVA FEEDBACK

Observational resultsChemical feedbackEnergy feedback

Patricia B. Tissera IAG-Lenac Advanced School

Patricia B. TisseraPatricia B. Tissera

In the last years, studies of chemical elements obtained in the Local Universeand at high redshifts have improved dramatically.

Chemical patterns are the result of different mechanisms which contribute togalaxy formation

growth of the structure:collapse, mergers, infall, etc

gas cooling and condensation

star formation and stellar evolution

supernova feedback:chemical + energy release

environmental effects: starvationstrangulation, etc

IAG-Lenac Advanced School


Supernova feedback is one of the process that contribute to structure the Insterstellar Medium (ISM).

CHEMICAL ENRICHMENT HYDRODYNAMICAL HEATING

•SN: Main source of heavy elements

•Change the cooling time

•evaporates cold-dense gas •galactic winds which can results in outflows or galactic fountains

•Regulates the star formation activity and enriches the ISM and IGM•Affects the gas dynamics



WHY DO WE CARE ABOUT METALLICITY ?

Chemical abundances and dynamical properties provide more stringentconstrains for galaxy formation models.

Eggen, Lynden-Bell & Sandage (1962): “galactic archaeology “ proposingthe so-called monolithic collapse model from studies of halo stars.

The MCM was first challenged by Searle (1977): Galactic globular clusters: wide range of metals abundances essentially independent of radius from the Galactic Center.

The importance of fossil signatures in the chemical/dynamical patterns whichcan be related to the history of formation.


Patricia B. Tissera

THE MILKY WAY

BULGE

DM HALO

THIN DISCTHICK DISC

STELLAR HALO


Patricia B. Tissera

MILKY WAY:THIN DISC

Rotationally supported: σ/V <<1.

Scale-length ~ 2-2.5 kpc (Siegel et al. 2001), hz ~ 280 pc

Stellar age distribution ~ [2,14]Gyr and [Fe/H] peaks at ~ -0.2

(Nordstrom et al. 2004)


Patricia B. Tissera

scale-lenght ~ 3 kpc , hz ~ 1 kpc (assuming a double exponential) .

t_medio ~ 12.5 +- 1.4 Gyr (Liu &Chaboyer 2000)

-2.2 < [Fe/H] <0.5 with <[Fe/H]> ~ -0.6 (Chiba & Beers 2000)

higher [O/Fe] than the stars in the thin disc.

MILKY WAY:THICK DISC


thick

thin

Patricia B. Tissera

MILKY WAY: BULGE

hz ~ 2 kpc; averged age ~ 10 Gyr for stars with hz > 400 pc

metallicity peak: [Fe/H] ~ -0.3 dex ( Zoccali et al. 2003).

lower [O/Fe] with respect to halo stars

there are young stars and on-going star formation

( Van Loon et al. 2003)


Patricia B. Tissera

MILKY WAY: STELLAR HALO

J/M ~ 0 (Freeman 1987); sopported by dispersion

<[Fe/H] > ~ -1.5 dex (Ryan & Nories 1991;Chiba & Beers (2000)

(σr, σphi, σz ) ~ (141, 106, 94) km/s


halo

Patricia B. Tissera

Freeman & Bland-Hawthorn 2002



The Local Group

Two massive galaxies and at least 38 small galaxies (dSphs, DEs and DIrrS)

DIrrs: gas rich galaxies, Mtot < 10^10 Mo; stars from 10 Gyr to recent born. age-metallicity relation.

DEs: ellipticals, Mtot < 10^9 Mo; mainly old and intermediate age stars; small gas fractions; high mass central concentration.

DSphs:diffuse, gas-deficient, little central concentration; Mtot ~ 10^7 Mo Lower [α /Fe] than stars in the galactic halo (Grebel et al. 2003; Grebel 2005)



The Local Group



GalaxiesLuminosity-metallicity and mass-metallicity relations:

There are well-known LMR and MMR in the local Universe.Observations suggest evolution in the zero point and slope of both relations.


SDSS: Tremonti et al .2004

Erb et al 2006: z~2.5


Damped Lymanα Systems



Damped Lymanα Systems



The study of the Formation and Evolution of Galaxies requires to be able tofollow the evolution of the structure in large scale, which is mainly determined

by gravitation,gravitation, and to describe the action of other processes such as gas cooling, star formation, stellar evolutiongas cooling, star formation, stellar evolution, etc.

Smooth Particle Hydrodynamics simulations are one of the most popular techniques to study galaxy formation.

However, the complex interaction of the non-linear gravitationalevolution and dissipative gas dynamics plus the action ofseveral physical process which introduce their own lengthand time-scales make the modelling of galaxy formation a severechallenge.



Simple one-zone model was discussed by Van den Bergh (1962) , Schmidt (1963)Four hypothesis:

The system is isolated: no inflows or outflows.The systems is well mixed at all timesThe systems starts from primordial abundances: Z(0)=0IMF and nucleosynthesis yields are unchanged

Instantaneous recycling

CHEMICAL FEEDBACK

There are numerous chemodynamical models for galaxy formation which have sofisticated the Simple Model (e.g. Larson 1976; Tinsley & Laron 1979; Burkert & Hensler 1988; Ferrini et al. 1992; Chiappini, Matteucci & Gratton 1997):

sophisticated stellar evolutionpoor initial conditions for galaxy formation



CHEMICAL FEEDBACK

Including chemical enrichment by individual elements provides a powerful tool to study galaxy formation in cosmological scenarios:

First attempts to introduce chemical feedback in SPH simulationsof MilkyWay type galaxies:

There are implementations that follow the metallicity Z (Springel & Hernquist 2003 and references therein)

Steinmetz & Muller (1994) SNII; global metallicity ZRaiteri et al. (1996; also Berczik 1999) SNII & SNIa; Fe & HCarraro et al. (1998)

Mosconi, Tissera, Lambas & Cora. (2001): SNII & SNIa, 13 ele. Lia, Portinari & Carraro (2002):detailed SE; difusion Kawata & Gibson (2003):SNII, SNIa,IS; Eth +Ekin Kobashashi (2004):detail SE; Eth +Ekin Scannapieco et al. 2005: SNII & SNIa, 13 ele + Multiphase+SNE



CHEMICAL FEEDBACK

Numerical space Physical space Need

Star particles Stellar populations

↔

IMF:SNe

long-lived stars

M* > 10 Mo; typical life-times: ~ 106 yrType II Sne

Main source of iron (Fe)

Typical life-times: ~ GyrType Ia Sne


Produce most O, Si, Ca, etc

YIELDS

Patricia B. Tissera

When SN explosions take place, they distribute metals according to the SPH technique. For a given chemical element x at a particle i,

Exploding star particle

Gaseous neighbours

CHEMICAL FEEDBACK

Mxi = ∑ j mj/ρj Mxi W(rij,hij)

Mxj =mj/ρj Mxi W(rij,hij)

Each neigbhour will receive

i j


Patricia B. Tissera

Sutherland & Dopita (1993).

At T= 10000 and ρ= ρ*:τ cool for primordial gas is 50 largerthan that of [Fe/H]=0.5 gas.

CHEMICAL FEEDBACK

τ cool ∝ T /ρ Λ(T)


Patricia B. Tissera

ENERGY FEEDBACK

Formation of spiral galaxies: angular momentum content, dynamical and chemical properties.

Galactic outflows: transportation of enriched material into the intergalactic and the intercluster media.

Formation of dwarf galaxies.

Regulation of the star formation process.



A more realistic representation of the ISM can be achieved if smallvolumes are studied so that supernova explosions can be describedtemporarily and spatially (e.g. Rosen & Bregman 1995; Avillez 2000).

Problems: relevant space and temporal scales are not adequately resolved.For Supernova Feedback in SPH simulations, this has been a main issue for years

If galaxy-scales want to be studied subgrid modellization of physical processes which are relevant at kpc scales

ENERGY FEEDBACK



SPH uses the information of the neighbouring fluid elements to the estimatethe hydrodynamical properties ρj = ∑ mi Wij

works against the coexistence of cold clumps and hot gas. artificially boost the cooling rate of hot gas close to dense cold media

overcooling of gas.

Overestimation of stellar massIf the SN energy is pumped directly into the sorrounding

gas of a star particle then , because of the short cooling times, it is radiated away producing no impact on the dynamics..

Multiphase representation of the ISM



Scannapieco et al modifies the way GADGET-2 estimates the neighbouring fluidelements based on Pearce et al. (1999, 2001) and Marri & White (2003).

Decoupling ModelDecoupling Model::

We estimate the hydrodynamical properties of the gas from the information of neighbours selected according to their thermodynamical

gas particles are prevented to interact with colder material.

particle j decouples from those particles i if Si > αSj, where S is the entropy of a gas particle.

non shock

Multiphase representation of the ISM


Patricia B. Tissera

with decoupling

without decoupling

Pearce et al. (1999, 2001)


Coexistence ofdifferent

phases in thegas

component.



Isolated Disc Galaxy Test

Idealized Initial Conditions: A spherical grid with superposed dark matter and gaseous particles is perturbed

giving rise to a r-1 profile.

The sphere is initially in solid body rotation with angular momentum

characterized by a spin parameter of λ ≈0.1.

Both the gas and the dark matter components are resolved with 9000 particles. The tests correspond to a 1012 Mo h

-1 (h=0.7) system with 10% of baryonic mass.

Fraction of gas in the different media defined as:

HOT GAS: T ≥ 8 x 104 K

WARM GAS: T < 8 x 104 K and ρ < 0.1 ρ∗

COLD GAS: T < 8 x 104 K and ρ ≥ 0.1 ρ∗


Patricia B. Tissera

SN energy (1051 ergs each SN) released by a star particle is distributed within its gaseous neighbours.

The Cold/DiffuseCold/Diffuse neighbours of a star particle: T < 8 × 104 K and ρ > 0.1 ρ*

ε rad radiated away

ε cold cold and densecold and dense neighbours

ε hot =1- ε hot - ε cold diffusediffuse neighbours

Patricia B. Tissera

ENERGY FEEDBACK



Cold gas particles accumulates it in a ReservoirReservoir until it is high enough to ensure that the gas particle will join “its own hot phaseits own hot phase” according to the decoupling scheme.

Diffuse gas particles thermalize the energy “instantaneously”.

ENERGY FEEDBACK


Patricia B. Tissera

Isolated galaxy

Hot/Diffuse Gas Cold Gas Stars


Patricia B. Tissera

Milky Way Type galaxy: Multiphase ISM


Patricia B. Tissera

Milky Way Type galaxy: Multiphase ISM

25kpc/h


Patricia B. TisseraPatricia B. Tissera11th Latin-American Regional IAU Meeting December 12 – 16 2005IAG-Lenac Advanced School

Patricia B. TisseraPatricia B. Tissera IAG-Lenac Advanced School



10^12Mo/h

10^9Mo/h

NO FEEDBACK FEEDBACK

Star formation is regulated without introducing anymass scale parameter.




No

Feed

GAS

Feed

No

STAR

Patricia B. Tissera

~4.4 reduction in the B/D ratio

Increase of Rd from 4.3 kpc/h to 7.3 kpc/h

No feedback

Feedback



100 kpc/h

Patricia B. Tissera

f

FINAL COMMENTSChemical properties of baryons together with dynamical andkinematical information can provide clues for unveiling the history of formation of the structure.


Supernova feedback is a key process in the formation of the structure.

Modelling SN feedback is tricky but it is possible if a multiphaseISM is also modelled.

Numerical simulations provide a tool to interpret observational data within a cosmological model.

Patricia B. Tissera- Galaxy Formation and Supernova Feedback

Documents