1

The Schmidt Law is 50 years old, but we still talk about it

Marteen Schmidt at the conferenceSFR@50:Filling the Cosmos with StarsJuly 6-10 2009, Spineto

Samuel Boissier, Laboratoire d'Astrophysique de Marseille

Slid

e pa

rtly

tak

en f

rom

R.

Ken

nic

utt

intr

odu

ctio

n

Star Formation and its Efficiency

2

Star Formation and its Efficiency

1) Current status of the Schmidt Laws

2) Toy models relating the various Schmidt Laws

3) Some studies at “low densities”

4) Krumholz et al.: a simple theory for star formation ?

3

Normal disks

Circum-nuclear starburst

Centers of normal galaxies

SFR = A GASSFR = A GAS1.41.4

SFR = A GAS /SFR = A GAS / dyndyn

The Schmidt-Kennicutt law is a nice relationship...

1-Schmidt Law (global averages)

Kennicutt et al. 1998

4

F. Bigiel talk

Resolved nearby spirals:THINGS survey : HIBima/Heracles : COGALEX : UV SFRSpitzer : IR SFR

1-Schmidt Law (local)

5

F. Bigiel talk

Resolved nearby spirals:THINGS survey : HIBima/Heracles : COGALEX : UV SFRSpitzer : IR SFR

1-Schmidt Law (local)

Barry Madore proposes that this shape results from the combination of two timescales:

- the star formation timescale- the star formation tracer timescale

(e.g. 100 Myr for FUV, 10 Myr for Ha)

6

WORK IN

PROGRESS

F. Bigiel talk

1-Schmidt Law (efficiency in outer disks)

7

GALEX UV observations allowed to extend the Schmidt law at low densities(below the “H threshold”)

Radial Schmidt Law, Boissier et al. 2007

1-Schmidt Law (radial)

8

Newest radial profiles: UV and H in the SINGS sample

(Juan Carlos Muños Mateos thesis with Armando Gil de Paz)

1-Schmidt Law (radial)

9

Kennicutt changes profiles

70%

15%

15%

1-Schmidt Law (radial)

10

Kennicutt changes profiles

70%

15%

15%

WARNINGWay to truncate in

both or in Hα and not UV:ram-pressure (e.g.

NGC4569, Boselli et al. 2008), unrelated to star formation !

1-Schmidt Law (radial)

11

“The decreasing of SFE in the outer parts of the disks is just the result of the radial increasing of thickness of gas layer, which in turn is caused by the low density of stellar discs.”

Zasov:

1-Schmidt Law (volume)

12

2-Toy models(toy model #1)

Toy-model #1:- Exponential Gas Disk - Assuming a local star formation law: SFR = a x GAS Local Threshold (10 Msol / pc2)

Computing the azimuthal relationship

log SFR

log GAS

Gas

SFR

13

log SFR

log GAS

Double arm spiral system, over-density=3, decreasing covering fraction with R

The azimuthal law present a break at much larger radius and a steeper slope

SFR

Gas

2-Toy models(toy model #1)

14

Comparison to the observations of Thilker et al. 2007: - azimuthally averaged law steeper than local one- Local “Threshold” invisible in the azimuthally averaged law

RADIAL LOCAL

2-Toy models(toy model #1)

15

Toy-model #2: - Gaseous disk Inner disk : = 0 exp(-R/Rco) ; Rco ~ R* ~ R25/4.6 Outer disk : = 3 Msol / pc2 (profiles similar to those in Bigiel et al. 2008; Leroy et al. 2008)

Local assumptions:- 2 phases : H2 and HI GAS < 10 Msol / pc2 : H2=0 GAS > 10 Msol / pc2 : H2 / HI =Rmol Rmol=10.6 exp (-R / 0.21 R25) (e.g. Leroy et al. 2008)

- SFR H2 (Leroy et al., Bigiel et al.)

2-Toy models(toy model #2)

16

Rmol

Decreasing molecular fraction

0= 80 Msol/pc

Extended star formation

PROFILESSCHMIDT LAWS

Molecular gas

Atomic Gas

Total Gas

2-Toy models(toy model #2)

17

H2:Azimuthal & Global = Local

SF “below” threshold

0= 20 to 200 Msol/pc2

HI: Azimuthal = dispersed Global = relationship, steeper than H2

HI+H2: Global nice relationship, Azimuthal = dispersed at low density, below thelocal threshold...

SCHMIDT LAWS

Molecular gas

Atomic Gas

Total Gas

2-Toy models(toy model #2)

18

A crucial point is the Covering Fraction of the arms.

B. Madore made similar models based on the covering fraction of star forming regions.

Back to observations : can we relate the dispersion in the Schmidt Law to the covering fraction of SF regions ?

Is there covering fraction laws ?

-> Coming investigation with B. Madore

2-Toy models(coming next)

19

XUV galaxies (e.g. Thilker et al. 2005, 2007): N4625 seems to follow a Schmidt law ...

HI + UV profiles in Gil de Paz et al. 2005

NGC4625

(See also Dong et al. 2008 for some outer regions of M83)

Radial Schmidt Law,Boissier et al. 2007

3-Low densities (XUV galaxies)

20

Tidal Dwarf Galaxies (*) also follow the Schmidt Law(Lisenfeld et al. 2001)

3-Low densities (TDG)

21

Study of a sample of massive Low Surface Brightness galaxies

- Lower star formation efficiency than normal spirals if UV taken as direct indicator of SFR with standard calibrations

Wyder et al. (2009) also suggests a lower SF efficiency in his dwarfs/LSBs

Boissier et al. 2008

log HI

log SFR

3-Low densities (LSB galaxies)

22

Compared to OBSERVED spirals

Compared to EXTINCTION-CORRECTED spirals

Some LSB galaxies have red FUV-NUV colors

Log HI MASS

FU

V-N

UV

FU

V-N

UV

3-Low densities (LSB galaxies)

Boissier et al. 2008, updated with GR4 data

23

Comparison to other studies:

Wyder et al. (2009) : not really red

Roychowdhury et al. (2009): kind of red

FU

V-N

UV

FU

V-N

UV

Log HI MASS

3-Low densities (LSB galaxies)

24

How to explain red FUV-NUV colors ?

- Extinction ?Spitzer observations of a few LSB galaxies show no dust (Rahman et al. 2008, Hintz et al. 2008)

- Initial Mass Function ?Pflamm-Altenburg, Kroupa et al. suggest SFR-dependent Krumholz & McKee (2008) suggest Density-dependent(but we have red colors for the largest SFR)

- Aging burst

3-Low densities (LSB galaxies)

25

The red FUV-NUV color can be explained with a time effect :

Bursty star formation histories, with latest burst within a few 100 Myr.

Boissier et al. 2008, updated with GR4 data

H FUV NUV

FUVNUV

NUVF

UV

-NU

V

3-Low densities (LSB galaxies)

26

H FUV NUV

FUVNUV

NUVCould a similar time effect also explain

the H vs UV discrepancy in low density regions/dwarfs ? (Boselli et al.submitted)

Assuming SFRH of bursts each T

Prediction: H/UV distribution

3-Low densities

27

A temporal effect could explain: - Large dispersion of H/UV- Small trend from dwarfs (bursty) to spirals (continuous SF)

(see also Fioc & Rocca-Volmerange, 1999)

What about the IMF ? (Pflamm-Altenburg, Kroupa et al.) ?

Boselli et al., submitted

3-Low densities

28

However, Calzetti showed that some studies support IMF variations

3-Low densities (IMF effect)

Pflamm-Altenburg et al.

29

However, Calzetti showed that some studies support IMF variations

Warning : Extinction Corrections may create

this trend !(Boselli et al., submitted)

Pflamm-Altenburg et al.

3-Low densities (IMF effect)

30

The SFR depends on:

The molecular fraction fH2 — determined by radiation and chemistry, depends on galaxy Σ, Z

The free-fall time in molecular clouds tff — determined by SF feedback in low Σ galaxies, by galaxy Σ in high Σ galaxies

The star formation rate per free-fall time ε ff — this is always ~1% due to the physics of turbulence

Krumholz:

4- Krumholz theory for star formation

31

Atomic and Molecular Star Formation Laws

HI H2 Total

Contours: THINGS (Bigiel et al. 2008)Lines: Theory for metallicities from 0.1 x solar to 3 x solar

(Krumholz, McKee, & Tumlinson 2009)

4- Krumholz theory for star formation

32

Successful Model Predictions● H2 fractions seen by THINGS

(Krumholz, McKee, & Tumlinson 2009)● Maximum HI columns of DLAs

(Krumholz et al. 2009)● When ram-pressure stripping

causes galaxies to lose H2 (Fumagalli et al. 2009; poster at the conf.)

4- Krumholz theory for star formation

33

Bolatto: it also works in the Magellanic Clouds

4- Krumholz theory for star formation

34

Bolatto: it also works in the Magellanic Clouds

4- Krumholz theory for star formation

35

Summary

- Schmidt Laws : local, radial, global.

- Radial variation of the Star Formation Effiency: - surface density ?- metallicity ?- time-scales effects ?

- Hα/UV : temporal + IMF effect

- Threshold or not threshold- local vs radial

- Hα vs UV

Star Formation and its Efficiency