1. Dani Maoz, Tel-Aviv University
Deriving metal enrichment history directly from observed supernova rates and yields.
Dani Maoz, Tel-Aviv University
(Maoz, Mannucci, Nelemans ARAA 2014)

2. Bottom line: cosmic iron accumulation history

3. Core-collapse SNe: progenitors are M>8 Mʘ stars, stellar evolution pretty clear, details of explosion shaky.

4. Core-collapse SNe:
We do know:
essentially “instantaneous” after star formation  CC SN rate must track the cosmic SFR.
For standard IMF: 0.01 SNe per formed Msun

5. Madau & Dickinson 14

6. Madau & Dickinson 14

7. CC iron yields are measurable directly from the SN light curves
0.2 Msun 
0.02 Msun 
Kushnir 15

8. Ratio of 3:1 Types II to Ibc …. Most Type II are IIP
Li+ 2011
Mean iron yield pr CC SN = ¾ * ¼ * 0.2 = Msun

10. Thermonuclear (“Type-Ia”) supernovae are wonderful cosmic distance indicators!
A. Howell, SNLS

12. Nobody knows exactly WHAT is exploding and HOW!
What?: (who are the progenitors)
How?: (physics leading up to, and of, explosion )

13. Type-Ia SNe: What do we know?
thermonuclear explosions
C,O Ni

14. How do we know this?
1. no H, He in the spectrum.
2. optical luminosity from radioactive decay of ~0.7 Mʘ of
56Ni Co Fe
t=9 d t=114 d
2.5 log flux
Time (days)

15. Hβ Hα
Core-collapse SN
Type-Ia SN

16. How do we know this?
1. no H, He in the spectrum.
2. optical luminosity from radioactive decay of ~0.7 Mʘ of
56Ni Co Fe
t=9 d t=114 d
2.5 log flux
Time (days)

17. The most recent certain type-Ia SN nearby: SN 1572 (Tycho’s)

18. Tycho’s SN Ia remnant, after 440 years
Badenes et al. 2006

19. Howell+09
0.7 Msun 

20. So, how to create?:
~0.7 Mʘ of 56Ni,
no traces of H,He,
kinetic energy of ejecta ~1051 erg
in regions with no massive stars.
Blow up a ~1.4 Mʘ white dwarf! (Hoyle & Fowler 1960)
WD is made almost entirely of C,O.
EOS: degenerate electron gas. If ignited, unstable to thermonuclear runaway!

21. C/O
But, how to ignite the WD?

22. “single degenerate” (“SD”) (Whelan & Iben 1974)WD
Main sequence, subgiant, red-giant, or “helium star”

23. “double degenerate” (“DD”) (Webbink 1984; Iben & Tutukov 1984)

24. Also: “collisional double degenerate”
(Benz+, Hawley+, Loren-Aguilar+, Raskin+, Rosswog+, Thompson, Katz & Dong, Kushnir+, Garcia-Senz+…)

25. Measuring SN Ia Rates
Can give clues to progenitors
But also can reveal the timescales of metal enrichment by SNe Ia

26. SN Ia “delay time distribution” (DTD): =
the hypothetical SN Ia rate vs. time following a short burst of star formation.
Different progenitor scenarios predict different DTD
Star formation rate SN DTD
SFR
t= time
SN Rate
t= time

27. e.g., Double-Degenerate scenario.
Consider population of binary WDs.
Time until merger of each pair (gravitational wave losses):
DTD ~ t -1 expected generically

28. double-degenerate: DTD ~ t -1 expected generically

29. single-degenerate: DTD cutoff at few Gyr
similarly:
single-degenerate: DTD cutoff at few Gyr
MS secondaries M<2 Mo cannot transfer mass stably
Decreasing secondary mass

30. Recovering the delay time distribution (many different ways to do it)
e.g. SN rates in galaxy clusters

31. SDSS z=0.68
Sharon et al. (2010)

33. The SN rate vs. redshift in galaxy clusters
B10
Cosmic time
Maoz, Sharon, Gal-Yam (2010)

34. Time-integrated # of SNe-Ia must produce observed mass of Fe in clusters (minus mass from CC-SNe)
Maoz, Sharon, Gal-Yam (2010)
SN rates in galaxy clusters + iron/star mass ratio

35. Time-integrated # of SNe-Ia must produce observed mass of Fe in clusters
Maoz, Sharon, Gal-Yam (2010)
SN rates in galaxy clusters + iron/star mass ratio

36. How to recover the delay time distribution
or… SN rates vs. redshift in field, compared to cosmic SFH

37. * = Star-formation history (z) SN delay time distribution (t)
SN rate (z) =
time

38. SN rate SFH delay time dist.

39. SN rate vs. redshift e.g.: SN rate at high z from the Subaru Deep Field
Poznanski et al. 2007,
Graur et al. 2011
4x(2-night) runs
Stare at the 0.25 deg2 SDF:
r, i ~ 27 mag, z ~ 26 mag

40. To measure a SN rate, need to discover some SNe.
1 SN SN
200 yr galaxy ,000 galaxies x yr =
SNe visible at a given time
SN is visible for ~2 weeks

45. SNSDF , z=1.66

48. SN rates out to z=2 and beyond with HST CLASH/CANDELS

52. How to recover the delay time distribution
or… SN Rates vs. individual galaxy star-formation histories

53. SN rate SFH delay function
True also in an individual galaxy!
N = r ∙ t
expec. value for # SNe in given galaxy
visibility time
expect. value visibility time

54. Compare observed number of SNe (0 or 1) in each galaxy to expectation value for given model DTD1 1 1 1

55. SDSS-II SNe Ia in Stripe 82 galaxies with SDSS spectra and SFHs
Maoz, Brandt, Mannucci 2012
SDSS-II SNe Ia in Stripe 82 galaxies with SDSS spectra and SFHs

56. A SN survey among 700,000 SDSS spectra: 90 SNe Ia (Graur & Maoz 12)

57. t -1
Maoz+11, Maoz+12, Graur & Maoz 12

58. Emerging Picture:
* Wide distribution of delay times, looks like ~ t -1 (DD?)
* >1/2 of SNe-Ia prompt (< 1 Gyr)

60. Cosmic iron accumulation history

61. Cosmic iron accumulation history

62. Summary
Measured SN rate and SF histories, plus observed iron yields, plus observational SN Ia DTD, permit a purely empirical Fe enrichment history.
CC SNe are dominant iron source until z=1, then become comparable to SN Ia.
today’s iron is 50:50 from Ia and CC.
Mean cosmic iron density today: (Z/Zo)Fe=0.09
10% of baryons in stars (solar metallicity?), so little left for IGM. But, ICM in clusters has (Z/Zo)Fe=0.3, unexplained by SNe.
6. Future IGM detection maybe possible thru absorption at FeXVII l 17 A

65. ~ 3 - 7%
~(1-2)x10-3
SN-Ia/Msun
~33
Msun

67. Present-day
mean Fe abundance solar
SN rates now give direct measure of metal accumulation over most of cosmic history
Graur+ 11