probing the milky ways oxygen gradient with planetary nebulae dick henry h.l. dodge department of...

PROBING THE MILKY WAY’S OXYGEN GRADIENT WITH PLANETARY NEBULAE

Dick HenryH.L. Dodge Department of Physics & Astronomy

University of Oklahoma

Collaborators: Karen Kwitter (Williams College)

Anne Jaskot (University of Michigan)Bruce Balick (University of Washington)Mike Morrison (University of Oklahoma)Jackie Milingo (Gettysburg College)

Thanks to the National Science Foundation for partial support.

Homer L. Dodge Department of Physics & AstronomyUniversity of Oklahoma

Astrophysics and CosmologyAtomic and Molecular Physics

Condensed Matter PhysicsHigh Energy Physics

ASTRONOMY AT

Eddie BaronSupernova studies

David BranchSupernova studies

John CowanChemical evolutionMilky Way studiesSupernova remnants

Dick HenryChemical evolutionGalaxiesNebular abundances

Bill RomanishinSolar system

Yun WangCosmologyDark matterDark energy

Karen LeighlyActive Galactic nuclei

ASTRONOMY AT

Eddie BaronSupernova studies

John CowanChemical evolutionMilky Way studiesSupernova remnants

Dick HenryChemical evolutionGalaxiesNebular abundances

Bill RomanishinSolar system

Yun WangCosmologyDark matterDark energy

Karen LeighlyActive Galactic nuclei

ASTRONOMY AT

Eddie BaronSupernova studies

Dick HenryChemical evolutionGalaxiesNebular abundances

Yun WangCosmologyDark matterDark energy

Karen LeighlyActive Galactic nuclei

OUTLINE

1. Introduction to chemical evolution of galaxies

2. Abundances and abundance gradients3. Planetary Nebula abundance study4. Statistics and the inferred gradient5. Conclusions

MILKY WAY MORPHOLOGY

• Halo• Bulge• Disk• Dark Matter Halo

Galactic Chemical EvolutionThe conversion of H, He into metals over time

Stars produce heavy elements

Stars expel products into the interstellar medium

New stars form fromenriched material

CHEMICAL EVOLUTION OF A GALAXY

Stars produce heavy elements

Stars expel products into the interstellar medium

INTERSTELLARMEDIUM

Stellar Evolution

Stellar Evolution

Stellar Evolution

Gas pressure outward Gravity inward

Stellar Evolution

Gas pressure outward Gravity inward

4 1H --> 4He

3 4He --> 12C

12C + 4He --> 16O

16O + 4He --> 20Ne

20Ne + 4He --> 24MgStellar Nucleosynthesis Reactions

Stellar Evolution

Gas pressure outward Gravity inward

4 1H --> 4He

3 4He --> 12C

12C + 4He --> 16O

16O + 4He --> 20Ne

20Ne + 4He --> 24MgStellar Nucleosynthesis Reactions

Supernova

Stellar Evolution

Gas pressure outward Gravity inward

4 1H --> 4He

3 4He --> 12C

12C + 4He --> 16O

16O + 4He --> 20Ne

20Ne + 4He --> 24MgStellar Nucleosynthesis Reactions

Supernova

Planetary Nebula

Local Results of Galactic Chemical Evolution

1. INTERSTELLAR MEDIUM BECOMES RICHER IN HEAVY ELEMENTS

2. NEXT STELLAR GENERATION CONTAINS MORE HEAVY ELEMENTS

Time

Heavy element abundances

Age-Metallicity Relation

Global Results of Chemical EvolutionOxygen Abundance Gradient

Abundance gradient Star formation history

Abundance gradients constrain:

1. Star formation efficiency

2. Star formation history

3. Galactic disk formation rate

WHAT DO ABUNDANCE GRADIENTS TELL US?

Project Goal

•Measure the oxygen gradient in the ISM of the Milky Way disk

•Employ planetary nebulae as abundance probes

•Perform detailed statistical treatment of data

Abundance Probes of the Interstellar Medium

•Stellar atmospheres: absorption lines

•H II Regions: emission lines

•Planetary Nebulae: emission lines

•Planetary Nebula

•Expanding envelope from dying star

•Contains O, S, Ne, Ar, Cl at original interstellar levels

•C, N altered during star’s lifetime

•Heated by stellar UV photons

•Cooled through emission line losses

PLANETARY NEBULAE

THE PN SAMPLE

• Number: 124• Location: MWG disk• Distance range: 0.9-21 kpc (~3-60 x 103 ly) from

center of galaxy• Data reduced and measured in homogenous

fashion• Oxygen abundances for all 124 PNe• Galactocentric distances from Cahn et al. (1992)

Data Gathering

CTIO 1.5m KPNO 2.1m APO: 3.5m

Emission Spectrum

The Physics of Emission Lines

• Bound-bound transition

• Inelastic ion-e- collision

• Radiative de-excitation• Photon production

h

Calculating Abundances from Emission Lines

I(el)

I(H) f (t,n)C

N(el)

N(H)

I(el)

I(H)

N(el)

N(H)

Abundance Software

Measure

Results: 12+log(O/H) vs. Rg

Statistical AnalysisLeast squares fitting

Input:

•Stats program: fitexy (Numerical Recipes, Press et al. 2003)•Data points: 124 (122 degrees of freedom)•Errors: 1 σ errors in both O abundances and distances•O errors: propagated through abundance calculations•Distance errors: standard 20%

Output:•Correlation coefficient and its probability•Slope (b) & intercept (a) •Χ2, reduced X2, and X2 probability

RESULTS: Trial #1

• a = 9.15 (+/- .04)• b = -0.066 (+/- .006)• r = -0.54 (r2=.29)• χν = 1.46

• qχ2 = 0.00074 (<.05)

Gradient = -0.066 dex/kpc

Improving the Linear Model

• Assume statistical errors don’t account for all of the observed scatter in O abundances

• Add natural scatter to statistical O/H abundance errors

• σtotal = 1.4 x σstat

Natural Scatter

• Poor mixing of stellar products in the ISM• Stellar diffusion: stars migrate from place of

birth to present location• Age spread among PN progenitors

• a = 9.09 (+/- .05)• b = -0.058 (+/- .006)• r = -0.54 (r2=.29)• χν = 1.00

• qχ2 = 0.49 (>.05)

2

RESULTS: Trial #2

Gradient = -0.058 dex/kpc

Different Models

•Gradient steepens in outer regions (Pedicelli et al. 2009; Fe/H)

•Gradient flattens in outer regions (Maciel & Costa 2009; O/H)

2-part linearquadratic

Two-part Linear Fit

Quadratic Fit

12+log(O/H) = 8.81 – 0.014Rg -0.001Rg2

Compare with Stanghellini & Haywood

Comparisons with Other Object Types

COMPARISONS

CONFUSION LIMIT

• Observed range in O/H gradient: -0.02 to -0.06 dex/kpc

Improvement will depend upon knowing:

1.Better distances to abundance probes2.Origin of natural scatter

Is Improving Gradient Accuracy Worth the Effort?

STAR FORMATION THRESHOLD (M pc-2) PREDICTED GRADIENT (dex kpc-1)

7.0 -0.059

4.0 -0.025

Observed gradient range: -0.02 to -0.06 dex kpc-1

Marcon-Uchida (2010): Sensitivity to star formation threshold

DISK FORMATION TIMESCALE PREDICTED GRADIENT RANGE (dex kpc-1)

Begins at galaxy formation, disk-wide -0.009 to -0.027

Increases with distance from center -0.056 to -0.091

Fu et al. (2009): Sensitivity to the timescale for disk formation

CONCLUSIONS1. We obtain a new O/H gradient of -0.058 +/- .006 dex kpc-1.

2. A good linear model of the data requires the assumption of natural scatter.

3. Observed gradient range ~ -0.02 to -0.06 dex kpc-1. We are at the confusion limit.

4. Improvements will come with better distances and the understanding of the natural scatter.

5. The endeavor is worthwhile for understanding the evolution of our Galaxy.

SN 1987A: 2/23/87

Distance from galaxy’s center

Heavy element abundances

Disk Abundance Gradient

OTHER SPIRALS

NEBULAE AS PROBES OF THE INTERSTELLAR MEDIUM

H II REGIONS• Photoionized and heated by young

hot central star(s)

• Radiatively cooled via emission lines

• Te ~ 104 K

• Density ~ 10-102

• 90% H, 8% He, 2% metals

Measuring Abundances: Spectra

• Emission spectrum

• Absorption spectrum