PyParadise Developed by: Bernd Husemann (MPIA), Omar Choudhury - - PowerPoint PPT Presentation

PyParadise

Developed by: Bernd Husemann (MPIA), Omar Choudhury (AIP)

• C. Jakob Walcher

Leibniz Institut für Astrophysik Potsdam (AIP)

24.11.2016 SELGIFS school, Walcher

Physical properties

• Physics 101: A physical property is described by

– a number – a unit – an errorbar

• Only well-defined quantities can be measured.
• The “Star Formation History” is not a well-defined

quantity.

• The “contribution of stars aged between 1*109

and 5*109 yrs to the total luminosity in the V-band” IS a well-defined quantity.

2

24.11.2016 SELGIFS school, Walcher

Fitting Spectral Energy Distributions

3

“Minimizing χ2 is a maximum likelihood estimation of the fitted parameters if the measurement errors are independent and normally distributed. “ χ2 is a measure of probability: P(D|M) ∝ e−χ2/2

Press+, Numerical Recipes

χ2 compares a model and a dataset A model is a prior! There is no fitting without prior!

24.11.2016 SELGIFS school, Walcher

Fitting optical galaxy spectra: decompose into three Χ2

4

Kinematics of stars Stellar populations Emission lines

χ2 =

n

X

i=0

Fi − PM

k=1 akSi[tk, Zk, Tk]

σi !2

A amplitude, v velocity, σkin velocity dispersion

χ2 =

n

X

i=0

✓Fi − Si ∗ G(A, v, σkin) σi ◆2

ak weights, tk age, Zk Metallicity, Tk Extinction Ak amplitude, λk line ID, v velocity, σkin velocity dispersion

χ2 =

n

X

i=0

Fi − PL

k=1 AkGi[λk, v, σkin]

σi !2

Fi, Si, sigmai, n Number of pixels

24.11.2016 SELGIFS school, Walcher

Stellar populations considerations

• Linear, but ill-posed problem

– Regularization assumes something – Bootstrap your way through the problem

• Transmission - Extinction issue

– Normalize spectra (but restrict wavelength range)

• Stellar populations accuracy ≠ continuum

subtraction precision = kinematics determination

– First case stelpop models – Second case stars better!

5

24.11.2016 SELGIFS school, Walcher

Pyparadise three main modules

6

Non-linear (MCMC) fit of kinematics v, sigma, (h3, h4) Linear (NNLS) inversion for stellar populations <age>, <feh>, etc. (also per age bins) Non-linear (MCMC) fit of emission lines F , v, sigma

24.11.2016 SELGIFS school, Walcher

Pyparadise fiducial run

7

Kinematics fit Stelpop fit Initial stelpop Emission line fit Kinematics fit Stelpop fit

24.11.2016 SELGIFS school, Walcher

Pyparadise bootstrap run

8

Kinematics fit Stelpop fit Disturb spectrum, subsample template basis Do N times Emission fit (Optional) (Optional)

24.11.2016 SELGIFS school, Walcher

Pyparadise bootstrap run

8

Kinematics fit Stelpop fit Disturb spectrum, subsample template basis Do N times Emission fit Mean, variance (Optional) (Optional)

24.11.2016 SELGIFS school, Walcher

Overall results

• Stellar populations

– Mean light weighted ages, Z, etc. for all templates – Mean light weighted ages, Z, etc. in bins of ages – Bootstrap-based errorbars – Weights of each SSP

• Kinematics

– Light weighted v, sigma – Bootstrap-based errorbars (incl. stelpop degeneracies)

• Emission lines

– Fluxes, v, sigma – Bootstrap-based errorbars (incl. stelpop and kin degeneracies)

9

24.11.2016 SELGIFS school, Walcher

Fit quality examples

10

24.11.2016 SELGIFS school, Walcher

Fit quality examples

11

Things to do when you fit spectra (and that you can do with PyParadise - but actually with any good software!)

24.11.2016 SELGIFS school, Walcher

Verify convergence!

13

MCMC iteration number parameter value

24.11.2016 SELGIFS school, Walcher

Spectrum of NGC6553: Schiavon+05 Walcher+2009

Test your model!

14

Globular Cluster Spectrum Literature value Our result

See also Conroy and Vazdekis models

24.11.2016 SELGIFS school, Walcher

Test your method on the relevant parameter!

15

Age [α/Fe] [Fe/H] Z

Recovered Input

24.11.2016 SELGIFS school, Walcher

Test applicability of Χ2 statistics!

16

Walcher et al., 2015

theory data

24.11.2016 SELGIFS school, Walcher

Test applicability of Χ2 statistics!

17

Walcher et al., 2008

theory data

24.11.2016 SELGIFS school, Walcher

Test applicability of Χ2 statistics!

17

Walcher et al., 2008

theory data

NO!

24.11.2016 SELGIFS school, Walcher

Look at your residuals!

18

Walcher et al., 2015

Rest-frame Observed- frame

24.11.2016 SELGIFS school, Walcher

Look at your residuals!

19

24.11.2016 SELGIFS school, Walcher

Look at your residuals!

19

CALIFA Pipeline vs1.4 / DR2 Template mismatch dominates!

24.11.2016 SELGIFS school, Walcher

Look at your residuals!

19

CALIFA Pipeline vs1.4 / DR2 Template mismatch dominates! Some molecular feature

24.11.2016 SELGIFS school, Walcher

There is science in template mismatch!

20

24.11.2016 SELGIFS school, Walcher

Do not trust good fits!

21

• bserved spec

fit around 4000A fit around 5000A

24.11.2016 SELGIFS school, Walcher

Do not trust good fits!

21

• bserved spec

fit around 4000A fit around 5000A

24.11.2016 SELGIFS school, Walcher

Do not over-interpret your results

22

Do not trust details in your “star formation history”! Peaks are a consequence of Regularization is a possible solution, but imposes smooth SFHs, which may not be true

24.11.2016 SELGIFS school, Walcher

Do not over-interpret your results

22

Do not trust details in your “star formation history”! Peaks are a consequence of Regularization is a possible solution, but imposes smooth SFHs, which may not be true This MUST be wrong!

Installing and running PyParadise

24.11.2016 SELGIFS school, Walcher 24