Tweet
Tweet

Slide 1

Slide 1 text

Yuhan Yao 2017-04-21@KIAA SNID: Supernova Identification!

Slide 2

Slide 2 text

https://people.lam.fr/blondin.stephane/graphics/index.html

Slide 3

Slide 3 text

B band Maxima.

Slide 4

Slide 4 text

Type II: Hydrogen Type Ia: Si Type Ic Type Ib: He Fillippenko, 1997 And many subtypes…

Slide 5

Slide 5 text

‘pec’=peculiar Ia-pec: 2002cx… II-pec: 1987A… And many subtypes… SNID Database

Slide 6

Slide 6 text

SNID Database 879 spectra of 65 SNe Ia 332 spectra of 19 SNe Ib/c 353 spectra of 10 SNe II Also galaxies, active galactic nuclei, stars, and novae. ⬆ In 2007 Larger now….

Slide 7

Slide 7 text

SNID Database

Slide 8

Slide 8 text

SNID Database Intrinsic Variance of https://people.lam.fr/blondin.stephane/graphics/index.html light blue: standard deviation dark blue: maximum deviation SNID Ia normal templates.

Slide 9

Slide 9 text

https://people.lam.fr/blondin.stephane/graphics/index.html Difficulty of observing SN Ia at high redshifts

Slide 10

Slide 10 text

Definition Supernova spectrum with redshift Template spectrum , Zero redshift Cross-correlation Maximize! n = 0, 1, 2, … , N-1

Slide 11

Slide 11 text

Preprocessing the SNe spectrum

Slide 12

Slide 12 text

Preprocessing the SNe spectrum

Slide 13

Slide 13 text

Definition Supernova spectrum with redshift Template spectrum , Zero redshift Cross-correlation — maximize

Slide 14

Slide 14 text

Redshift Estimate The r-Value · Spectrum Overlap Reliability

Slide 15

Slide 15 text

r-Value: correlation height-noise ratio

Slide 16

Slide 16 text

r-Value: correlation height-noise ratio

Slide 17

Slide 17 text

lap contours Redshift Estimate ❌ The r-Value · Spectrum Overlap

Slide 18

Slide 18 text

r · lap contours Redshift Estimate ❌ The r-Value · Spectrum Overlap

Slide 19

Slide 19 text

Initial and revised redshift estimates 1. All templates correlate with input, selects 10 highest peaks in . 2. Compute initial redshift , based on rlap-weighted median of all 3. Shift s(n) to , trim, search for highest peak in , find the revised redshift This time non-rlap-weighted median.

Slide 20

Slide 20 text

Save the width of the correlation peak. Redshift error 3. Shift s(n) to , trim, search for highest peak in , find the revised redshift This time non-rlap-weighted median.

Slide 21

Slide 21 text

Accuracy of Redshift & Age Simulation: choose normal Ia template spectra. Shifted: Contaminated: Added:

Slide 22

Slide 22 text

Accuracy of Redshift

Slide 23

Slide 23 text

Accuracy of Redshift

Slide 24

Slide 24 text

Accuracy of age

Slide 25

Slide 25 text

Accuracy of age

Slide 26

Slide 26 text

Type Ia normal

Slide 27

Slide 27 text

Type determination Normal vs. 1991T-like SNe Ia SNe Ia vs. SNe Ib/c Identifying SN ‘‘Oddballs’’ Using SNID for SN Identification

Slide 28

Slide 28 text

Type determination Normal vs. 1991T-like SNe Ia SNe Ia vs. SNe Ib/c Identifying SN ‘‘Oddballs’’ Using SNID for SN Identification

Slide 29

Slide 29 text

SN 1991T-like events: Local: up to 20% of SNe Ia. High redshift: goes down. Avishay, arXiv:1611.09353v1 SN 2011fe: regular Ia SNe.

Slide 30

Slide 30 text

no constraint ±0.01 ±3 both

Slide 31

Slide 31 text

Type determination Normal vs. 1991T-like SNe Ia SNe Ia vs. SNe Ib/c Identifying SN ‘‘Oddballs’’ Using SNID for SN Identification

Slide 32

Slide 32 text

No content

Slide 33

Slide 33 text

Type determination Normal vs. 1991T-like SNe Ia SNe Ia vs. SNe Ib/c Identifying SN ‘‘Oddballs’’ Using SNID for SN Identification

Slide 34

Slide 34 text

Avishay, arXiv:1611.09353v1 SN 2002cx-like events: Early: resemble 1991T, Late: resemble normal. SN 1991T-like events: Local: up to 20% of SNe Ia. High redshift: goes down. SN 2011fe: regular Ia SNe.

Slide 35

Slide 35 text

No content

Slide 36

Slide 36 text

– Dong, 2016, Science “SNID, a commonly used SN classification software that has a spectral library of most types of supernovae except SLSN, failed to find a good SN match. ”

Slide 37

Slide 37 text

Type determination Normal vs. 1991T-like SNe Ia SNe Ia vs. SNe Ib/c Identifying SN Ia ‘‘Oddballs’’ Using SNID for SN Identification

Slide 38

Slide 38 text

Interested? Download: https://people.lam.fr/blondin.stephane/software/snid/index.html Other methods (other than cross-correlation): Chi-square: superfit (Howell, 2005); SN-fit (Sainton, 2004); SFA (Riess, 1997) Bayesian: Kuznetsova & Connolly, 2007; Kunz, 2006 PCA: James, 2006 And improvements: Silverman, 2012, MNRAS, 425, 1789 Thank You!