CSP-II Stripped-Envelope SNe Spectroscopy In The NIR
CSP-II Stripped-Envelope SNeSpectroscopy in the NIRMelissa ShahbandehFlorida State UniversityMay 2019E. Hsiao, P. Hoeflich, C. Ashall, M. PhillipsCarnegie Supernova Project II
Outline Introduction Sample characteristics Why do we care? Optical vs NIR Methods to identify the lines Preliminary ConclusionM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
What is a SE-SN? SN explosion from a progenitor which has lost most of itsenvelope Found in the H II regions and spiral arms Core Collapse explosion of massive stars (ZAMS 8 - 60 M ) The envelope is lost due to: ‣Radiation driven winds‣Common envelope phase in a binary system‣Fast rotation (Be stars)W49BBAsymmetric explosionImage Credits: X-ray: NASA/CXC/MIT/L.Lopez et al.; Infrared: Palomar; Radio: NSF/NRAO/VLAM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
The Carnegie Supernova Project II (CSP-II)0.861.01.21.41.61.82.0 2.2 2.4 2.6 2.8SNe in the NIRMg II 52011-2015SLSN ICa IIOne of the emphases wason the NIR spectra (0.82 2.5 µm)Las CampanasObservatory: Magellan II(Baade FIRE), Swope, duPont4SN 2015bn19dMg II0.84980.85420.86621.6787Mg IISi IIIHe I2.15691.2523 1.33951.2601 1.34971.36441.0830log10(F λ ) constant 0.92271.00921.0927SN IaiPTF13ebh3dFe/CoHe IMg I32.05811.18281.2083Mg I1.5033SN IIbSN 2011hs31dMg I1.7109Paγ1.0941Paβ1.28222SN IIPaα1.8756SN 2012aw45dBrγ2.16611He I1.0830PascheSN IInnSN IbnLSQ13ddu15d00.8Hsiao et al. 20191.0BraSN 2012ca70dckett1.21.41.6 1.8rest wavelength (µm)He I2.05812.0 2.2 2.4 2.6 2.8M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
CSP-II sample of IIb, Ib, Ic, Ic-BL Largest sample of SE-SNe NIRspectra109 spectra of 40 SNe7654321000.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08z25 Average redshift of 0.015 Best covered is SN 2013ge with18 spectra20NumberofofSNeNumberSNe NumberNumberofofSNeSNe8152014L2013ge10500 2 4 6 8 10 12 14 16 18Number of SpectraM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Why NIR?M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
What is the most interesting featurein our dataset?5 The strongest feature is λ 1.083 µm42013ak Is it He I?Is it C I?(why did I jump to carbon?)log10(F ) constantIIb313cumIb22014L Is it 2.4Rest Wavelength [µm]M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
OpticalvsNear-Infrared6 0 d 0 dCa IIPP5Ca IIP HePI2014btIIlog10(F ) constant4Ca IIOIHe I Mg IHe IHe I2013akIIbHe I3Ca IIHe IC I or He IOIMg IFe II13cumFe IIC I or He 02.22.4Rest Wavelength [µm]Modjaz et al. 2014Shahbandeh et al. 2019 in prepM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
66Ca II 0 dCa II 30 dPPPP55Ca IIP P HePICa II2014btHe IP2014btIICa IIOIHe I Mg IHe IHe I2013akIIbHe I3Ca IIMg I4He IC I or He IOIMg IFe II13cumIb2log10(F ) constantlog10(F ) constant4IICa IIOIHe I Mg IHe IHe IIIb3Mg I He ICa IIHe IC I or He IFe IIMg IFe IIC I or He IMg I1OICa IIIc2014LFe II C I or He IIc1Mg I2013dkMg II/Co IIIc-BL00.80.91.01.11.21.41.61.8Rest Wavelength .41.61.8Rest Wavelength [µm]2.02013dkIc-BL2.22.4
In order to identify the featureswe used three different methodsM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Method I:Fitting the feature and measuring velocities Using 2 gaussians to fit the two components Left component as either HV He I or C I Right component as He IDo the two components belong to the same element? Ortwo different elements?M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Absorption profileat λ1.05 µm gaussianExamples of doublecomponentsandfitsgaussian fitsIc-2 dfluxFlux2013ge1.021.031.041.051.061.07Ic85 001.011.02-2 d1.031.041.05Ib2012au0.991.0081 d1.011.021.031.041.051.061.07rest wavelength[µm]RestWavelength[µm]M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Velocity vs PhaseAbsorption profile at λ1.05 µm velocity measurements2013ge (Ib/c)2014az (Ib/IIb)2012au (Ib)2014L (Ic)2013ak (IIb)2014akx (IIb)180001800018000HV He 1.0830 µm1600016000[Km/s]16000C 1.0693 µm140001400014000velocityHe 1.0830 75phase ese identifications would imply that C is outside the He layer!M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9150
Method IIComparing the strongest features in the velocity space Compared Ib to Ic‣He I: λ 0.5876 µm, 1.083 µm, 2.058 µm‣C I: λ 0.9093 µm, 0.9658 µm,1.0693 µm, 1.1330 µm, 0.9658 µmUsing grotrian diagrams to confirm the resultsM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
What is a grotrian diagram? To choose the linesthat are more likely tohappen (ΔE) To eliminate theprofiles that don’tmatch in the velocityspaceM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
1.13300.90930.96581.0693CIM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Method IIIUsing the line strength, gf values or intensity gf values (LTE assumption) Line strength [ gf . exp (-ΔE/KT)]He I λ 1.083 µm vs 2.058 µm‣ The strength of λ 2.058 µm should be roughly half ofthe strength of λ 1.083 µmM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
4.54.0SN 2012au (Ib)SN 2013ak (IIb)3.5log10(F ) constant3.028 d22 d2.52.081 d70 d1.51.0318 d0.5255 d0.00.80.91.01.11.21.41.6Rest Wavelength [µm]1.82.02.22.5
5SN 2012au (Ib)SN 2014L (Ic)4log10(F ) constant28 d22 d331 d33 d281 d48 d1318 d121 d00.80.91.01.11.21.41.6Rest Wavelength [µm]1.82.02.22.5
LSQ13cum (Ib)SN 2013dk (Ic-BL)4log10(F ) constant36d-4 d212 d17 d122 d30 d00.80.91.01.11.21.41.6Rest Wavelength [µm]1.82.02.22.5
20 Carbon monoxide SN 2013ge18-2 d8d1617 d22 d1428 d Appears as early as78 daysPresent in 10spectra1240 dlog10(Fλ) constant 35 d48 d1053 d78 d885 d92 d697 d108 d4116 d123 d2152 d196 d00.80.91.01.11.21.41.61.82.02.22.5rest wavelength [µm]M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Why do we care about CO? CO formation timescale tells us where CO forms Velocity Local temperature and its time evolutionEffective cooler Dust formationM e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
How does CO profile change withtemperature?2000 K3000 K4000 KSharp, Hoeflich 1990M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Application to observations SN 2013ge Phase 123 days T 4000 K Radius 1014 cm Density gradient (n) 9 V 1000 Km/s SN 2014L Phase 121 days T 5000 K Radius 1014 cm Density gradient (n) 7 V 1000 Km/s[𝝆 r -n][𝝆 r -n]M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
Preliminary Conclusion NIR spectra provided new insight into the nature of SE-SNe There is a strong P Cygni profile at λ1.08 µm present in all of thespectra of this dataset We combined 3 different constraints to identify the strongest profiles We find that SN Ib shows He rich layers whereas in SN Ic we see C/O We identify the C I lines as an effective way to potentially determinethe C/O core mass in SE-SNe CO formation is common and can be used as diagnostics to lead dustformation studies.M e l i s s a S h a h b a n d e h , F. O. E , M a y 2 0 1 9
CSP-II Stripped-Envelope SNe Spectroscopy in the NIR Melissa Shahbandeh Florida State Univers
niques (except t-SNE) perform strongly on ar-tificial data sets but their visualization of real, high-dimensional data sets is poor. Whereas, t-SNE outperforms the other techniques and cap-tures most of the local structure while revealing global structure like presence of clusters at vari-ous scales. t-SNE uses Gaussian distribution for
X-TYPE 2001 - 2010 Engine Stripped - 2.5 V6 C2S34460E NEW** S-TYPE 1999 - 2008 Engine Stripped - 2.5 V6 XR842791E NEW** XJ 2010 Engine Stripped - 3.0 V6 Gen 2 C2D44392E NEW** XF 2016 Engine Stripped - 3.0 V6 Gen 2 C2D46701E NEW** F PACE Engine Stripped - 3.0 V6 Gen 2 T4A8683E XJ 2003 -
Additional Visualizations with t-SNE and PCA Figure 2: t-SNE (1st column) and PCA (2nd column) visualizations on Sines, and t-SNE (3rd column) and PCA (4th column) visualizations on Stocks. Each row provides the visualization for each of the 7 benchmarks, ordered as follows: (1) TimeGAN, (2) RCGAN, (3) C-RNN-GAN, (4) T-Forcing, (5)
VISUALIZING DATA USING T-SNE 2. Stochastic Neighbor Embedding Stochastic Neighbor Embedding (SNE) starts by converting the high-dimensional Euclidean dis-tances between datapoints into conditional probabilities that represent similarities.1 The similarity of datapoint xj to datapoint xi is the conditional probability, pjji, that xi would pick xj as its neighbor
Visualizing Data using t-SNE An Intuitive Introduction Simon Carbonnelle Universit e Catholique de Louvain, ICTEAM 12th of May, 2016. Visualization and Dimensionality Reduction Intuition behind t-SNE Visualizing representations. Visualization and Dimensionality Reduction
when applied to high-dimensional clustered data, t-SNE tends to produce a visualization with more separated clusters, which are often in good agreement with the clusters found by a dedicated clustering algorithm (Kobak and Berens, 2019). See Figure 1 for an example of data visualization using such a basic t-SNE algorithm.
PID/SKU CSP-5216 CSP-5228 CSP-5436 CSP-5444 CSP-5456 Rack Size 1RU 2RU CPU Cores 16 28 36 44 56 On Board Processors 2 Redundant Power (110/220 VAC)
American Math Competition 8 Practice Test 8 89 American Mathematics Competitions Practice 8 AMC 8 (American Mathematics Contest 8) INSTRUCTIONS 1. DO NOT OPEN THIS BOOKLET UNTIL YOUR PROCTOR TELLS YOU. 2. This is a twenty-five question multiple choice test. Each question is followed by
