Multi-wavelength Properties of Supernova Remnants

Transcription

1 Multi-wavelength Properties of Supernova Remnants Jacco Vink University of Amsterdam Anton Pannekoek Institute/GRAPPA

2 Supernova classification Simple CSM (?) But see Kepler (Chiotellis+ 12) Complex CSM: - strong winds ϱ 1/r 2 -starforming regions & MCs 2

3 SNR structure & evolution Two shock structure: blast wave in CSM reverse shock heating ejecta (stellar debris) Truelove & McKee 99 Once Mej << Mswept up: self-similar Sedov-Taylor phase Late phase radiative Early on reverse shock moves outward Approaching the ST phase it starts moving to center Rev. shock leaves a hot low density center 3

4 Non-equilibrium ionization Densities in SNR plasmas low (ne= cm -3 ) Plasmas recently shocked (~100-~10000 yr) often not enough time to reach equilibrium ionization (ion. rate=recomb. rate) 4

5 An age sequence of Type Ia SNRs SNR 0509 SNR 0519 Dem L71 Evolutionary sequence higher Fe ionization, more Fe is shocked (reverse shock further in) Chandra images (Fe in green) XMM/RGS high resolution X-ray spectra Warren et al Kosenko, et al.2008 Van der Heyden

6 Linking young SNRs with SNe: X-rays G oxygen neon silicon Core collapse SNRs are rich in O, Ne, Mg Type Ia SNRs are iron-rich

7 Mature SNRs For low V/T<5x10 5 K: cooling becomes very strong (oxygen line emission) Downstream of shock: rapid cooling, energy loss, very strong compression radiative shocks Schure et al Low plasma temperatures: bright emission from optical/uv forbidden lines a.o H-alpha line blended with [NII] SNR no longer adiabatic: R~t 0.25 (momentum conservation) 7

8 Mixed-morphology/Thermal composite SNRs W28 (ROSAT/VLA) W44 AGILE: Guiliani+ 11 -Many MM SNRs detected by Fermi/Agile - W28 has offset HESS source: CR escape (Aharonian c.s. ;91,96, 01,...) Mixed-morphology SNRs are centrally bright in X-ray and shell-type in radio X-ray emission (center) is thermal, and often enhanced in abundances Evidence for over-ionization! adiabatic cooling (e.g Yamaguchi+ 09) Idea: shell too cool for X-rays, but center hot enough for X-rays (Cox+ 99) Shell bright in radio (compressed electron cosmic rays, vd Laan mechanism?) Many associations with masers 8

9 Young versus Old SNRs in radio SN1006 PKS Dickel & Milne 96 Radial magnetic fields Emission due to recently accelerated electrons Tangential magnetic fields Flux can be explained by Van der Laan mechanism (compression of pre-existing electron cosmic rays) 9

10 Shocks, temperatures & cosmic rays A strong shock in a monatomic gas heats the plasma to kt i = 2(γ 1) (γ +1) 2 m ivs 2 = 3 ( 16 m ivs 2 mi )( V ) 2 s =2.0 kev m p 1000 km/s For young SNRs (Vs>3000 km/s) expect kt > 10 kev No SNR known with kt > 4 kev!! Possible solution: non-equilibration: electrons are colder than protons! Alternative: efficient cosmic ray acceleration (e.g. Hughes+ 00) X-rays: measure in general electron temperature not proton temperature! 10

11 Temperature (Non-)Equilibration Simplified plane parallel shock model, no equilibration: Protons Electrons Also scales with net!! 11

12 Proton temperatures Comparing electron and proton temperatures Proton temperature: thermal Doppler broadening of Hα Electron temperature: X-ray spectra Tprotons > Telectrons only for Vs> 300 km/s Ghavamian

13 Shock heating & acceleration: 2 fluid approach Standard shock continuity relations with some modifications: - Three regions 0: far upstream;1: precursor, just upstream of main shock; 3: shocked gas - cosmic ray pressure continuous from cosmic rays can escape (no energy flux conservation!)

14 Shock heating and compression Compression ratio Energy/cosmic ray escape Agrees with kinetic/monte carlo simulations (Ellison, Blasi, Jones) - disadvantage: have to assume adiabatic index - advantage: no assumption on how particles are accelerated Efficiency (cr pressure/total pressure): g (1 1 )+ gm w = g M

15 Limits on particle acceleration For relativistic dominated particles (γ=4/3): Mach nr M > 5.88 Relaxed for non-relativistic cosmic rays to M > 5=2.24 Note different behavior of curves for γ=4/3 and γ=5/3 (no solutions with low w and low escape) No efficient acceleration for M<6: important implications for clusters of galaxies and heliospheric shocks!!! 15

16 Plasma Temperatures correction w.r.t. standard standard result kt = 3 16 mv s 2 Vink et al Post-shock temperatures much lower in cosmic-ray dominated shocks 16

17 X-ray synchrotron filaments Cas A Tycho Kepler SN 1006 All historical shell SNRs emit X-ray synchrotron (1st detected SN1006: Koyama et al. 95) Implies Emax,electron ~ TeV Requires fast shocks (hνmax~0.5(v/2000) 2 η -1 kev -> Aharonian&Atoyan 99) Emission regions often narrow (few ) Given rapid cooling times -> active accelerations Evidence for small diffusion coefficients (rapid acc.) 17

18 X-ray synchrotron profiles Fits with spherical models with exponential synchr. emissivity profiles Fits not that well (except Vela Jr) Helder,JV,

19 Interpreting narrow X-ray rims Rim widths determined by interplay diffusion, advection and synchrotron losses length scale = v syn / B? h ch = µg v B 2 E E 100 TeV 2 kev Combining diffusion and advection: l adv 2/3 1/3 B cm 1 1/3 µg 4 Rim width can be used to measure B-field: B 110 (L/10 17 cm) -2/3 μg Vink&Laming 03 Cas A/Tycho/Kepler: ~ μg (Vink&Laming 03, Berezhko+ 03, Völk+ 04, Bamba+ 04, Warren+ 05, Parizot+ 06, Helder+ 12) High B fast acceleration protons beyond ev? High B-field likely induced by cosmic rays (e.g. Bell 04) High B-fields are a signature of efficient acceleration 19

20 Tycho s SNR (western rim) uniform exponential shell profile + precursor Exponential profile not a good fit Much better (but still preliminary): -add synchrotron precursor (with X=2.5) - requires small B-field jump sub-shock with low compression - alternatives: deviations from spherical symmetry Vink 2012 See also Bamba

21 Cas A reverse shock 4-6keV Deprojection Γ= -3.2 Spectral index: 2 regions of hard emission: X-ray synchrotron emission Deprojection: Most X-ray synchrotron from reverse shock! Surprising: ejecta should have low magnetic field! Prominence of West: No expansion ejecta shocked with V>6000km/s Helder&Vink 08 Final amplified B-field insensitive to initial field!? Uchiyama

22 Magnetic Field Amplification There is a clear correlation between ρ, V and B, in rough agreement with theoretical predictions (e.g. Bell 2004) Relation may even extend to supernovae (B 2 ρvs 3?) (Völk+ 05, Vink 08) 22

23 X-ray synchrotron decline of Cas A X-ray synchrotron flux (4-6 kev) declines strongly: Whole SNR: (1.5 ± 0.17)% yr 1 Western part: (1.9 ± 0.10)% yr 1 Accompanied by steepening of spectral index Γ Critical check: no decline in line rich band (1.5-3 kev) no 4-6 kev decline cluster A1795 Decline more than in radio: not adiabatic cooling Likely cause: shock deceleration 1 F ( ) df ( ) dt = 2 d dt r d c dt = 4 c c d dt Confirms basic interpretation of synchrotron model Decline somewhat high, may imply small diffusion coefficient, hence very fast acceleration Patnaude, Vink, Laming, Fesen,

24 Heating vs Acceleration GeV/TeV + X-ray synchrotron: particles accelerated to high energies (but not yet the knee or beyond) CR efficiency seems to be near or below required 10% efficiency of total expl. energy Contradicts other signs of efficient acceleration: B-field amplification, high compression ratios (Tycho), curved spectra What about another consequence of efficient acceleration: low plasma temperatures? 24

25 Balmer lines from non-radiative shocks Neutral hydrogen entering shock: - direct ionization no line emission - excitation followed by ionization narrow line hydrogen emission - charge exchange (H+p p+h*) H* has kt of shocked plasma May also be important for shock-precursor (Alfven damping)! (Raymond 11,Blasi+ 12) X-ray synchrotron Compression/pre-heating Heng 2010 Fig: Eveline Helder Thermal X-ray Thin layer (~10 15 /np cm): neutrals excite, charge X-change, ionize 25

26 Hα from X-ray synchrotron region RCW 86 X-ray (XMM) Vink+ 06 RCW 86 likely SN185 RCW 86 is a TeV source, but not GeV (Aharonian+ 08, Lemoine-Goumard+ 12) Non-radiative H-α emission from several shock locations in RCW 86 Faint H-α associated with X-ray emitting region in NE -> suggests high Vshock VLT H-α spectroscopy: FWHM=1100 ± 63 km/s ktp = 2.2 kev X-ray expansion measurement: Vs = (5900±1200) d2.5 km/s Discrepancy: >50% shock acceleration efficiency (Helder, JV+ 09) 26

27 New expansion measurement RCW 86 Helder, JV+ in prep. New expansion measurements based on two H-α images ( ) Two regions NE (X-ray synchrotron) and SE (no X-ray synchrotron) Both regions same H-α width -> same plasma temperature H-α proper motion for NE H-α filament ~1400 d2.5km/s For SE similar but somewhat lower Conclusion: - no need for (very) efficient acceleration - puzzle: why does NE (low density) have X-ray synchrotron and SE (denser) not? - was X-ray proper motion wrong, or is H-α from denser, slowed down regions? 27

28 H-ɑ from fastest known SNR shock SNR (LMC) Helder, Kosenko, Vink 10 Distance known (LMC, 50 kpc) Shock velocity: X-ray line broadening + Chandra expansion: Vs> 5000 km/s One of the fastest shocks in a known SNR J.P. Hughes private communication Vs=6500 km/s H-alpha broad line widths: 2680 ± 70 km/s (SW), 3900 ± 800 km/s 28

29 LMC SNR Southwest: ktp=16 kev, expect ~50 kev Curves: model by Van Adelsberg+ 08 Preliminary! (Arne de Laat master s thesis) -Inferred CR fractional pressure SW: w>15%, -But perhaps w >50% if indeed Vs=6500 km/s -Needed: better models for Hα emission with cosmic rays (see Blasi+ 2012) Helder, Kosenko, Vink 10 29

30 GeV emission from young SNRs Recall: Broad spectral range needed for distinguishing leptonic vs hadronic radiation Fermi detected several young SNRs: Cas A, Tycho, RX J1713,... RX J1713 ->leptonic domination likely Cas A -> undecided Tycho -> hadronic likely Only Tycho seems to have ~10% of energy in CRs (e.g. Aciari+ 11, Morlino&Caprioli 12) Abdo+ 10 RX J1713 (Fermi) Tycho Cas A Giordano+ 12 Abdo

31 Cas A versus Tycho Cas A may have been/be the better accelerator! Escape likely important Escape in Cas A in 1/r 2 profile: less interactions of cosmic rays Type Ia Core collapse (SN II(b)) Density structure (R) = constant (R) = Ṁ 4 R 2 v w Emission from escaped CRs (with diffusion radius Rd) L 0 / Z Rdiff 0 n cr n H (R)4 R 2 dr L 0 / N cr n H L 0 / 3N cr Ṁ w /(v w R 2 di ) - In stellar winds: Luminosity evolution 1) total number of CRs 2) decreases with time Rd ~t 1/2 - Type Ia: emission only depends on Ncr, surface brightness limits detection - For Tycho pion decay emission may come partly from ambient medium! Vink 12 (arxiv: ) 31

32 Summary Supernova remnants are interesting in many ways: - explosion physics - chemical evolution - plasma physics - cosmic-ray acceleration Multi-wavelengths aspects important for all: - Gamma-rays: see many other talks during this conference: - X-rays: thermal: composition, electron temperatures, ionization synchrotron: -proof of active acceleration -B-field amplification or even creation (talk by Tony Bell) - Optical observations (H-alpha) some evidence for low temperatures: non-linear acceleration more modeling needed 32