Neutrinos and the stars Supernova Neutrinos Georg Raffelt, MPI for Physics Lectures at the Topical Seminar Neutrino Physics & Astrophysics 17-21 Sept 2008, Beijing, China

Sanduleak -69 202 Supernova 1987A 23 February 1987 Tarantula Nebula Large Magellanic Cloud Distance 50 kpc (160.000 light years)

Supernova Neutrinos 20 Jahre nach SN 1987A

Crab Nebula

Stellar Collapse and Supernova Explosion Onion structure Main-sequence star Hydrogen Burning Collapse (implosion) Helium-burning star Helium Burning Hydrogen Degenerate iron core: r  109 g cm-3 T  1010 K MFe  1.5 Msun RFe  8000 km

Stellar Collapse and Supernova Explosion Newborn Neutron Star ~ 50 km Proto-Neutron Star r  rnuc = 3  1014 g cm-3 T  30 MeV Collapse (implosion) Neutrino Cooling

Stellar Collapse and Supernova Explosion Newborn Neutron Star ~ 50 km Proto-Neutron Star r  rnuc = 3  1014 g cm-3 T  30 MeV Neutrino Cooling Gravitational binding energy Eb  3  1053 erg  17% MSUN c2 This shows up as 99% Neutrinos 1% Kinetic energy of explosion (1% of this into cosmic rays) 0.01% Photons, outshine host galaxy Neutrino luminosity Ln  3  1053 erg / 3 sec  3  1019 LSUN While it lasts, outshines the entire visible universe

Neutrino Signal of Supernova 1987A Kamiokande-II (Japan) Water Cherenkov detector 2140 tons Clock uncertainty 1 min Irvine-Michigan-Brookhaven (US) Water Cherenkov detector 6800 tons Clock uncertainty 50 ms Baksan Scintillator Telescope (Soviet Union), 200 tons Random event cluster ~ 0.7/day Clock uncertainty +2/-54 s Within clock uncertainties, signals are contemporaneous

SN 1987A Event No.9 in Kamiokande Kamiokande Detector Hirata et al., PRD 38 (1988) 448

Thermonuclear vs. Core-Collapse Supernovae Thermonuclear (Type Ia) Core collapse (Type II, Ib/c) Carbon-oxygen white dwarf (remnant of low-mass star) Accretes matter from companion Degenerate iron core of evolved massive star by nuclear burning at its surface Chandrasekhar limit is reached - MCh  1.5 Msun (2Ye)2 C O L L A P S E S E T S I N Nuclear burning of C and O ignites  Nuclear deflagration (“Fusion bomb” triggered by collapse) Collapse to nuclear density Bounce & shock Implosion  Explosion Powered by gravity Powered by nuclear binding energy Gain of nuclear binding energy ~ 1 MeV per nucleon Gain of gravitational binding energy ~ 100 MeV per nucleon 99% into neutrinos Comparable “visible” energy release of ~ 3  1051erg

Supernova Neutrinos 20 Jahre nach SN 1987A Explosion Mechanism for Core-Collapse SNe

Collapse and Prompt Explosion Velocity Density Movies by J.A.Font, Numerical Hydrodynamics in General Relativity http://www.livingreviews.org Supernova explosion primarily a hydrodynamical phenomenon

Why No Prompt Explosion? Dissociated Material (n, p, e, n) Collapsed Core Undissociated Iron Shock Wave 0.1 Msun of iron has a nuclear binding energy  1.7  1051 erg Comparable to explosion energy Shock wave forms within the iron core Dissipates its energy by dissociating the remaining layer of iron

Neutrinos to the Rescue Neutrino heating increases pressure behind shock front Picture adapted from Janka, astro-ph/0008432

Supernova Delayed Explosion Scenario

Standing Accretion Shock Instability (SASI) Mezzacappa et al., http://www.phy.ornl.gov/tsi/pages/simulations.html

Gravitational Waves from Core-Collapse Supernovae Müller, Rampp, Buras, Janka, & Shoemaker, “Towards gravitational wave signals from realistic core collapse supernova models,” astro-ph/0309833 Asymmetric neutrino emission Bounce Convection The gravitational-wave signal from convection is a generic and dominating feature

Supernova Neutrinos 20 Jahre nach SN 1987A Some Particle-Physics Lessons from SN 1987A

Neutrino Mass Sensitivity by Signal Dispersion Time-of-flight delay of massive neutrinos SN 1987A (50 kpc) E  20 MeV, Dt  10 s Simple estimate or detailed maximum likelihood analysis give similar results mn ≲ 20 eV Future Galactic SN at 10 kpc (Super-K) Rise-time of signal ~ 10 ms (Totani, PRL 80:2040, 1998) mn ~ 3 eV Full signal (Nardi & Zuluaga, NPB 731:140, 2005) mn ~ 1 eV With late black-hole formation Cutoff “infinitely” fast (Beacom et al., PRD 63:073011, 2001) mn ~ 2 eV Future SN in Andromeda (Megatonne) D  750 kpc, Dt  10 s few tens of events mn ~ 1-2 eV

Early Lightcurve of SN 1987A Expected bolometric brightness evolution Expected visual brightness evolution Neutrinos several hours before light Adapted from Arnett et al., ARAA 27 (1989)

Do Neutrinos Gravitate? Neutrinos arrive a few hours earlier than photons  Early warning (SNEWS) SN 1987A: Transit time for photons and neutrinos equal to within ~ 3h Shapiro time delay for particles moving in a gravitational potential Longo, PRL 60:173,1988 Krauss & Tremaine, PRL 60:176,1988 Equal within ~ 1 - 4 10-3 Proves directly that neutrinos respond to gravity in the usual way because for photons gravitational lensing already proves this point Cosmological limits DNn ≲ 1 much worse test of neutrino gravitation Provides limits on parameters of certain non-GR theories of gravitation Photons likely obscured for next galactic SN, so this result probably unique to SN 1987A

The Energy-Loss Argument Neutrino sphere SN 1987A neutrino signal Neutrino diffusion Late-time signal most sensitive observable Emission of very weakly interacting particles would “steal” energy from the neutrino burst and shorten it. (Early neutrino burst powered by accretion, not sensitive to volume energy loss.) Volume emission of novel particles

Too much hot dark matter Axion Bounds 103 106 109 1012 [GeV] fa eV keV meV ma Experiments Tele scope CAST Direct search ADMX Too much hot dark matter Too much cold dark matter Globular clusters (a-g-coupling) Too many events Too much energy loss SN 1987A (a-N-coupling)

Sterile Neutrinos sin2(2Qes) ≲ 3  10-10 Active-sterile mixing Electron neutrino appears as sterile neutrino in ½ sin2(2Qes) of all cases Average scattering rate in SN core involving ordinary left-handed neutrinos To avoid complete energy loss in ~ 1 s sin2(2Qes) ≲ 3  10-10

Sterile Neutrino Limits See also: Maalampi & Peltoniemi: Effects of the 17-keV neutrino in supernovae PLB 269:357,1991 Hidaka & Fuller: Dark matter sterile neutrinos in stellar collapse: alteration of energy/lepton number transport and a mechanism for supernova explosion enhancement PRD 74:125015,2006

Supernova 1987A Limit on Large Extra Dimensions SN core emits large flux of KK gravity modes by nucleon-nucleon bremsstrahlung Large multiplicity of modes RT ~ 1011 for R ~ 1 mm, T ~ 30 MeV Cullen & Perelstein, hep-ph/9904422 Hanhart et al., nucl-th/0007016 SN 1987A energy-loss argument: R < 1 mm, M > 9 TeV (n = 2) R < 1 nm, M > 0.7 TeV (n = 3) Originally the most restrictive limit on such theories, except for cosmological arguments

Supernova Neutrinos 20 Jahre nach SN 1987A Neutrinos from the Next Galactic Supernova

Local Group of Galaxies Events in a detector with 30 x Super-K fiducial volume, e.g. Hyper-Kamiokande 30 60 250

Core-Collapse SN Rate in the Milky Way SN statistics in external galaxies Core-collapse SNe per century 1 2 3 4 5 6 7 8 9 10 van den Bergh & McClure (1994) Cappellaro & Turatto (2000) Gamma rays from 26Al (Milky Way) Diehl et al. (2006) Historical galactic SNe (all types) Strom (1994) Tammann et al. (1994) No galactic neutrino burst 90 % CL (25 y obserservation) Alekseev et al. (1993) References: van den Bergh & McClure, ApJ 425 (1994) 205. Cappellaro & Turatto, astro-ph/0012455. Diehl et al., Nature 439 (2006) 45. Strom, Astron. Astrophys. 288 (1994) L1. Tammann et al., ApJ 92 (1994) 487. Alekeseev et al., JETP 77 (1993) 339 and my update.

Nearby Galaxies with Many Observed Supernovae M83 (NGC 5236, Southern Pinwheel) D = 4.5 Mpc NGC 6946 D = (5.5 ± 1) Mpc Observed Supernovae: 1923A, 1945B, 1950B, 1957D, 1968L, 1983N Observed Supernovae: 1917A, 1939C, 1948B, 1968D, 1969P, 1980K, 2002hh, 2004et, 2008S

Large Detectors for Supernova Neutrinos MiniBooNE (200) LVD (400) Borexino (100) Baksan (100) Super-Kamiokande (104) KamLAND (400) In brackets events for a “fiducial SN” at distance 10 kpc IceCube (106)

SuperNova Early Warning System (SNEWS) Neutrino observation can alert astronomers several hours in advance to a supernova. To avoid false alarms, require alarm from at least two experiments. Super-K IceCube Coincidence Server @ BNL Alert LVD Supernova 1987A Early Light Curve Others ? http://snews.bnl.gov astro-ph/0406214

Simulated Supernova Signal at Super-Kamiokande Accretion Phase Kelvin-Helmholtz Cooling Phase Simulation for Super-Kamiokande SN signal at 10 kpc, based on a numerical Livermore model [Totani, Sato, Dalhed & Wilson, ApJ 496 (1998) 216]

Supernova Pointing with Neutrinos 95% CL half-cone opening angle Neutron tagging efficiency None 90 % SK 7.8º 3.2º SK  30 1.4º 0.6º Beacom & Vogel: Can a supernova be located by its neutrinos? [astro-ph/9811350] Tomàs, Semikoz, Raffelt, Kachelriess & Dighe: Supernova pointing with low- and high-energy neutrino detectors [hep-ph/0307050]

IceCube as a Supernova Neutrino Detector Each optical module (OM) picks up Cherenkov light from its neighborhood. SN appears as “correlated noise”. About 300 Cherenkov photons per OM from a SN at 10 kpc Noise < 500 Hz Total of 4800 OMs in IceCube IceCube SN signal at 10 kpc, based on a numerical Livermore model [Dighe, Keil & Raffelt, hep-ph/0303210] Method first discussed by Halzen, Jacobsen & Zas astro-ph/9512080

LAGUNA - Approved FP7 Design Study Large Apparati for Grand Unification and Neutrino Astrophysics (see also arXiv:0705.0116)

Supernova Neutrinos 20 Jahre nach SN 1987A Neutrinos From All Cosmic Supernovae

Diffuse Background Flux of SN Neutrinos 1 SNu = 1 SN / 1010 Lsun,B / 100 years Lsun,B = 0.54 Lsun = 2  1033 erg/s En ~ 3  1053 erg per core-collapse SN 1 SNu ~ 4 Ln / Lg,B Average neutrino luminosity of galaxies ~ photon luminosity Photons come from nuclear energy Neutrinos from gravitational energy For galaxies, average nuclear & gravitational energy release comparable Present-day SN rate of ~ 1 SNu, extrapolated to the entire universe, corresponds to ne flux of ~ 1 cm-2 s-1 Realistic flux is dominated by much larger early star-formation rate  Upper limit ~ 54 cm-2 s-1 [Kaplinghat et al., astro-ph/9912391]  “Realistic estimate” ~ 10 cm-2 s-1 [Hartmann & Woosley, Astropart. Phys. 7 (1997) 137] Measurement would tell us about early history of star formation

Experimental Limits on Relic Supernova Neutrinos Super-K upper limit 29 cm-2 s-1 for Kaplinghat et al. spectrum [hep-ex/0209028] Upper-limit flux of Kaplinghat et al., astro-ph/9912391 Integrated 54 cm-2 s-1 Cline, astro-ph/0103138

DSNB Measurement with Neutron Tagging Future large-scale scintillator detectors (e.g. LENA with 50 kt) Inverse beta decay reaction tagged Location with smaller reactor flux (e.g. Pyhäsalmi in Finland) could allow for lower threshold Pushing the boundaries of neutrino astronomy to cosmological distances Beacom & Vagins, hep-ph/0309300 [Phys. Rev. Lett., 93:171101, 2004]

Supernova Neutrinos 20 Jahre nach SN 1987A Oscillations of Supernova Neutrinos

Structure of Supernova Neutrino Signal 1. Collapse (infall phase) 2. Shock break out 3. Matter accretion 4. Kelvin-Helmholtz cooling Traps neutrinos and lepton number of outer core layers

Neutronization Burst as a Standard Candle Different Mass Neutrino Transport Nuclear EoS If mixing scenario is known, perhaps best method to determine SN distance, especially if obscured (better than 5-10%) Kachelriess, Tomàs, Buras, Janka, Marek & Rampp, astro-ph /0412082

Flavor-Dependent Fluxes and Spectra Prompt ne deleptonization burst Broad characteristics Duration a few seconds En ~ 10-20 MeV En increases with time Hierarchy of energies Approximate equipartition of energy between flavors nx _ However, in traditional simulations transport of nm and nt schematic Incomplete microphysics Crude numerics to couple neutrino transport with hydro code ne ne Livermore numerical model ApJ 496 (1998) 216

Flavor-Dependent Neutrino Fluxes vs. Equation of State Wolff & Hillebrandt nuclear EoS (stiff) Lattimer & Swesty nuclear EoS (soft) Kitaura, Janka & Hillebrandt, “Explosions of O-Ne-Mg cores, the Crab supernova, and subluminous Type II-P supernovae”, astro-ph/0512065

Level-Crossing Diagram in a SN Envelope Normal mass hierarchy Inverted mass hierarchy Dighe & Smirnov, Identifying the neutrino mass spectrum from a supernova neutrino burst, astro-ph/9907423

Spectra Emerging from Supernovae Primary fluxes for After leaving the supernova envelope, the fluxes are partially swapped Normal Inverted sin2(2Q13) ≲ 10-5 ≳ 10-3 Any Mass ordering sin2(Q12)  0.3 cos2(Q12)  0.7 Case A B C Survival probability

Oscillation of Supernova Anti-Neutrinos Measured spectrum at a detector like Super-Kamiokande Assumed flux parameters Flux ratio No oscillations Earth effects included Oscillations in SN envelope Mixing parameters P(Dighe, Kachelriess, Keil, Raffelt, Semikoz, Tomàs), hep-ph/0303210, hep-ph/0304150, hep-ph/0307050, hep-ph/0311172

Model-Independent Strategies for Observing Earth Effects One detector observes SN shadowed by Earth Case 1: Another detector observes SN directly Identify Earth effects by comparing signals Case2: Identify “wiggles” in signal of single detector Problem: Smearing by limited energy resolution If 13-mixing angle is known to be “large”, e.g. from Double Chooz, observed “wiggles” in energy spectrum signify normal mass hierarchy Scintillator detector ~ 2000 events may be enough Water Cherenkov Need megaton detector with ~ 105 events Dighe, Keil & Raffelt, “Identifying Earth matter effects on supernova neutrinos at a single detector” [hep-ph/0304150]

Supernova Shock Propagation and Neutrino Oscillations Schirato & Fuller: Connection between supernova shocks, flavor transformation, and the neutrino signal [astro-ph/0205390] Resonance density for R. Tomàs, M. Kachelriess, G. Raffelt, A. Dighe, H.-T. Janka & L. Scheck: Neutrino signatures of supernova forward and reverse shock propagation [astro-ph/0407132]

Shock-Wave Propagation in IceCube Inverted Hierarchy No shockwave Inverted Hierarchy Forward & reverse shock Inverted Hierarchy Forward shock Normal Hierarchy Choubey, Harries & Ross, “Probing neutrino oscillations from supernovae shock waves via the IceCube detector”, astro-ph/0604300

Supernova Neutrinos 20 Jahre nach SN 1987A Collective Supernova Neutrino Oscillations

Neutrino Density Streaming off a Supernova Core Typical luminosity in one neutrino species Corresponds to a neutrino number density of Current-current structure of weak interaction causes suppression of effective potential for collinear-moving particles Nu-nu refractive effect decreases as Appears to be negligible Equivalent Neutrino density ∝ R-2 Nu-nu refraction ∝ R-4

Collective Effects in Neutrino Flavor Oscillations Collapsed supernova core or accretion torus of merging neutron stars: Neutrino flux very dense: Up to 1035 cm-3 Neutrino-neutrino interaction energy much larger than vacuum oscillation frequency Large “matter effect” of neutrinos on each other Non-linear oscillation effects Assume 80% anti-neutrinos Vacuum oscillation frequency w = 0.3 km-1 Neutrino-neutrino interaction energy at nu sphere (r = 10 km) m = 0.3105 km-1 Falls off approximately as r-4 (geometric flux dilution and nus become more co-linear)

Self-Induced Flavor Oscillations of SN Neutrinos Survival probability ne Normal Hierarchy atm Dm2 Q13 close to Chooz limit Inverted No nu-nu effect Realistic nu-nu effect MSW effect Realistic nu-nu effect Bipolar collective oscillations (single-angle approximation) MSW

Mass Hierarchy at Extremely Small Theta-13 Using Earth matter effects to diagnose transformations Ratio of spectra in two water Cherenkov detectors (0.4 Mton), one shadowed by the Earth, the other not Dasgupta, Dighe & Mirizzi, arXiv:0802.1481

Collective SN neutrino oscillations 2006-2008 (I) “Bipolar” collective transformations important, even for dense matter Duan, Fuller & Qian astro-ph/0511275 Numerical simulations Including multi-angle effects Discovery of “spectral splits” Duan, Fuller, Carlson & Qian astro-ph/0606616, 0608050 Pendulum in flavor space Collective pair annihilation Pure precession mode Hannestad, Raffelt, Sigl & Wong astro-ph/0608695 Duan, Fuller, Carlson & Qian astro-ph/0703776 Self-maintained coherence vs. self-induced decoherence caused by multi-angle effects Sawyer, hep-ph/0408265, 0503013 Raffelt & Sigl, hep-ph/0701182 Esteban-Pretel, Pastor, Tomàs, Raffelt & Sigl, arXiv:0706.2498 Theory of “spectral splits” in terms of adiabatic evolution in rotating frame Raffelt & Smirnov, arXiv:0705.1830, 0709.4641 Duan, Fuller, Carlson & Qian arXiv:0706.4293, 0707.0290 Independent numerical simulations Fogli, Lisi, Marrone & Mirizzi arXiv:0707.1998

Collective SN neutrino oscillations 2006-2008 (II) Three-flavor effects in O-Ne-Mg SNe on neutronization burst (MSW-prepared spectral double split) Duan, Fuller, Carlson & Qian, arXiv:0710.1271 Dasgupta, Dighe, Mirrizzi & Raffelt, arXiv:0801.1660 Theory of three-flavor collective oscillations Dasgupta & Dighe, arXiv:0712.3798 Identifying the neutrino mass hierarchy at extremely small Theta-13 Dasgupta, Dighe & Mirizzi, arXiv:0802.1481 Second-order mu-tau refractive effect important in three-flavor context Esteban-Pretel, Pastor, Tomàs, Raffelt & Sigl, arXiv:0712.1137 But for high density, conversions suppressed by geometric effect Esteban-Pretel, Mirizzi, Pastor, Tomàs, Raffelt, Serpico & Sigl, arXiv:0807.0659 Collective oscillations along flux lines for non-spherical geometry Dasgupta, Dighe, Mirizzi & Raffelt, arXiv:0805.3300

Neutrino Oscillations in a Neutrino Background Neutrinos in a medium suffer flavor-dependent refraction (Wolfenstein, PRD 17:2369, 1978) f W, Z f Z n n n n If neutrinos form the background, the refractive index has “offdiagonal elements” (Pantaleone, PLB 287:128, 1992) n Z n n One can not operationally distinguish between “beam” and “background” Problem is fundamentally nonlinear

Matrices of Density in Flavor Space Neutrino quantum field Spinors in flavor space Destruction operators for (anti)neutrinos Variables for discussing neutrino flavor oscillations Quantum states (amplitudes) “Matrices of densities” (analogous to occupation numbers) Neutrinos Anti- neutrinos Sufficient for “beam experiments” “Quadratic” quantities, required for dealing with decoherence, collisions, Pauli-blocking, nu-nu-refraction, etc.

General Equations of Motion Vacuum oscillations M is neutrino mass matrix Note opposite sign between neutrinos and antineutrinos Usual matter effect with Nonlinear nu-nu effects are important when nu-nu interaction energy exceeds typical vacuum oscillation frequency (Do not compare with matter effect!)

Oscillations of Neutrinos plus Antineutrinos in a Box Equal and densities, single energy E, with ≫ Equal self terms Opposite vacuum oscillations “Pendulum in flavor space” Inverted mass hierarchy  Inverted pendulum  Unstable even for small mixing angle Normal mass hierarchy  Small-amplitude oscillations

Flavor Conversion Without Flavor Mixing? Equal ne and ne densities in a box (inverted hierarchy) _ Inverted pendulum: Time to fall depends logarithmically on small initial angle Q Stays up forever only for Q = 0 Unstable by quantum uncertainty relation (“How long can a pencil stand on its tip?”) This is no real “flavor conversion”, rather a “coherent pair conversion” Occurs anyway at second order GF Coherent “speed-up effect” (Sawyer) Not clear (to me) if coherent transformations can be triggered by quantum fluctuations alone (mixing angle Q = 0)

Supernova Neutrino Conversion Neutrinos in a box Permanent pendular oscillations Neutrinos streaming off a supernova core Complete conversion Nu-nu interaction energy decreases Pendulum’s moment of inertia m-1 increases Conservation of angular momentum  kinetic energy decreases  amplitude decreases ∝ m1/2 Envelope declines as ∝ m1/2 ∝ r -2

Flavor Conversion in Toy Supernova Assume 80% anti-neutrinos Vacuum oscillation frequency w = 0.3 km-1 Neutrino-neutrino interaction energy at nu sphere (r = 10 km) m = 0.3105 km-1 Falls off approximately as r-4 (geometric flux dilution and nus become more co-linear) Pendular Oscillations Decline of oscillation amplitude explained in pendulum analogy by inreasing moment of inertia (Hannestad, Raffelt, Sigl & Wong astro-ph/0608695)

Synchronized vs. Pendular Oscillations Ensemble of unequal densities (antineutrino fraction a < 1) Equal energies (equal oscillation frequency w = Dm2/2E) Interaction energy Synchronized oscillations ≪ Pendular oscillations ≪ Free oscillations ≪

Synchronized vs. Pendular Oscillations Supernova Core R = 40-60 km R  200 km Synchronized oscillations ≪ Pendular oscillations ≪ Free oscillations ≪

Pendulum in Flavor Space Polarization vector for neutrinos plus antineutrinos Precession (synchronized oscillation) Nutation (pendular oscillation) Very asymmetric system - Large spin - Almost pure precession - Fully synchronized oscillations Perfectly symmetric system - No spin - Simple spherical pendulum - Fully pendular oscillation [Hannestad, Raffelt, Sigl, Wong: astro-ph/0608695] Spin (Lepton Asymmetry) ≫ Mass direction in flavor space

Multi-Energy and Multi-Angle Effects Different modes oscillate with different frequencies  kinematical decoherence Self-maintained coherence by nu-nu interactions Can lead to “spectral split” Multi-angle effects for non-isotropic nu distribution (streaming from SN): Different modes should oscillate differently  kinematical decoherence However, nu-nu interaction can lead to “Angular synchronization” (quasi-single angle behavior) Self-accelerated multi-angle decoherence Isotropic matter background affects all modes the same

Spectral Split (Stepwise Spectral Swapping) Initial fluxes at nu sphere After collective trans- formation For explanation see Raffelt & Smirnov arXiv:0705.1830 0709.4641 Duan, Fuller, Carlson & Qian arXiv:0706.4293 0707.0290 Fogli, Lisi, Marrone & Mirizzi, arXiv:0707.1998

Spectral split in terms of the w variable Collective conversion of thermal spectra of ne and ne as in a supernova _ Energy spectrum Spectrum in terms of w = Dm2/2E Flavor lepton number conservation: Equal integrals Raffelt & Smirnov, arXiv:0709.4641

Lots of theoretical work to do! SN 1006 No problem May take a long time Lots of theoretical work to do! Looking forward to the next galactic supernova http://antwrp.gsfc.nasa.gov/apod/ap060430.html