Supernova Neutrinos John Beacom, Ohio State University

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 1 Why are Supernovae Special Stars?

Stars Supernovae Slow release of energy Fast release of energy Gravity > Pressure: contracon Pressure >> Gravity: explosion

Can only reveal the interior condions of collapsing stars with neutrinos

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 2 Why are Neutrinos Special Parcles?

Charged leptons Neutral leptons ± τ ντ ,ν τ ± νµ ,ν µ decay µ mixing

± e νe,ν e € € Can only probe all neutrino properes with extremes of astrophysics € John Beacom, Ohio State University € Academic Lectures, Fermilab, February 2014 3 € € Why are Cosmic Backgrounds Special Data? Photons Neutrinos

?

Lacki (2010) Can only detail energy budget of the universe with cosmic backgrounds

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 4 Talk Outline

Introducon: Basics and Movaons

Introducon: Detecon Modes

DSNB: Theorecal Predicons

DSNB: Experimental Limits

DSNB: Detecon Strategy

Concluding Perspecves

(DSNB = Diffuse Supernova Neutrino Background)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 5 Introducon: Basics and Movaons

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 6 Core-Collapse Supernova Basics

Type Ia (thermonuclear, few neutrinos)

Type II (core collapse, many neutrinos)

Neutrinos carry away the change in gravitaonal potenal energy 2 2 53 Δ(P.E.) ∼ (-GM /R)neutron-star - (-GM /R)stellar-core ∼ -3x10 erg approximately shared among all six flavors Neutrinos are trapped by scaering interacons and diffuse out quasi-thermal with ∼ 15 MeV τ ∼ few seconds John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 Importance of Supernova Neutrino Detecon

How do core-collapse supernovae explode? How do they form neutron stars and black holes? What are the nucleosynthesis products of supernovae? What are the acons and properes of neutrinos? What is the cosmic rate of black hole formaon? Which supernova-like events make neutrinos? What else is out there that makes neutrinos? ….

We cannot solve key problems without detecng supernova neutrinos

The required detecons are – surprisingly – within our reach

Detecng even a few neutrinos could oen give decisive answers

Will open new froners in observaonal neutrino astrophysics

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 8 SN 1987A: Our Rosea Stone

IMB

KamII

Observaon: Type II supernova Observaon: The neutrino progenitors are massive stars precursor is very energec

Theory: Core collapse makes a proto-neutron star and neutrinos

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 9 Supernova Neutrino Detecons Since 1987

(This page intenonally kept blank.)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 10 Introducon: Three Detecon Modes

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 11 Distance Scales and Detecon Strategies

N >> 1 : Burst N ∼ 1 : Mini-Burst N << 1 : DSNB

Gpc kpc Mpc

Rate ∼ 0.01/yr Rate ∼ 1/yr Rate ∼ 108/yr high stascs, object identy, cosmic rate, all flavors burst variety average emission

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 12 Simple Esmate: Milky Way Burst Yields Super-Kamiokande (32 kton water) ~ 104 inverse beta decay on free protons ~ 102 - 103 CC and NC with oxygen nuclei ~ 102 neutrino-electron elasc scaering (crude direconality) KamLAND, MiniBooNE, Borexino, SNO+, etc (~ 1 kton oil) ~ 102 inverse beta decay on free protons ~ 102 neutron-proton elasc scaering ~ 10 - 102 CC and NC with carbon nuclei ~ 10 neutrino-electron elasc scaering IceCube (106 kton water) Burst is significant increase over background rate Possibility of precise ming informaon

Much larger or beer detectors are being proposed now

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 13 Simple Esmate: Extragalacc Mini-Burst Yields

Yield in Super-Kamiokande ∼ 1 (Mpc/D)^2

A 5000-kton detector could see mini-bursts from galaxies within several Mpc, where the supernova rate is above one per year

New consideraons for such a detector as a dense infill for IceCube!

Kistler, Ando, Yuksel, Beacom, Suzuki (2011); builds on Yoichiro Suzuki’s ideas for Deep-TITAND

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 14 Simple Esmate: DSNB Event Rate

Super-Kamiokande rate in Kamiokande-II rate in a ∼ 1 s-1 every 10 second interval special 10 second interval

NSN Mdet dN dN 4⇥D2 DSNB * dt ⇠ dt NSN Mdet  DSNB  87A ⇥ 4⇥D2 ⇤ 87A ⇥ ⇤ For the DSNB relave to SN 1987A: 2 -10 NSN up by ∼ 100 Mdet up by ∼ 10 1/D down by ∼ 10

DSNB event rate in Super-Kamiokande is a few per year

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 15 Present: Standard Model of Predicted DSNB

See my 2010 arcle in Annual Reviews of Nuclear and Parcle Science

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 16 Theorecal Framework

Signal rate spectrum in detector in terms of measured energy dN 1 cdt e (E )=N (E ) (1 + z) ⇥[E (1 + z)] R (z) dz dE e p SN dz e Z0  h ih i Third ingredient: Detector Capabilies Second ingredient: Supernova (well understood) Rate (formerly very uncertain, but now known with good precision) First ingredient: Neutrino spectrum (this is now the unknown)

Cosmology? Solved. Oscillaons? Included. Backgrounds? See below.

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 17 First Ingredient: Supernova Neutrino Emission

Nonparametric reconstrucon from SN 1987A data Core collapse releases 10. ∼ 3x1053 erg, shared by six flavors of neutrinos -1 1 Spectra quasi-thermal Model / MeV ]

with average energies of 56 -1 ∼ 15 MeV 10 ) [ 10

! IMB Combined

( E Kam-II Neutrino mixing surely " -2 10 important but actual effects unknown -3 10 0 10 20 30 40 50 60 E [ MeV ] Goal is to measure the ! received spectrum Yuksel, Beacom (2007)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 18 Importance of the Neutrino Spectrum

Experiment Theory SN 1987A data Supernova simulaons (inial spectra)

Experiment ? +

DSNB data ? Neutrino flavor change (effects of mixing) +

Experiment Nucleosynthesis yields (neutrino interacons) SN 2012? data

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 19 Second Ingredient: Cosmic Supernova Rate

Number of massive stars unchanging due to short lifemes

dN dN dN dN =0=+ dt dt star dt bright dt dark ✓ ◆ ✓ ◆ birth ✓ ◆ collapse ✓ ◆ collapse

Measured from N/τ Measured from Inferred from mismatch; using luminosity and the core collapse can be measured by star spectrum of galaxies supernova rate disappearance; can be measured by DSNB (now high precision) (precision will improve rapidly) (froner research area)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 20 Predicons from Cosmic Star Formaon Rate

Total star formaon rate deduced from massive stars using inial mass funcon (IMF)

Impressive agreement among results from different groups, techniques, and wavelengths

Integral of RSF agrees with EBL

R (z) R (z) SF SN ' 143M Horiuchi, Beacom (2010); see also Hopkins, Beacom (2006) IMF uncertainty on RSN small

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 21 Measured Cosmic Supernova Rate

Some measured cosmic supernova rates are half as big as expected, a greater deviaon than allowed by uncertaines

Why?

There must be missing supernovae – are they faint, obscured, or truly dark?

Horiuchi et al. (2011) plus updates; see also Hopkins, Beacom (2006), Bocella et al. (2008), Mala et al. (2012)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 22 What About the Supernova Rate Nearby?

Within ∼ 10 Mpc, more supernovae than expected are found

Lots of faint supernovae

Prediction from cosmic SFR

-4 10 Cosmic SNR measurements ] -3 Mpc -1

-5 10 SNR [yr Catalog SNRs: Total Luminous (M < -15) Dim (M > -15) -6 10 10 100 Distance [Mpc] Horiuchi et al. (2011)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 23 Third Ingredient: Neutrino Detecon Capabilies

Only Super-Kamiokande has large enough mass AND (nearly) low enough backgrounds

¯ + p e+ + n e ! Free proton targets only 2 Cross secon grows as σ ∼ Eν Kinemacs good, Ee ∼ Eν Direconality isotropic

Vogel, Beacom (1999); Strumia, Vissani (2003) Super-Kamiokande

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 24 Predicted Flux and Event Rate Spectra

10 Reactor ! ] e 8 MeV 8 MeV -1 Excluded 6 MeV 6 MeV 0.5 4 MeV ] 4 MeV by SK -1 SN1987A

SN1987A MeV (2003) -1 1 0.4 MeV

-1 Invisible ! s

-2 0.3 [cm ! 0.1 0.2 [ (22.5 kton yr) e / dE " d Invisible ! 0.1 +

Reactor !e Spallation dN / dE 0.01 0 10 20 30 40 0 10 20 30 40 E [MeV] E [MeV] ! e Horiuchi , Beacom, Dwek (2009)

Bands show full uncertainty range arising from cosmic supernova rate

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 25 Neutrino Emission with Black Hole Formaon When core collapse fails (no opcal supernova), DSNB spectrum could be more detectable the neutrino emission can be larger in total and NS BH total average energy 1 E>19.3 MeV 0.33 0.56 0.89 1

s E>11.3 MeV 1.9 1.5 3.4

1 0.1 BH The collapse goes farther Total and faster, but must shed MeV 0.01 much thermal energy by 2 neutrino emission cm 0.001 ⇥ NS 4 Sumiyoshi et al. (2007) 10 0 10 20 30 40 50 Nakazato et al. (2008) E MeV Fischer et al. (2008) O’Connor, O (2011) Lunardini (2009) …. John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 26 Limits on the Black Hole Formaon Rate Low visible supernova rate would require large black hole fracon, up to ∼ 50%

Standard models predict at least ∼ 10% black holes

This can be resolved

“Survey About Nothing” (Kochanek et al., 2008) can see massive stars disappear; ASAS-SN for nearby SN rate

Lien, Fields, Beacom (2010) Large DSNB a crucial test

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 27 Present: Limits from Super-Kamiokande

See Bays et al. [Super-Kamiokande] (2012)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 28 Measured Spectrum Including Backgrounds

Amazing background rejecon: 3.5

] nothing but neutrinos despite

-1 Atm. ν + ν 3 µ µ huge ambient backgrounds

MeV -1 2.5 Amazing sensivity: factor Atm. ν + ν 2 € e e ∼100 over Kamiokande-II limit and first in realisc DSNB range 1.5 [ (22.5 kton yr) e 1 € No terrible surprises

0.5 dN / dE Challenges: Decrease 0 20 30 40 50 60 70 80 backgrounds and energy Visible Energy E [MeV] e threshold and increase Malek et al. [Super-Kamiokande] (2003); efficiency and parcle ID energy units changed in Beacom (2011) – use with care

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 29 Limits on Supernova Neutrino Emission

35 2003 Super-Kamiokande limit: A

-2 -1 erg] IMB Φ < 1.2 cm s (90% CL) 52 30

for nuebar with Eν > 19.3 MeV [10 ! 25

Supernova rate uncertainty is 20 Excluded by SK now subdominant; this limits the effecve nuebar spectrum 15 that includes mixing effects 10 B

Within range of expectaons 5 D from theory and SN 1987A! Kam-II C Time-Integrated Luminosity L 0 0 5 10 15 20 25 30 Average Energy < E > [MeV] Also limits from KamLAND ! (lower energy) and SNO (nue) Yuksel, Ando, Beacom (2006); SN 1987A fits from Jegerlehner, Neubig, Raffelt (1996)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 30 New Analysis of Super-Kamiokande Data

2003: factor ∼ 100 improvement 2012: all details down to ∼ 10% over Kamiokande-II limit More data Full reanalysis Three detector periods Backgrounds in more detail New backgrounds included Lowered energy threshold Improved efficiency Detailed systemacs Beer treatment of stascs Improved cross secon Conservave choices …. Who got the beer Ph.D. thesis project?

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 31 Examples of the New Work on Details

Background separaon by topology: Improvement in SK-I Efficiency:

1

0.9

0.8

Signal efficiency 0.7 Signal Efficiency

0.6 2003 Efficiency unnormalized LMA flux 0.5

0.4

0.3

0.2

0.1

0 20 30 40 50 60 70 80 90 Energy (MeV)

Bays et al. [Super-Kamiokande] (2012)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 32 Energy Spectrum Fits

20 30 relic all background 18 relic SK-I all background 25 ! CC 16 SK-II

e CC ! CC 14 NC elastic CC 20 e / ! 12 NC elastic !/ 15 10

8 10 6

4 5 2

0 0 20 30 40 50 60 70 80 20 30 40 50 60 70 80 38-50 degrees C. angle (MeV) 38-50 degrees (MeV)

14 relic SK-III all background 12 ! CC

e CC SK-I: 10 NC elastic !/ Best fit is slightly negave DSNB 8 SK-II and SK-III: 6 Best fit is slightly posive DSNB 4 2

0 20 30 40 50 60 70 80 Bays et al. [Super-Kamiokande] (2012) 38-50 degrees (MeV)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 33 New Super-Kamiokande Limits Much improved analysis and more data To be conservave, new limits are a factor ∼ 2 worse than before

Bays et al. [Super-Kamiokande] (2012)

Must further decrease detector backgrounds and energy threshold

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 34 Emerging: Gadolinium in Super-Kamiokande?

See talks by Mark Vagins at HAνSE 2011 and LowNu 2011

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 35 GADZOOKS! Proposal

The signal reacon produces a neutron, but most backgrounds do not

Beacom and Vagins (2004): First proposal to use dissolved gadolinium in large light water detectors showing it could be praccal and effecve

Neutron capture on protons Gamma-ray energy 2.2 MeV SK Hard to detect in SK + ¯e + p e + n ! SK+Gd Neutron capture on gadolinium Gamma-ray energy ∼ 8 MeV Easily detectable coincidence New general tool for parcle ID separated by ∼ 4 cm and ∼ 20 µs Rich new physics program

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 36 Benefits of Neutron Tagging for DSNB

Solar neutrinos: 3 eliminated 10 GADZOOKS!

-1 2 Reactor ! Spallaon daughter decays: 10 e essenally eliminated 1 Supernova !e 10 (DSNB) Reactor neutrinos: Atmospheric !! !e

[(22.5 kton) yr MeV] 0 now a visible signal e 10

dN/dE -1 Atmospheric neutrinos: 10 significantly reduced -2 10 0 5 10 15 20 25 30 35 40 Measured E [MeV] DSNB: e More signal, less background! Beacom, Vagins (2004) (DSNB predicons now at upper edge of band)

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 37 Research and Development Efforts

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 38 EGADS Proposal

Masayuki Nakahata, Mark Vagins, others

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 39 EGADS Detector

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 40 Mad Scienst at Work in Underground Lair

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 41 Water and Gadolinium Filtraon System

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 42 Recent News from Vagins Filtraon System – Pure Water: Transparency of filtered pure water in EGADS equivalent to SK

Gadolinium Water Small-Batch Brew System: Gadolinium dissolved with no problems in 15-ton holding tank

Gadolinium Removal System: Demonstrated factor 106 removal of gadolinium in a single pass

Filtraon System – Gadolinium Water: Gadolinium water circulaon already has 99.97% efficient return

Gadolinium Water Transparency: Transparency for filtered gadolinium water is already very high Now tesng EGADS with gadolinium-loaded water

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 43 Possible Future for Water + Gd Detectors

;.

EGADS: Super-K: Hyper-K: 0.2 kton total 22.5 kton fiducial 560 kton fiducial Almost completed Working well LOI submied

Gd addion soon Gd decision soon Gd discussed

Dramacally increased reach for many physics topics with Gd

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 44 Concluding Perspecves

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 45 Prospects for First Detecon of the DSNB

Guaranteed signal: SK has a few DSNB nuebar signal interacons per year Astrophysical uncertaines are small and shrinking quickly

Super-Kamiokande upgrade: Possibility of adding gadolinium is seriously considered Research and development work very promising so far

Supernova implicaons: New measurement of cosmic core-collapse rate (and more?) Direct test of the average neutrino emission per supernova

Broader context: Possible first detecons besides Sun and SN 1987A Non-observaon of a signal would require a big surprise

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 46 Summary of Three Detecon Modes

Milky Way Nearby Galaxies DSNB Mostly ready Need new detectors Need an upgrade

Lifeme events Annual events Steady detecons

Only way to measure Only way to check Only way to measure precise details, neutrino presence, average emission, me structure, burst variaon, cosmic emissivity, all flavors, isolate dark bursts, new sources, etc. etc. etc.

We cannot solve key problems without detecng supernova neutrinos

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 47 Broader Vision for Core-Collapse Science Core-collapse paradigm Opcal supernova types

Core-collapse simulaons Supernova explosions

Failed collapses Disappearing stars

Extreme stellar outbursts Supernova “impostors”

Exact collapse me Gravitaonal wave searches

Nuclear maer theory Neutron star properes

Condions above core Nucleosynthesis yields

Massive star fates Missing supernova problem

Variety in collapse Variety in neutron stars SN neutrino detecon modes Mulple New parcle physics Extreme condions

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 48 Center for Cosmology and AstroParcle Physics

New faculty hires Linda Carpenter, Chris Hirata, and Annika Peter

Postdoctoral Fellowship applicaons welcomed in Fall

ccapp.osu.edu

Some (rough) stascs that may surprise Columbus, Ohio: 0.8 million people (city), 1.8 million people (metro) Ohio State University: 56,000 students on Columbus campus Physics: 55 faculty, Astronomy: 20 faculty CCAPP: 20 faculty, 10 postdocs from both departments

John Beacom, Ohio State University Academic Lectures, Fermilab, February 2014 49