Why accelerator-based oscillation physics may be the best place to find

The NextThe Next RealRealSurprise!Surprise!

Janet Conrad,Columbia University

Because...Neutrinos are a relatively unexplored frontierandAccelerators are developing so rapidly

There is a special opportunity hereto find something really unexpected

Recent technology revolutions for accelerator-based neutrinos:

Nearly MegaWatt beams (with the promise of Multi-MegaWatts soon:

SC RF Linacs, etc.) Improved secondary focusing Selectable energy ranges, narrow band beams GPS High statistics 2ndary production and xsec experiments

(e.g. HARP, BNL-E910, MIPP, FINeSSE...) Innovative ideas for even better accelerators:

FFAG, neutrino factory, beta beams

Opening Up OpportunitiesUsing Accelerator-basedOscillations Searches....

Soon: MiniBooNENear Term: Minos, Opera, T2K, Nova

Long Term: Neutrino Factory, Beta Beam

The opportunity for surpriseextends beyond oscillation searches too...

See talk by Mike Shaevitz

It would be a big surprise if ....

LSNDwas

RIGHTI'll eat my hat

Nearly 49,000 Coulombs of protons on upstream target

Baseline 30 m

Neutrino Energy 20-55 MeV,

167 tons Liquid scintillator

1220 phototubes

combined analysis allowed region

The LSND Experiment at LANL(1993-1998)

:87.9 ± 22.4 ± 6.0 (4.σ)

νe events observed in aνµ beam ???

P(νµ νe)= sin22θ sin2(1.27∆m2L/Ε)

∆m122

∆m232

ν1

ν2

ν3

Problem:

∆m132 = ∆m12

2 +∆m23

2

It must be new physics. It could be new muon-decay physics (TWIST) It could be new oscillation physics (MiniBooNE)

If LSND is not due to a fluctuation or an unknown systematic

(or one of the other results must change due to systematics or new physics)

MiniBooNE:

Booster K+

target and horn detectordirt (~450 m)decay region absorber

primary beam tertiary beamsecondary beam(protons) (mesons) (neutrinos)

Keep L/E same while changing systematics

P(νµ νe)= sin22θ sin2(1.27∆m2L/Ε)

E~1GeV

The MiniBooNE Beam:

µ → e νµ νe

K→ π e νe

Produced via 8 GeV protons on tgt

Secondary production measuredat the HARP experiment:

Focusing via a horn

50 m decay path for secondaries

in-situ kaon decay monitor

The MiniBooNE Detector

• 12 meter diameter sphere

• Filled with 950,000 liters of undoped mineral oil

(events produce both Cerenkov and Scintillation γ's)• Light tight inner tank with 1280 photomultiplier tubes• Outer veto region with 240 PMTs. •We measure charge and time for each tube

Many in situ monitoring systems...

resonant:

coherent:

Preliminary

+0

Coming soon...Resonant π+

production studies

Crucial if we plan to do precision ocillation physics!

Right now: Calibration & Cross Section Studies:

Updated Appearance Sensitivity

● νe signal events

● NC π0 misIDs

● Beam νe events

● νe signal and background breakdown

● Reasonable signal separation with 1021 POT

Monte Carlo

Monte Carlo

There aresterile

neutrinosNot a chance!

It would be a big surprise if ....

Z decay indicates 3 active neutrinos!

Another neutrino with m<45 GeVwill have to be sterile

νeνµ

ντ

νs

Ue1 Ue2 Ue3 Ue4 Uµ1 Uµ2 Uµ3 Uµ4

Uτ1 Uτ2 Uτ3 Uτ4Us1 Us2 Us3 Us4

ν1ν2

ν3

ν4

=

But it could mix with the others...

...thus be observed via oscillations

We know...

Why search?

Interesting from a theoretical point of view:The simplest extension to a model with neutrino massSterile neutrinos ''fall out'' of GUTS and Extra Dimensions Theories

Why should the sterile neutrino be light? maybe a new symmetry Why not massless? maybe Planck scale corrections

They can improve models for the R-process in supernovae (which requires oscillations)

Ye

Ye = 1/(1+(n/p)(Ye small has neutron excess)

present theoriesFrom Strumia, Neutrino '04:

Luckily, Experiments are more restrictive/suggestive!

Treat LSND as a positive signal

CDHS sees an effect in LSND mass range (near detecto2σ)

(Sorel, Conrad , Shaevitz., hep-ph/0305255)

Limits ~90% CL from atmospheric < 30%solar < 10%

Short baseline experiments:

This pulls the fit at about 20 eV2

Compatibility Level = 4%3+1 Models do not show high compatibility...

3+2 Models do showcompatibility....

NSBL LSND

90% CL

95% CL

99% CL

Best Fit: ∆m412=0.92 eV2 , Ue4=0.121 , Uµ4=0.204 , ∆m51

2=22 eV2 , Ue5=0.036 , Uµ4=0.224

Compatibility Level = 30%Allowed region for 3+2 "joint analysis"@90% CL

Comparing''Null Short Baseline''to LSND

1σ, 2σ

What does this mean for cosmology?

Hannestad astro-ph/0303076

Need to vary Ho in global fitsas you vary the mass,,,,

K. Abazajian hep-ph/0307266 J. Beacom, astro-ph/0404585 R. Mohapatra, hep-ph/047194

The latest from BBN fits:Cyburt, Fields, Olive, Skillman, astro-ph/0408033

at 68% CL (1σ) 2.67<Nν<3.85

...It would mean we'd have a lotof surprised cosmologists!

Including CMB/LSS too:

There are mechanisms whichsubstantially reduce cosmic ν's

NeutrinosDecay

That can'tbe true!

It would be a big surprise if ....

ν2 →ν3 + J Implies a non-zero mass difference between 2 neutrino states:

Which means mixing is possible as well:

So the most general formula for survival probability would be:

Where the lifetime appears as:

hep-ph/9810121hep-ph/9907421

And is assumed to be a function of the 2 neutrino masses

Isn't neutrino decay already ruled out by Super K?

Best fit oscillation: Neutrino Decay:

Errors are statistical

only

Errors are statistical

only

χ2/dof=37.9/40 (Prob=56%)

Fit, including systematic errors...

χ2/dof=49.1/40(Prob=15%)

hep-ph/0404034

There is 3σ betweenχ2 min...but...

Long Base Line experiments: An idea pioneered by K2K... who showed a 3.9σ osc. signal at Nu'04!

Expected: 150 ± 11.6

The ideal place to look:

But not enough statistics todifferentiate decay from oscillations...

8m octagonal steel & scintillator tracking calorimeter

Magnetized Iron (B~1.5T)

484 planes of scintillator

Beam begins Jan '05

running with cosmic ν's now....

The Next to Turn On: NuMI/Minos

Near DetectorDet. 2

Det. 1

Far Detector:segmented Iron

calorimeter detector

upward goingneutrino ratescompared to Bartol'96 flux

Neutrino decay in NuMI/Minos

Solid histogram: Flux at Soudan with no osc or other new physics,Dashed: Flux with oscillations (SK best fit)Points: Flux for neutrino decay

There is a big differencebetween the models atlow energy!

θ23

is exactly

45o

No Way!

It would be a big surprise if ....

None of the ''well−measured'' anglesin the quark sector or the neutrino sector,

are ''maximal''

solar only

solar+KamLAND...Except, for, possibly,the atmospheric oscillation

WHY MAXIMAL MIXING?see Grimus, hep-ph/0405261

A new symmetry group, eg: Z2, D4 models

max

max

Expected Minos capability:solid = 90% CL

d(sin2 2θ23)~5% with 25E20 protons on target (~6 years)

Is it maximal mixing?

SK: 90% CL and best fit (✭)

OPERA data can be combined also νµ → ντ appearance

NuMI runningwill be 4E20 p.o.t/year

Begins in 2006

And T2K can improve on this further!

Approved Dec, 2003Expected to run: 2009

Plan is to use a 2.5o off-axis beam& the existing Super K detector

Soudan

Macro

SK

CP violation is HUGE

in the lepton sectorYou're

kiddingme!

It would be a big surprise if ....

From Atmosphericand Long Baseline

DisappearanceMeasurements From Reactor

DisappearanceMeasurements

From Solar NeutrinoMeasurements

From Long BaselineAppearance

Measurements

The Mixing Matrix:The CP Violation Parameter

s13

has to be large (>1%)to have a chance at all!

CP violation

Posc(να→ νβ) ≠ Posc(να→ νβ)

∆m2<0∆m2>0

Posc(να→ νβ)

P osc

( να→

νβ)

CP

CP + matter,∆m2 <0

CP + matter, ∆m2 >0

δCP parameter

To untangle,you need different baselines!e.g.

T2K and Nova(Proposed off-axis FNAL-Minn.)

0

π

Can we tell at 3σ that δ ≠ 0 or π???

largemedium& small reactor expt.sensitivity

With a Multi-MegaWProton Driver

Nature'sreal value

For nature'sreal values of δ and sin2 2θ13,the experimentcan see that P

ν ≠ P⎯ ν at 3σ

2450km baseline

1MW source

½MT detector

5×107sec exposure

If the CP Violation signal is not 90 or 270 degrees,How can we see it?

VLBνO

We sawsomething

evenstranger!

Try to bepractical!

It would be a big surprise if ....

To go furtherwe will need

The Next Generationof accelerators....

A Neutrino Factory ..... or ..... A Beta Beam Facility

High Intensity Well defined flavor content Well defined energy distribution

Not in mylifetime!

You're kidding

me! No Way!

Not a chance!

Gimme a break!

I'll eat my hat

.

.

MiniBooNE

Minos

CNGS

T2K

Nova

VLBνO

Beyond!

LSN

D

ster

iles

deca

y

max

. mix

CPV

Bey

ond!

Accelerator-based oscillation studies may be the best,most versatile way to search for the next BIG surprise!

Backup Slides

Theoretical Justification for 3+2 .... examples from the last 6 months

THE STERILE NEUTRINO: FIRST HINT OF 4TH GENERATION FERMIONS?By Stephen Godfrey, Shouhua Zhu. May 2004. 4pp. ** Temporary entry ** e-Print Archive: hep-ph/0405006

LARGE MIXING FROM SMALL: PSEUDODIRAC NEUTRINOS AND THE SINGULAR SEESAW.By G.J. Stephenson,Jr. (New Mexico U.), T. Goldman (Los Alamos), B.H.J. McKellar, M. Garbutt (Melbourne U.),. LA-UR-04-1736, Apr 2004. 26pp. Extension of hep-ph 0307245. e-Print Archive: hep-ph/0404015

TWO LIGHT STERILE NEUTRINOS THAT MIX MAXIMALLY WITH EACH OTHER AND MODERATELY WITH THREE ACTIVE NEUTRINOS.By Wojciech Krolikowski (Warsaw U.),. IFT-04-7, Feb 2004. 12pp.Published in Acta Phys.Polon.B35:1675-1686,2004e-Print Archive: hep-ph/0402183

(3+2) NEUTRINO SCHEME FROM A SINGULAR DOUBLE SEESAW MECHANISM.By K.L. McDonald, B.H.J. McKellar, A. Mastrano (Melbourne U.),. Jan 2004. 5pp.e-Print Archive: hep-ph/0401241

SIMPLE MODEL FOR (3+2) NEUTRINO OSCILLATIONS.By K.S. Babu, Gerhart Seidl (Oklahoma State U.),. OSU-HEP-03-15, Dec 2003. 12pp.e-Print Archive: hep-ph/0312285

CPT Violation

Mass Spectrum Model:LSND

atmos, Kamland atmospheric

solar

Antiν: ν:

hep-ph/0210411 Barenboim, Lykken

Quantum Gravity Decoherence Model:

disfavored unless steriles are also invoked

Additional mixing induced by singular space-time configurations(wormholes, microscopic black holes, geons = ''space time foam'')

fit to data: χ2/DOF = 60.7/56 hep-ph/0404014, 0406035 Barenboim, Mavromatos

Lorentz Invariance Violation:Kostelecky and Mewes, hep-ph/0308300

Fits to neutrino data can, in principle, accommodate an LSND signal

hep-ph/0308299, Barger, Marfatia, Whisnant

TWIST:

& MiniBooNE antineutrino running:

Sorel and Whisnant, preliminary

Posc(να→ νβ) ≠ Posc(να→ νβ)

CP Violation in 3+2

How the Nova sensitivity was calculated:

Code: a package written by Mike Shaevitz for APS Nu Study.(includes osc. prob. code from S.Parke)

Purpose: Study relative contributions of Reactor, T2K, Novato atmospheric ∆m2 studies individually, in groups, as fn of time.

Agreement between Groups:A meeting between representatives of the SuperBeams and Reactor APS Study Groups (SW, JC, DM, BM/MD, GB, EB, MS, GF)led to agreement on this code, statistical methods & presentatation layout.

How the code works, in general:

1) Generate data (osc. probs) for a given point in δ and sin2θ13 space.2) Find the minimum χ2 demanding δ=0 but allowing θ13 and θ23 to vary3) The 2σ limit curve is where the χ2=4

The main focus of MiniBooNE right now:Precision modeling/understanding of beam and detector...

Optical Model:● Light Creation

– Cerenkov – Scintillation

● Light Propagation– Fluorescence

– Scattering– Absorption

Secondary Production Model:new data from BNL 910

Calibration:● Michel electrons● Tracker & Cube system