Super Beams

Takashi Kobayashia

a Institute for Particle and Nuclear Studies, High Energy Accelerator Research Organization (KEK),1-1 Oho, Tsukuba, 305-0801, Japan

High intensity conventional horn-focused neutrino beam produced by MW-class proton accelerator, “superbeam”, provides opportunity to further develop neutrino physics, especially long baseline (LBL) oscillation ex-periments. Several super beam LBL experiments are proposed as next generation high sensitivity, high precisionexperiments before neutrino-factory era. Sensitivities of the experiments are one orders of magnitude or morehigher than the current ones. The experiments and their physics potential are introduced.

1. Introduction

“Super beam” means high intensity neutrinobeam obtained from decays of horn-focused sec-ondary pions which are produced using a (Multi-)MW proton beam. The beam produces pure νµ

beam with νe contamination of the order of 1%from muon and Kaon decays. The sign of neu-trino, νµ/ν̄µ, can be switched by flipping the po-larity of the horn focusing system. It is stronglymotivated by future high precision long baseline(LBL) neutrino oscillation measurements. There-fore, in this presentation, I focus on LBL experi-ments with super beam while there are other in-terests on super beam such as study of neutrinointeractions, spin structure of nucleons and so on.

The LBL neutrino oscillation experimentscould be classified into 3 generations [1].

1st generation experiments (Present):Most important goal of the experimentsis the confirmation of the neutrino os-cillation found in the atmospheric neu-trino [2]. K2K [3], MINOS [4] andICARUS [5]/OPERA [6] belong to thisgeneration.

2nd generation experiments (∼10 years):The experiments are designed and opti-mized after the evidence of neutrino os-cillation and their most important goal isdiscovery of νµ → νe appearance. The ex-periments typically use a MW proton beamand ∼50 kt detector and include T2K 1stphase (T2K-I) [7], NOνA [8] and C2GT [9].

3rd generation experiments (10∼20 years?):Main purpose of the experiments will besearch for CP violation in lepton sectorand/or determination of mass hierarchythrough νe appearance measurements. Thediscovery of νe appearance is a prerequisiteto go this phase. Projects in this generationtypically assume a multi-MW proton beamand a Mton detector.

Usually “super beam” experiments in a widesense cover 2nd and 3rd generation experiments.The 2nd generations experiments T2K-I andNOνA are described in detail in other talks[10, 11]. Therefore, the 3rd generation projectsare mainly described in this presentation.

1.1. νe appearance probabilityMass hierarchy and CP violation can be probed

by measuring νe appearance. In Fig. 1, the rele-vant oscillation probabilities are plotted. The sizeof the probabilities are driven by a CP-even termproportional to sin2 θ13, while the CP-odd term isproportional to sin θ13. Therefore, the statisticalmerit on CP asymmetry, A = (P − P̄ )/(P + P̄ ) ∝

1sin θ13

, does not depend of the size of θ13, where(−)

P is(−)ν µ → (−)

ν e oscillation probability. Mattereffect, which is proportional to Eν , also producesthe difference between P and P̄ . As shown inFig. 1, at the 1st oscillation maximum in 295 kmcase, the size of the matter effect is much smallerthan the CP violation effect, but in the case of730 km, those sizes becomes comparable. This, inturn, means that, at higher energies, there would

Nuclear Physics B (Proc. Suppl.) 143 (2005) 303–308

0920-5632/$ – see front matter © 2005 Published by Elsevier B.V.

www.elsevierphysics.com

doi:10.1016/j.nuclphysbps.2005.01.122

νµ→νe

0

0.2

0.4

0.6

0.8

1

1.2

0 0.2 0.4 0.6 0.8 1 1.2 1.4Eν (GeV)

Pro

babi

lity (

%)

νµ→νe

0

0.2

0.4

0.6

0.8

1

1.2

0 0.5 1 1.5 2 2.5 3 3.5Eν (GeV)

Pro

babi

lity (

%)

(a) (b)

Figure 1. Oscillation probabilities for νµ →νe(red) and ν̄µ → ν̄e(blue) for baseline length of(a) 295 km and (b) 730 km. The solid curvesincludes asymmetry due to matter effect. Forthe dashed curves, the matter effect is subtractedand the difference between νµ → νe(red) andν̄µ → ν̄e(blue) are all due to CP effect.

be a chance to decide the sign of ∆m2 throughthe matter effect.

1.2. Off-axis beamIn many of the future super beam experiments,

“off-axis (OA)” beam [12] plays a key role toachieve high sensitivity. In the OA scheme, axisof the beam line components is declined a fewdegree from the direction to a far detector. Inthis way, high intensity low energy narrow bandbeam can be obtained. In order to maximize thesensitivity of the experiments, the peak of the en-ergy spectrum is adjusted close to the oscillationmaximum.

2. Super-beam experiments

2.1. CNGS to Gulf of Taranto (C2GT)An idea to detect off-axis νµ beam from CERN

Neutrino Beam to Gran Sasso (CNGS) [13] by amovable underwater Cherenkov detector in theMediterranean sea (Gulf of Taranto) [9]. Theexpected spectrum is peaked at ∼800 MeV, thebaseline length can be variable in the range1,100∼1,700 km and possible fiducial mass couldbe 1.5 Mt. The expected number of νµ CC in-teractions without oscillation is estimated to be∼700/year. The purpose of the experiment isthe precise measurement of νµ disappearance and

the search for the νe appearance. Sensitivity forsin2 2θ13 is about 0.0066 at 90%CL.

2.2. Second phase of T2K (T2K-II)The Tokai-to-Kamioka (T2K) experiment is a

next generation LBL experiment in Japan [7].The νµ beam is produced at Japan Proton Ac-celerator Research Complex (J-PARC) [14] anddetected at Kamioka, 295 km apart. The neu-trino beam is produced by using a 50-GeV protonsynchrotron (PS) in J-PARC.

In the first phase of T2K (T2K-I), the designbeam power of the 50-GeV PS is 0.75 MW and thefar detector is Super-Kamiokande (SK) of 22.5-ktfiducial mass [10]. The main purpose of T2K-I is discovery of νe appearance. T2K-I startedconstruction of neutrino beam line in 2004 andwill start measurement in 2009.

Furthermore, as a future extension of T2K, up-grading the PS to 4 MW and constructing 1-Mt“Hyper-Kamiokande” are also envisaged (T2K-II) [15]. With ∼5 times higher intensity andabout 25 times larger fiducial mass, statistics atSK will be 2 order of magnitude higher thanT2K-I. Expected number of νµ CC interactionsis ∼ 360, 000/year with 2◦ OA beam. The goalsof T2K-II are discovery of CP violation and pre-cise measurement of νe appearance.

Preliminary study on possible upgrade of the50-GeV PS to 4 MW has been made by the J-PARC accelerator group. It is based on (1) in-creasing repetition rate by doubling number ofRF cavities and eliminating idling time in acceler-ation cycle (factor ∼2.5), (2) doubling number ofcirculating protons by adopting “barrier bucket”method [16] (factor ∼2).

The expected sensitivity on CP violation inT2K-II based on full detector simulation is plot-ted in Fig. 2. With the present level of back-ground rejection, the sensitivity is very much de-pend on the size of systematic error. If 2% erroris achieved, then the CP violating phase δ can beexplored down to ∼ 20◦ for sin2 2θ13 greater than0.01.

2.3. CERN-Frejus projectIn Europe, there is an idea of super beam LBL

experiment in which the neutrino beam is pro-

T. Kobayashi / Nuclear Physics B (Proc. Suppl.) 143 (2005) 303–308304

0

2

4

6

8

10

12

14

0 0.2 0.4 0.6 0.8 1sinδ

sin2 2θ

13 (x

10−2

)∆m23

2=2.8x10-3

∆m122=6.9x10-5

θ23=π/4

θ12=0.594

4MW, 540kt

2+6.3years

Figure 2. Expected 3σ discovery regions of sin δas a function of sin2 2θ13 after 2 (νµ) and 6.3 (ν̄µ)years of exposure in T2K-II. The (blue) dottedcurve is the case of no background and only statis-tical error of signal, (red) dashed one is 2% errorfor the background subtraction, and (black) solidcurve is the case that systematic errors of bothbackground subtraction and signal detection are2%. The values of other parameters are shown inthe plot.

duced by Super Proton Linac (SPL) and detectedby a detector at Modane laboratory in Frejus tun-nel, 130 km from CERN [17]. The proposed SPLis a 2.2 GeV linac with 4 MW beam power op-erated at 75-Hz repetition rate and 1.5 × 1014

protons/pulse. The neutrino beam is a conven-tional wide-band beam and the expected neutrinospectrum is plotted in Fig. 3. The main compo-

106

107

108

109

1010

1011

1012

1013

ν /1

00

m/2

0 M

eV

2

νµ

νe

νµνe

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

Eν (GeV)

Figure 3. Expected flux of CERN-Frejus projectfor π+ focused beam.

nent is below 600 MeV and the peak energy isat around 300 MeV. Less NC π0 background isexpected for νe appearance search with this lowenergy spectrum. The far detector is a UNO typewater Cherenkov detector with 440 kton fiducialmass [18]. Expected number of νµ CC interac-tions is ∼15,000/440kt/year in the case of no os-cillation.

Sensitivity on νe appearance with 5 years ofrunning is θ13 = 1.5◦ (sin2 2θ13 = 0.0027) at 90%C.L. for LMA, δ = 0, ∆m2

23 = 2.5 × 10−3eV2

case as shown in Fig. 4. For sensitivity on CPV,

0.15

0.2

0.25

0.3

0.35

0.4x 10

-2

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

θ13

∆m

2 23

S MA

LMA δ=0

LMA δ=90

LMA δ=-90

Figure 4. θ13 sensitivity (90%CL) of CERN-Frejus project. Five years of running is assumed.

2 year run with νµ beam and 8 year run withν̄µ are assumed. Systematic errors of 2% onboth background normalization and signal effi-ciency are taken into account. No CPV (δ = 0)and δ = 90◦ can be discriminated at 99%CL at∆m2

12 = 7.3 × 10−5eV2 if θ13 � 1.8◦ (sin2 2θ13 �0.004).

2.4. BNL-Homestake projectThere is an interest to conduct a (very) long

baseline experiment using a neutrino beam fromBNL [19]. The neutrino beam is produced by28 GeV proton beam from AGS and is detectedby a Mton UNO type water Cherenkov detector inHomestake mine at 2540 km from BNL. The AGSbeam power is supposed to be upgraded to 1 MWfrom present 0.1 MW by introducing 1.2 GeV su-perconducting LINAC for direct injection and in-creasing repetition rate [20].

The beam is horn-focused on-axis wide bandbeam with the spectrum ranging up to about

T. Kobayashi / Nuclear Physics B (Proc. Suppl.) 143 (2005) 303–308 305

νµ DISAPPEARANCE

0

50

100

150

200

250

0 1 2 3 4 5 6 7 8 9 10

Reconstructed ν Energy (GeV)

Eve

nts/

bin

BNL-HS 2540 km

sin22θ23 = 1.0

∆m322 = 2.5e-3 eV2

1 MW 0.5 MT 5yr

No oscillations: 13290 evts

With oscillations: 6538 evts

Bckg in oscillated signal: 1211 evts

Figure 5. Expected spectrum of detected eventsin a 0.5 MT detector at 2540 km from BNL in-cluding quasi-elastic signal and CC-single pionbackground with 1.0 MW of beam power in 5years of running. The top histogram is withoutoscillations; the middle error bars are with oscil-lations and the bottom histogram is the contri-bution of the background to the oscillated signalonly. This plot is for ∆m2

32 = 0.0025 eV2.

6 GeV and the peak is at around 2 GeV(Fig. 5). Off-axis option is also a possibilitywhen higher background rejection becomes nec-essary for νe appearance search. Expected num-ber of νµ CC interactions without oscillation is∼13,000/500kt/year. An alternative detector op-tion based on very large mass liquid Argon TPCis being considered, following the approach de-scribed in [21]. Unique feature of this proposalis the large L/E, where L is the flight distanceand E is the neutrino energy. Not only the 1stoscillation maximum, but also higher oscillationmaxima are covered. The physics goals of theproject include precise measurement of νµ disap-pearance, νe appearance search, determination ofmass hierarchy, search for CPV. Moreover, thanksto the large L/E, the solar parameters (θ12 and∆m12) can be probed.

The expected sensitivities on mass hierarchyand CP violation are drawn in Fig. 6. Onlywith neutrino running, wrong hierarchy can beruled out at more than 2.5σ, and by adding anti-neutrino running, the separation becomes more

-150

-100

-50

0

50

100

150

0 0.02 0.04 0.06 0.08 0.1 0.12

Reversed hierarchy S olutions

>10σ

> 2.5σ

Sin22θ

13

δ CP

(de

g.)

STAT+SYSνµ Running O nly

68, 90 % C.L.

BNL-HS 2540 kmsin

22θ

ij (12,23,13) = 0.86/1.0/0.04, δ

CP=45

o

∆mij

2 (21,32) = 7.3e-5/2e-3 eV2

1 MW, 0.5 MT, 5e7 sec

(a)

-150

-100

-50

0

50

100

150

0 0.02 0.04 0.06 0.08 0.1 0.12

Sin22θ

13

δ CP (

deg

.)

(b)

Figure 6. BNL-Homestake sensitivity on (a) masshierarchy and (b) CP violation for normal hierar-chy case.

than 10σ. The resolution on CP violating phaseδ is 20 ∼ 30◦ at sin2 2θ13 = 0.01 which meansthat if δ is greater than 20 ∼ 30◦, CPV can bediscovered.

2.5. Projects at FNALAt Fermilab, the construction of the MINOS

experiment is almost completed and the firstbeam is planned at the beginning of 2005. Theexperiment produces νµ beam using the 120-GeV0.4-MW Main Injector (MI) and detects it atSoudan mine, 730km from Fermilab. Recently,an off-axis experiment, NOνA, using the samebeam as MINOS is proposed [11]. The plan isto construct a new 50-kt fine-grained detector ataround 14 mrad off axis and at about 810 kmfrom Fermilab. The peak position, ∼2 GeV, ofthe spectrum is tuned at the oscillation maximumfor ∆m2 � 2.5 × 10−3eV2. The most importantgoal of the experiment is to discover νe appear-ance.

As a possible future extension of theMINOS/NOνA experiments, increasing the beamintensity by upgrading the proton driver has beenactively discussed. Original idea was to replacethe present 8 GeV booster by a new one withlarger aperture and doubled repetition rate toachieve 1.9 MW MI power. Recently a newidea to construct a 8-GeV 2-MW superconduct-ing LINAC and directly inject to MI came intodiscussion. This option enables flexible operationof MI, because full 2 MW power is available atall energies from 40 to 120GeV due to negligibleinjection time compared to the ramping time.

With the upgrade, expected sensitivity of the

T. Kobayashi / Nuclear Physics B (Proc. Suppl.) 143 (2005) 303–308306

Figure 7. Sensitivity of NOνA experiment tomass hierarchy. The solid lines are with thepresent 0.4 MW beam and the dashed lines arewith upgraded 2MW proton driver [11].

experiments are improved. As an example, the ef-fect of the upgrade on the sensitivity of the NOνAexperiment to mass hierarchy is drawn in Fig. 7.

As one of more future possibility with the newproton driver, a case study, FeHo, to send a neu-trino beam produced by 120 GeV 2 MW beamand 8 GeV 2 MW beam at the same time toHomestake mine at 1290km apart from Fermi-lab [22]. The neutrino energy spectrum ranges upto about 3 GeV with the peak around 1 ∼ 2 GeV.With a Mton detector (500kt fiducial), about50,000 νµ CC interactions are expected in oneyear.

2.6. Summary of experimentsCurrent and future (super beam) LBL exper-

iments are summarized in Table 1. As can beseen in the 1.27∆m2L/E column, many of thefuture super beam experiments optimize the neu-trino energy spectrum to cover the first oscil-lation maximum with a off-axis beam. Amongthe experiments at the first maximum, T2K isabout 3 times closer distance or smaller energythan NOνA. Therefore, T2K has less sensitiveon matter effect and hence on mass hierarchywhile NOνA has higher sensitivity on the mattereffect. This difference may help to disentangle

the oscillation parameters. In that sense, theseexperiments can play complementary role. Theother group of experiments (BNL-Homestake andFeHo) aim to cover higher oscillation maxima aswell as the 1st one. The energy dependence of CPviolation effect (∝ 1/E), matter effect (∝ E) aredifferent.The wide coverage of the energy regiongive a chance to separate these two effects.

3. Summary

The next generation super beam experi-ment will provide opportunities to understandthe whole view of the neutrino mixing, such asθ13, CP violation and mass hierarchy. In the 2ndgeneration experiments, sensitivity to the νe ap-pearance will be improved by more than an or-der of magnitude from the present upper bound.The CP violating phase δ can be probed down to∼ 20◦ in the 3rd generation experiments. Severalideas of super beam experiments have been dis-cussed in US, Europe and Japan. The first 2ndgeneration experiment T2K in Japan has beenapproved and started construction in 2004. Theexperiment start measurement in 2009.

REFERENCES

1 G. Feldman, Presentation at a jointBNL/UCLA - APS Workshop “Neu-trino Super Beam, Detectors andProton Decay”, Mar. 2004 at BNL,http://www.bnl.gov/physics/superbeam.

2 Y.Fukuda et al , Phys. Rev. Lett. 81, 1562(1998), Phys. Rev. Lett. 85, 3999 (2000)T. Kajita, in these proceedings.

3 S.H. Ahn et al , Phys. Lett. B511, 178 (2001);M.H. Ahn et al , Phys. Rev. Lett. 90: 041801(2003); M.H. Ahn et al , Phys. Rev. Lett. 93:051801 (2004); T. Nakaya, in these proceed-ings.

4 D. Michael, Nucl. Phys. Proc. Suppl. 118, 189(2003);

5 F. Arneodo et al., Nucl. Instrum. Meth. A461, 324 (2001).

6 M. Komatsu, “Opera Experiment,” Nucl. In-strum. Meth. A 503, 124 (2003).

T. Kobayashi / Nuclear Physics B (Proc. Suppl.) 143 (2005) 303–308 307

Tab

le1:

Sum

mar

yof

LB

L(s

uper

beam

)ex

peri

men

ts.

The

first

two

colu

mns

are

for

prot

onbe

am,w

here

Ep

ispr

oton

ener

gy.

The

rang

eof

neut

rino

ener

gyE

νan

dth

ede

tect

orm

ass

Mdetfo

rfu

ture

proj

ects

are

roug

hro

und

num

bers

whi

char

ere

adfr

omre

fere

nces

.T

he6t

hco

lum

n,1.

27∆

m2L

/Ear

eca

lcul

ated

at∆

m2

=2.

5×

10−

3eV

2.

Osc

illat

ion

max

ima

are

expe

cted

tobe

atod

din

tege

rfo

rth

isqu

anti

ty.

The

seco

ndco

lum

nfr

omri

ght

isth

eex

pect

ednu

mbe

rof

ν µC

Cin

tera

ctio

ns.

Ep

Pow

erL

Mde

tν µ

CC

(GeV

)(k

W)

(km

)(k

t)(/

yr)

K2K

125

WB

0.4

~2.

525

00.

2~

1.3

22.5

~50

Run

ning

MIN

OS(

LE

)12

040

0W

B1

~6

730

0.2

~1.

55.

4~2

,500

Con

stru

ctin

g. S

tart

in 2

005

CN

GS

400

300

WB

5~

3073

20.

0~

0.3

~2~5

,000

Con

stru

ctin

g. S

tart

in 2

006

T2K

-I50

750

OA

0.5

~1

295

0.6

~1.

222

.5~3

,000

Con

stru

ctin

g. S

tart

in 2

009

NO

νA12

040

0O

A1.

5~

381

00.

5~

1.1

~50

~4,6

00Pr

opos

ed.

C2G

T40

030

0O

A0.

4~

11,

200

2.4

~6.

1~1

,500

~700

R&

DN

OνA

+PD

120

2,00

0O

A1.

5~

381

00.

5~

1.1

~50

~23,

000

bein

g st

udie

dB

NL

-Hom

esta

ke28

1,00

0W

B/O

A0.

5~

62,

540

0.9

~10

.3~5

00~1

3,00

0L

OI

wri

tten

T2K

-II

504,

000

OA

0.5

~1

295

0.6

~1.

2~5

00~3

60,0

00L

OI

wri

tten

SPL

-Fre

jus

2.2

4,00

0W

B0.

1~

0.5

130

0.5

~2.

6~5

00~1

8,00

0be

ing

stud

ied

FeH

o8/

120

4,00

0W

B/O

A0.

5~

31,

290

0.9

~5.

2~5

00~5

0,00

0ca

se s

tudy

Stat

us

1st gen. 2nd gen. 3rd gen.Exp

erim

ents

Eν

(GeV

)B

eam

1.27

∆m2 L

/E(π

/2)

7 Y. Itow, et al , hep-ex/0106019,http://neutrino.kek.jp/jhfnu

8 D. Ayres et al , hep-ex/0210005,http://www-off-axis.fnal.gov/.

9 A.E. Ball et al , CERN-SPSC-2004-025.10 Y. Hayato, in these proceedings.11 M. Messia, in these proceedings.12 D. Beavis et al , Proposal of BNL AGS E-889

(1995).13 G. Acquistapace et al., CERN-98-02;

R. Baldy et al., CERN-SL-99-034-DI.14 http://j-parc.jp15 M. Koshiba, Phys. Rep. 220, 229 (1992); K.

Nakamura, Neutrino Oscillations and TheirOrigin, (Universal Academy Press, Tokyo,2000), p. 359.

16 M. Fujieda et al , Proc. 1999 Part. Accel.Conf., MOP81 (1999); K. Takayama et al ,Phys. Rev. Lett. 88: 144801 (2002).

17 J.J. Gomes Cadenas et al , (hep-ph/0105297), May 2001; A. Blondel etal , CERN-NUFACT-NOTE-095, Oct.2001; M. Mezzetto, Proc. of the NuFact02Workshop, hep-ex/0302005 (2003).

18 UNO Collaboration, hep-ex/00050046; StateUniversity of New York (SUNY), Stony BrookPreprint SBHEP01-3 (Jun. 2001); C.K. Jung,in these proceedings.

19 M. Diwan, hep-ex/0407047; D. Beavis et al ,BNL-69395 (hep-ex/0211001),(2002); M. Di-wan et al , Phys Rev D68: 012002 (2003).

20 M. Diwan, W. Marciano, W. Weng et al ,BNL-71228-2003-IR (hep-ex/0305105).

21 A. Ereditato and A. Rubbia, Proceedings of“The 3rd International Workshop on Neu-trino Nucleus Interactions in the Few GeVRegion (NuInt04)”, hep-ph/0409143.

22 D. Michael, Presentation at a jointBNL/UCLA - APS Workshop “Neu-trino Super Beam, Detectors andProton Decay”, Mar. 2004 at BNL,http://www.bnl.gov/physics/superbeam.

T. Kobayashi / Nuclear Physics B (Proc. Suppl.) 143 (2005) 303–308308