Progress in Particle and Nuclear Physics 64 (2010) 334–341

Contents lists available at ScienceDirect

Progress in Particle and Nuclear Physics

journal homepage: www.elsevier.com/locate/ppnp

Review

Determination of the mixing angle θ13R.D. McKeownW.K. Kellogg Radiation Laboratory, California Institute of Technology, Pasadena, CA 91125, USA

a r t i c l e i n f o

Keywords:Neutrino massNeutrino mixingNeutrino oscillations

a b s t r a c t

The phenomenon of neutrino flavor oscillations is now experimentally well established(R.D. McKeown and P. Vogel, 2004 [15]). The neutrino mixing matrix depends upon 3mixing angles and a CP-violating phase. Two of the mixing angles have been measuredbut the last mixing angle, known as θ13, has not yet been experimentally determined to benon-zero. A variety of experimental efforts are underway to attempt establishing a well-determined finite value for θ13, and future experiments to further our knowledge of thisangle are being proposed. I present an overview of the status of this experimental programto measure θ13.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

The phenomenon of neutrino oscillations occurs when the flavor eigenstates (i.e., states produced in weak interactionprocesses, νe, νµ and ντ ) are not identical with the mass eigenstates. The neutrino flavor eigenstate produced in a weakprocess (such as nuclear beta decay) is then a superposition of themass eigenstates, and the subsequent evolution of the stateas it propagates through space involves slippage of the relative phases of the mass eigenstates leading to flavor oscillations.This is easily demonstrated in a 2 flavor approximation where there are 2 flavor eigenstates (e.g. νe and νµ) that are

superpositions of 2 mass eigenstates (ν1 and ν2, with massesm1 andm2). The mixing is described by a matrix involving onemixing angle θ(

νeνµ

)=

(cos θ sin θ− sin θ cos θ

)(ν1ν2

). (1)

If a νe is produced in an experiment, it will develop a component of νµ as it propagates through space. The νµ probabilityoscillates in the propagation distance L. It is straightforward to show that the probability for the νµ state is given by

P(νe → νµ) = sin2(2θ) sin2[1.271m2(eV2)

L(m)Eν(MeV)

], (2)

where1m2 ≡ |m22 −m21| is the difference in squared masses and Eν is the neutrino energy.

The 3 flavor case is a straightforward generalization in which there are 3 mass eigenstates (with masses m1, m2, andm3) and a 3 × 3 mixing matrix that depends on 3 mixing angles (θ12, θ23, and θ13) plus a CP-violating phase δ. The mixingmatrix is denoted as UPMNS (for Pontecorvo [1], Maki, Nakagawa and Sakata [2]) and, for neutrino oscillation physics, can beconveniently written as a product of three matrices:

UPMNS =

(1 0 00 c23 s230 −s23 c23

)×

c13 0 s13e−iδ

0 1 0−s13eiδCP 0 c13

× ( c12 s12 0−s12 c12 00 0 1

)(3)

E-mail address: bmck@caltech.edu.

0146-6410/$ – see front matter© 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.ppnp.2009.12.043

R.D. McKeown / Progress in Particle and Nuclear Physics 64 (2010) 334–341 335

Eν (MeV)

(see

ann

otat

ions

)

(a)

(b)

(c)

(a)ν_

e interactions in detector [1/(day MeV)]

(b) ν_

e flux at detector [108/(s MeV cm2)]

(c) σ(Eν) [10-43 cm2]

0

10

20

30

40

50

60

70

80

90

100

2 3 4 5 6 7 8 9 10

Fig. 1. Reactor νe spectrum, inverse beta decay cross-section, and νe interaction spectrum, from [6].

where sij ≡ sin θij and cij ≡ cos θij. As discussed below, there are already significant experimental determinations of θ12 andθ23, so the matrix element in Eq. (3) involving the CP-violating phase s13e−iδ vanishes if θ13 = 0. Thus the angle θ13 can beviewed as the gateway to observation of CP violation in the lepton sector.The most significant determination of θ23 comes from measurements of the oscillations associated with atmospheric

neutrinos by the Super-Kamiokande collaboration [3]: sin2 2θ23 > 0.92 at 90% CL. The most precise value of 1m223 is fromthe MINOS accelerator-based oscillation measurement [4]:1m223 = 2.43± 0.13× 10

−3 eV2.Experimental information regarding the mixing angle θ12 and1m212 comes from combined fits of the KamLAND reactor

neutrino oscillation measurement and solar neutrino data (assuming CPT invariance) [5]:

1m212 = 7.59± 0.21× 10−5 eV2 (4)

tan2 θ12 = 0.47+0.06−0.05. (5)

2. Reactor experiments and θ13

Nuclear reactors are prolific sources of antineutrinos emitted by neutron-rich fission fragments. The flux and energyspectrum of the antineutrinos has been studied in great detail [6]. The energy spectrum of the νe is a steeply falling spectrum(see Fig. 1) and depends slightly on themix of fissionablematerialwhich varies during the fuel cycle of the reactor. Generally,if one has knowledge of the reactor power and fuel composition then one can predict the flux to 1%–2% accuracy.The reactor neutrino experiments utilize the inverse beta decay reaction on the proton

νe + p→ e+ + n (6)

which enables determination of the antineutrino energy through energy conservation Eν = E(e+) + 1.31 MeV + En. (Theaverage neutron recoil energy En ∼ 10 keV is a very small contribution.) The threshold for this reaction is Eν ' 1.8 MeV,and the cross-section increases roughly quadratically with Eν above this value, as shown in Fig. 1. The cross-section for thisprocess is strongly constrained by the neutron lifetime, and so is accurately known to better than 0.5% precision [7].The formula for survival of electron neutrinos (or antineutrinos) in the 3 flavor case is given by [8]

P(νe → νe) ∼= 1− sin2 2θ13 sin2131 − cos4 θ13 sin2 2θ12 sin2112 (7)

where 1ij ≡ 1m2ijL/4Eν . Note that the 2 terms oscillate with different ‘‘frequencies’’ depending on the values of the 1m2ij.

Thus one can choose the baseline L to maximize (or minimize) the sensitivity to particular 1m2ij. For an average reactorantineutrino energy of 4 MeV and a value of1m232 = (2.43± 0.13)× 10

−3 eV2 one finds that the optimum distance for thefirst minimum is L ' 2000 m.

2.1. CHOOZ and Palo Verde experiments

The best present experimental information on θ13 is due to the CHOOZ experiment [9]. The CHOOZ nuclear power plantin France contains 2 reactor cores for a total power of 8.5 GWth. The detector/target was 5 ton of Gd-doped (0.09% by

336 R.D. McKeown / Progress in Particle and Nuclear Physics 64 (2010) 334–341

10-1

Δm

2

2Θsin2

10-2

10-3

10-4

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

2Θsin2 (90% CL)SK

Palo Verde

Chooz

13

Fig. 2. Excluded regions for the Super-Kamiokande (outside dotted curve) and CHOOZ and Palo Verde experiments (right of solid curves, from [6]).

Table 1Parameters of the 3 new reactor experiments to study θ13 .

Experiment Thermal power Baselines near/far Overburden near/far Target mass Start datea Sensitivityb(GWth) (m) (mwe)c (ton) (Proj.)

Double Chooz 8.6 410/1050 115/300 8.8/8.8 2010–11 0.032RENO 17.3 290/1380 120/450 16/16 2010 0.02Daya Bayd 17.4 363/1985 255/910 40/80 2011 0.008

481/1613 292/910 40/80a The start dates are approximate based on current construction schedules.b At1m213 = 2.5× 10

−3 eV2 , 90% CL and 3 years running.c Meters water equivalent.d The Daya Bay site involves 2 near sites near 2 groups of reactors, so the baselines and overburdens of the 2 near sites are given separately.

weight to facilitate detection of the delayed neutron) liquid scintillator located 1050 m from the reactors under ∼300 mwater equivalent of overburden. The CHOOZ measurement of the antineutrino flux was in excellent agreement with theno-oscillation expectation, with the resulting ratio of observed to expected rate R = 1.01± 2.8%(stat)± 2.7%(syst).The Palo Verde experiment [10] provided important contemporaneous confirmation of the CHOOZ result, although with

less precision. The Palo Verde power station in Arizona consists of 3 reactor cores with a total power output of 11.6 GWth.The detector/target consisted of 11.24 ton of Gd-doped liquid scintillator located at an average of about 850 m from thereactors. The overburden was only 32 m water equivalent, so cosmic background was more of an issue for the Palo Verdesetup.The results of the CHOOZ and Palo Verde experiments are displayed in Fig. 2. The CHOOZ result is more restrictive,

constraining θ13 to sin2 2θ13 < 0.15, 90% CL, at1m213 = 2.43× 10−3 eV2 (the central value reported by MINOS [4]).

2.2. New reactor experiments to determine θ13

Three new experiments are now in preparation to study the quantity θ13 with higher precision. These all intend to utilizethe method of comparing two identical detectors at different baselines to reduce systematic errors, as first proposed byMikaelyan and Sinev [11]. The new experiments are Double CHOOZ [12] located at the CHOOZ site, a new experiment inChina, the Daya Bay experiment [13], and another called RENO in Korea [14]. These three experiments are the topics of threeother contributions to these proceedings where one can find more information on the experimental details, in addition tothe references given here. A summary of the experimental parameters, schedules, and projected sensitivities is presentedin Table 1.

3. Accelerator-based experiments

A variety of long baseline accelerator-based experiments are currently underway or in preparation. These experimentsall have capability to address the mixing angle θ13 through observation of the process νµ → νe.

R.D. McKeown / Progress in Particle and Nuclear Physics 64 (2010) 334–341 337

80

Medium Energy Time

on-axis7 mrad off-axis14 mrad off-axis21 mrad off-axis

60

40

CC

eve

nts/

kt /

IE2I

PO

T /

0.2

GeV

μ

20

00 2 4 6 8 10

Eν

ν

(MeV)

Fig. 3. Charged-current νµ event rates prior to oscillations calculated for a distance of 810 km from Fermilab and at various off-axis locations in the NuMIbeam [16].

The formula for νe appearance is given by a more complicated expression (here including matter effects to lowest order)[15]:

P(νµ → νe) ' 4c213s213s223 sin

2131 + 8c213s13s23c23s12c12 sin131[cos132 cos δ − sin132 sin δ] sin121− 8c213s

213s223s212 cos132 sin131 sin121 + 4c

213s212[c

212c223 + s

212s223s213 − 2c12c23s12s23s13 cos δ] sin

2121

− 8c213s213s223(1− 2s

213)aL4Eνsin131

[cos132 −

sin131131

](8)

where the notation

sij = sin θij, cij = cos θij, 1ij = 1m2ijL/4Eν (9)

is used and the matter effect involves the parameter

a ≡ 2√2GFNeEν = 1.54× 10−4Yeρ (g/cm3)Eν (GeV) (a is in (eV2)). (10)

The probability of the conjugate process P(νµ → νe) is obtained by the substitution δ→−δ and a→−a.The νe appearance probability clearly depends on the phase δ and the mass hierarchy [sign (1m231)] in addition to the

mixing angle θ13. Clearly observation of this process is of great interest, but to extract the physicswill require detailed studiesto separate the effects of CP violation and mass hierarchy.The experimental method generally involves an intense proton beam incident on a π production target, followed by a

focusing horn system and a decay regionwhere the neutrinos are produced viaπ → µ+νµ. The polarity of the horn systemcan be reversed to switch from π+ (for a νµ beam) to π− (for a νµ beam, generally with less intensity). The beam must bedirected downward, through the earth, towards the neutrino detector/target at a remote location. Some experiments utilizethe ‘‘on-axis’’ method where the detector is on the symmetry axis of the νµ beam and the energy spectrum is quite broad(‘‘wide band’’). Others use an ‘‘off-axis’’ geometry,where the beam ismoremonoenergetic (‘‘narrowband’’) butwith reducedflux. As an example, the spectra of event rates for the NOνA experiment (discussed below) are displayed in Fig. 3 for theon-axis case and several off-axis geometries. In addition to the ‘‘far’’ detector at the remote location, these experimentsalso utilize a ‘‘near’’ detector which can monitor the flux and spectrum of neutrinos before oscillations cause significantmodifications. The ‘‘near’’ detector is also useful in constraining the contamination of νe in the νµ beam.The three experiments that are presently in progress (running or under construction) are MINOS, T2K and NOνA. These

are briefly described in the following, and Table 2 summarizes the properties of these projects.

3.1. MINOS

The MINOS (Main Injector Neutrino Oscillation Search) experiment has been in operation for several years. A wide bandbeam generated at Fermilab is directed towards the Soudan mine (baseline 735 km) in Minnesota where the 5.4 kton fardetector is located. As discussed above,MINOS has observed a significant νµ disappearance effect, providing themost precise

338 R.D. McKeown / Progress in Particle and Nuclear Physics 64 (2010) 334–341

Table 2Parameters of the 3 accelerator-based experiments to study θ13 .

Experiment Beam power (MW) Baseline (km) Target mass (kton) Start date (Proj.) Sensitivity 90% CL

MINOS 0.25 735 5.4 2005 0.1T2K 0.1–0.75 295 22 2010 0.014NOνA 0.7 810 14 2013 0.010

3

2

num

ber

of s

igm

a

1

0

ϑsin213

ν oscillation parameter bounds on 3 ϑ13

0 0.02 0.04 0.06 0.08 0.1

Fig. 4. Global fit result for θ13 from [17].

value to date for1m223. MINOS has also been studying νe appearance, and one can expect the collaboration to report resultsof comparable sensitivity to CHOOZ when the complete dataset is analyzed in the near future.

3.2. T2K

The T2K (Tokai to Kamioka) experiment is currently under construction, and is expected to begin operation in 2010. It willutilize a neutrino beam generated by the new JPARC facility and the existing Super-Kamiokande water Cherenkov detectorat a distance of 295 km. The off-axis (2.5◦) geometry results in a narrow band neutrino beam peaked at about 0.5 GeV. Anear detector facility will be located at 280m from the pion production target. Running will begin with low-power, 0.1 MW,with the ultimate goal of reaching 0.75MW.With 5 years running at 0.75MW the sensitivity to sin2 2θ13 will approach 0.01.

3.3. NOνA

The NOνA experiment has just begun construction and is expected to begin operation in 2013. The existingmain injectorneutrino beam (NuMI) at Fermilab will be employed, but in an off-axis (14 mr) geometry to the new Ash River far detectorsite in Minnesota (811 km). The beam power will be upgraded to 0.7 MW. The detector is a 14 kton array of 4 cm× 6 cm×15 m liquid scintillator modules, enabling excellent electron identification. The run plan is for 3 years of νµ and 3 years ofνµ with a sensitivity to sin2 2θ13 ∼ 0.01 at 90% CL. If sin2 2θ13 is significantly larger than 0.01, NOνA could have sensitivityto the mass hierarchy as well as the CP-violating phase δ.

4. Current status ofΘ13

At present, the most significant data bearing on the value of θ13 is still the CHOOZ result sin2 2θ13 < 0.15 at 90% CL.However, it is possible to perform global 3 generation fits including the KamLAND and solar data, atmospheric ν data, anddata from long baseline accelerator experiments like MINOS. The result of such a recent analysis [17] is shown in Fig. 4. Theresult (1σ ) obtained in this analysis is

sin2 θ13 = 0.016± 0.010. (11)

R.D. McKeown / Progress in Particle and Nuclear Physics 64 (2010) 334–341 339

Fig. 5. Projected electron neutrino appearance signal at 1300 km from an on-axis wide band neutrino beam [18].

While this is considered an intriguing hint that θ13 6= 0, the statistical significance is not great enough to allow a firmconclusion.

5. Future proposed project at DUSEL

During the last few years a new future project to study long baseline neutrino oscillations has been developed in theUnited States. The basis for this concept was a joint Fermilab/Brookhaven study [18] that considered various options withdetailed simulations. In 2008, the Particle Physics Prioritization Panel (P5) endorsed the plan to site a large detector atthe proposed DUSEL (Deep Underground Science and Engineering Laboratory) in Lead, South Dakota and construct a newneutrino beam from Fermilab directed at this location [19].The primary goal of this experiment is to study the oscillation channels νµ → νe and νµ → νe (see Eq. (8)) and determine

the three quantities θ13, δ (CP violation) and the sign of1m213 (mass hierarchy). For the longer baseline of 1300 km, there areoscillation maxima at132 ' π

2 ,3π2 , corresponding to Eν ' 2.5, 0.8 GeV. In addition, the matter effects are quite significant

at this larger distance. The experiment strategy is to produce a wide band beam that will cover the energy region of thesetwo oscillation maxima. Fig. 5 shows the νe appearance signal from a wide band νµ beam at 1300 km indicating that thereare two oscillation maxima and that the energy spectrum provides significant information on δ and the mass hierarchy. Atthis larger baseline (compared to NOνA or T2K) andwith the additional capability to study ν, the appearance signals containenough information to determine all three oscillation parameters.A new beamline will need to be constructed at Fermilab to generate the neutrino beam directed towards DUSEL. The

60 GeV proton beam from the main injector will be transported to a new target location. In the current design, focusinghorns will direct the pions into a ∼2 m diameter decay pipe that will be about 300 m long. The beam power will initiallybe 700 kW, but with a possible new proton driver (known as Project X) the beam power could be increased to 2 MW. Theresulting wide band neutrino beam would span the energy range 0.5–4 GeV, providing simultaneous coverage of the 2maxima of the oscillation probability (Fig. 5). A near detector facility will be located on or near the Fermilab site to monitorthe flux and energy spectrum of the neutrino beam as well as contaminant species such as νe and ‘‘wrong sign’’ neutrinos.The detector will be located deep underground at DUSEL, and a very large detector mass must be employed to yield

sufficient statistical precision. In addition, the detector must provide particle identification to distinguish the electronappearance events from background neutral current π0 events. Two detector technologies are under consideration for thisproject: water Cherenkov and liquid argon Time Projection Chamber (TPC).The water Cherenkov detector technology is relatively well understood from the experience of the Super-Kamiokande

experiment [20]. The Cherenkov light associated with the produced charged lepton is incident on the photomultiplier tubesdeployed on the inner surface of the cylindrical container. This light forms characteristic ‘‘ring’’ patterns that can be usedto reconstruct the momentum direction and energy of a relativistic charged particle. Muons generate relatively sharp ringssince they propagate primarily as single penetrating particles. Electrons interact in thewater to generate an electromagneticshower which results in a ‘‘fuzzy’’ ring pattern. These properties enable effective separation of electron events from muonevents. In addition, π0 decays result in 2 rings and so can generally be distinguished from single electron events. However,the π0 identification is not 100% efficient, resulting in a significant neutral current π0 background that must be understoodand subtracted. The statistical precision of the νe appearance measurements is degraded by this background. Nevertheless,

340 R.D. McKeown / Progress in Particle and Nuclear Physics 64 (2010) 334–341

Fig. 6. Projected sensitivity of a 100 kton liquid argon TPC detector with a 2MWwide band beam for 6 years running. The sensitivity curves for discoveringsin2 2θ13 are plotted vs. the value of the CP-violating phase [18].

with sufficient detector mass the statistical precision required for sin2 2θ13 & 0.01 can be achieved. Current estimatesindicate that a total fiducial mass of ∼300 kton is appropriate for this level of sensitivity. The total mass would then beover 500 kton, about 15 times the scale of the Super-Kamiokande detector. Although it may be feasible to excavate a singlecavern of this size, it might be preferable (technically and/or economically) to construct the detector in several (possibly2 or 3) smaller modules. The presently envisioned photocathode coverage is about 25% which would require about 60,000photomultiplier tubes per 100 kton (fiducial mass) module.A liquid argon TPC detectorwould enablemore detailed reconstruction of the neutrino events. Charged particles generate

ionization along the particle trajectory. The high purity and noble liquid property of liquid argon enables the drift of theelectrons along an electric field for distances of several meters. The drift velocity is about 1 mm/µs so the drift time givesthe z coordinate (assuming EE ‖ z). The drifted electrons are then detected on a wire plane, and the x-y position is obtainedfrom the wire readout pattern. Few mm spatial resolution is possible with this method. Although small prototypes havebeen operated successfully, there is no experience with large kiloton size detectors. In addition to the cost of such a largedevice, the issues associated with such large volumes of cryogenic liquid must be managed. Nevertheless, if such detectorscan be demonstrated to be technically and economically feasible then they would offer superior performance (comparedto water Cherenkov detectors) for background rejection. This improved background rejection would improve the statisticalsensitivity of the TPC relative to a water Cherenkov detector (for the same mass). Therefore it is likely that a 50–100 ktonTPC would have comparable sensitivity to a 300 kton water Cherenkov detector. The projected sensitivity for sin2 2θ13 insuch an experiment is shown in Fig. 6.It should bementioned that such large underground detectors as described herewould enable very interesting additional

science goals. These additional capabilities include extending the reach for detection of nucleon decay, galactic and extra-galactic supernova neutrino detection, high-precision solar ν studies, and detailed studies of atmospheric neutrinos.

6. Conclusion

Clearly the last decade has produced remarkable progress in our understanding of neutrino oscillations and masses. Themass splittings are now well-determined and we have good quantitative knowledge of two of the three mixing angles:θ12 and θ23. The last unknown mixing angle θ13 has eluded experimental discovery thus far, with the CHOOZ experimentproviding only an upper limit. The presence of CP violation (through the phase δ, see Eq. (3)) in the neutrino mixingmatrix requires that θ13 6= 0. Such CP violation would lend support to the leptogenesis scenario [21] for generating thematter–antimatter asymmetry in the universe, and so its experimental study is of great interest. Determination of the valueof θ13 is essential in the planning and design of future neutrino oscillation experiments to determine δ as well as the masshierarchy.Establishing that θ13 6= 0 is now a major priority for the experimental program. The three reactor experiments (RENO,

Double CHOOZ, and Daya Bay) along with the new long baseline experiment T2K are preparing to address θ13 with higherprecision, down to the level of sin2 2θ13 . 0.01. This represents an order of magnitude improvement in sensitivity overpresent experiments.If it is indeed found that sin2 2θ13 & 0.01, then future long baseline experiments such as NOνA and the proposed Fermilab

to DUSEL experiment will have good sensitivity to the CP-violating phase δ and the mass hierarchy. However, if θ13 is muchsmaller then the future determination of θ13 will be much more difficult. The Fermilab to DUSEL experiment could have

R.D. McKeown / Progress in Particle and Nuclear Physics 64 (2010) 334–341 341

sensitivity down to sin2 2θ13 ∼ 2–3 × 10−3 (see Fig. 6), but measuring δ and the mass hierarchy would be much moredifficult with this experiment if θ13 is that small. Thus it seems that if sin2 2θ13 � 0.01 future experimental efforts to studyCP violation and the mass hierarchy will likely require new facilities with even greater capability. Present concepts include‘‘beta beams’’ from radioactive decay of nuclei in storage rings [22] or ‘‘neutrino factories’’ that produce neutrinos fromthe decay of stored muons [23]. These concepts are presently undergoing research and development to assess the ultimatefeasibility for future construction.

Acknowledgements

I would like to thank Petr Vogel and Bonnie Fleming for helpful comments on this manuscript.

References

[1] B. Pontecorvo, Sov. Phys. JETP 6 (1958) 429;B. Pontecorvo, Sov. Phys. JETP 33 (1967) 549.

[2] Z. Maki, M. Nakagawa, S. Sakata, Progr. Theoret. Phys. 28 (1962) 870.[3] Y. Ashie, et al., Phys. Rev. D71 (2005) 112005.[4] J.M. Paley, et al., arXiv:0901.2131.[5] S. Abe, et al., Phys. Rev. Lett. 100 (2008) 221803.[6] C. Bemporad, G. Gratta, P. Vogel, Rev. Modern. Phys. 74 (2002) 297.[7] P. Vogel, J.F. Beacom, Phys. Rev. D60 (1999) 053003;A. Kurylov, M.J. Ramsey-Musolf, P. Vogel, Phys.Rev. C67 (2003) 035502.

[8] V. Barger, D. Marfatia, K. Whisnant, Internat. J. Modern. Phys. E12 (2003) 569–647.[9] M. Appolonio, et al., Eur. Phys. J. C27 (2003) 331.[10] F. Boehm, et al., Phys. Rev. D64 (2001) 112001.[11] L.A. Mikaelyan, V.V. Sinev, Phys. Atomic Nucl. 63 (2000) 1002.[12] F. Ardellier, et al. Double Chooz: A search for the neutrino mixing angle θ13 , 2006 arXiv:hep-ex/0606025.[13] Daya Bay Collaboration, A precision measurement of the neutrino mixing angle θ13 using reactor antineutrinos at Daya Bay, 2007 hep-ex/0701029.[14] K.K. Joo, for the RENO collaboration, Nucl. Phys. B168 (2007) 125.[15] R.D. McKeown, P. Vogel, Phys. Rep. 394 (2004) 315.[16] D.S. Ayers, et al. NOνA Technical Design Report, FERMILAB-DESIGN-2007-01, 2007 http://www-nova.fnal.gov/nova_cd2_review/tdr_oct_23/tdr.htm.[17] G.L. Fogli, et al., arXiv:0905:3549.[18] Fermilab-0801-AD-E, BNL-77973-2007-IR, May 9, 2007 http://nwg.phy.bnl.gov/~diwan/nwg/fnal-bnl/report.pdf.[19] Particle physics project prioritization panel (P5) US particle physics: Scientific opportunities. A Strategic Plan for the Next Ten Years, June 2, 2008

http://www.science.doe.gov/hep/files/pdfs/P5_Report06022008.pdf.[20] S. Fukuda, et al., Nucl. Instrum. Methods Phys. Res. A 501 (2003) 418.[21] M. Fukugita, T. Yanagida, Phys. Lett. B174 (1986) 45.[22] P. Zucchelli, Phys. Lett. B532 (2002) 166;

M. Mezzetto, hep-ex/0302007.[23] J.J. Gomez-Cadenas, D.A. Harris, Ann. Rev. Nucl. Part. Sci. 52 (2002) 253;

P. Huber, M. Lindner, W. Winter, Nucl. Phys. B645 (2002) 3;V. Barger, et al., Phys. Rev. D 62 (2000) 073002;S. Geer, J. Phys. G. 29 (2003) 1485.