LAGUNA and LBNO: Towards the next generaon neutrino observatory in Europe

Thomas Patzak

ICFA Meeting, January 8 – 10, 2014, , Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 1 LAGUNA/LBNO consortium! Large Apparatus for Grand Unificaon and Neutrino Astrophysics and Long Baseline Neutrino Oscillaons • LAGUNA DS (FP7 Design Study 2008-2011) – ~100 members; 10 countries, 1.7 M€ – 3 detector technologies ⊗ 7 sites, different baselines (130 → 2300km) • LAGUNA-LBNO DS (FP7 DS Long Baseline Neutrino Oscillations, 2011-2014) Steering group: – ~300 members; 14 countries + CERN, 4.9 M€ Alain Blondel (UniGe) Ilias Efthymiopoulos (CERN) – Down selection of sites & detectors Takuya Hasegawa (KEK) Yuri Kudenko (INR) • LBNO (CERN SPSC EoI for a very long baseline Guido Nuijten (Rockplan, ) Lothar Oberauer (TUM) neutrino oscillation experiment, June 2012) Thomas Patzak (APC, Paris) – An incremental approach, based on the findings of LAGUNA Silvia Pascoli (Durham) Federico Petrolo (ETH Zürich) – ~230 authors; 51 institutions André Rubbia (ETH Zürich) Wladyslaw Trzaska (Jyväskyla) – CERN-SPSC-2012-021 ; SPSC-EOI-007 -> WA105 Large Alfons Weber (Oxford) scale prototype at CERN Marco Zito (CEA)

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 2 Towards a real experiment: SPSC-EoI-007: «Expression of Interest for a very long baseline neutrino oscillation experiment (LBNO)»

1. III. Physikalisches Institut, RWTH Aachen, Aachen, 2. Middle East Technical University (METU), Ankara, Turkey 3. Ankara University, Ankara, Turkey 4. LAPP, Université de Savoie, CNRS/IN2P3, F-74941 Annecy-le-Vieux, France 5. Institute of Nuclear Technology-Radiation Protection, National Centre for Scientific Research ”Demokritos”, Athens, Greece 6. INFN and Dipartimento interateneo di Fisica di Bari, Bari, 7. University of Bern, Albert Einstein Center for Fundamental Physics, Laboratory for High Energy Physics (LHEP), Bern, Switzerland 8. Faculty of Physics, University of Bucharest, Bucharest, Romania 9. INFN Sezione di Cagliari, Cagliari, Italy 10. INFN Sezione di Cagliari and Università di Cagliari, Cagliari, Italy 11. University of Cambridge, Cambridge, United Kingdom 12. Universita‘ dell’Insubria, sede di Como/ INFN Milano Bicocca, Como, Italy 13. Alan Auld Engineering, Doncaster, United Kingdom 14. Joint Institute for Nuclear Research, Dubna, Moscow Region, Russia 15. Institute for Particle Physics Phenomenology, Durham University, United Kingdom 16. Technodyne International Limited, Eastleigh, Hampshire, United Kingdom 17. INFN Laboratori Nazionali di Frascati, Frascati, Italy 18. University of Geneva, Section de Physique, DPNC, Geneva, Switzerland 19. University of Glasgow, Glasgow, United Kingdom 20. University of Helsinki, Helsinki, 21. Rockplan Ltd., Helsinki, Finland 22. Department of Physics, University of Jyväskylä, Finland 23. Physics Department, Lancaster University, Lancaster, United Kingdom 24. University of Liverpool, Department of Physics, Liverpool, United Kingdom 25. Imperial College, London, United Kingdom 26. Queen Mary University of London, School of Physics, London, United Kingdom 27. Dept. of Physics and Astronomy, University College London, London, United Kingdom 28. Institute for Nuclear Research of the Russian Academy of Sciences, Moscow, Russia 29. INFN Sezione di Napoli and Università di Napoli, Dipartimento di Fisica, Napoli, Italy 30. University of Manchester, Manchester, United Kingdom 31. INFN Milano Bicocca, Milano, Italy 32. University of Oulu, Oulu, Finland 33. Oxford University, Department of Physics, Oxford, United Kingdom 34. STFC, Rutherford Appleton Laboratory, Harwell Oxford, United Kingdom 35. AGT Ingegneria S.r.l., Perugia, Italy 36. UPMC, Université Paris Diderot, CNRS/IN2P3, Laboratoire de Physique Nucléaire et de Hautes Energies (LPNHE), Paris, France 37. APC, AstroParticule et Cosmologie, Université Paris Diderot, CNRS/IN2P3, CEA/Irfu, Observatoire de Paris, Sorbonne Paris Cité Paris, France 38. Università and INFN Roma Tre, Roma, Italy 39. INFN Roma Tre, Roma, Italy 40. IRFU, CEA Saclay, Gif-sur-Yvette, France 41. University of Sheffield, Department of Physics and Astronomy, Sheffield, United Kingdom 42. Department of Atomic Physics, Faculty of Physics, St.Kliment Ohridski University of Sofia, Sofia, Bulgaria 43. Petersburg Nuclear Physics Institute (PNPI), St-Petersburg, Russia 44. IPHC, Université de Strasbourg, CNRS/IN2P3, Strasbourg, France 45. INFN Trieste, Trieste, Italy 46. IFIC (CSIC & University of Valencia), Valencia, 47. Université de Lyon, Université Claude Bernard Lyon 1, IPN Lyon (IN2P3), Villeurbanne, France 48. National Centre for Nuclear Research (NCBJ), Warsaw, Poland 49. Institute of Experimental Physics, Warsaw University (IFD UW), Warsaw, Poland 50. University of Warwick, Department of Physics, Coventry, United Kingdom 51. ETH Zurich, Institute for Particle Physics, Zurich, Switzerland

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 3 Laguna-LBNO: Large Apparatus for Grand Unificaon and Neutrino Astrophysics and Long Baseline Neutrino Oscillaons

?

LAGUNA Physics: 1. Accelerator based: • Mass Hierarchy • δCP large θ13 • MSNP precision • 3 ν or 3+n ?

2. Non-Accelerator based: • Proton decay

3. Neutrino Astronomy: • Supernova neutrinos • Diffuse Supernova Neutrinos (DSN) • Solar Neutrinos • Atmospheric Neutrinos

4. Dark Maer

The highly important discovery of the Higgs at CERN July 4th 2012 has crowned the SM, Neutrino physics provides us with surprises! ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 4 Strategy

 Based on the findings from LAGUNA and LAGUNA-LBNO the collaboration decided to put forward a concrete proposal for the future neutrino observatory in 2012, EoI 007 to CERN.

 We have compared 7 locations in Europe and conducted precise estimations on the costs of the facility, of the detector and of the beam.

 We compared the physics potential of all possible combinations - detector - location - beam.

 The conclusion is to propose a neutrino observatory with a clear long-term strategy in a deep underground location at the longest baseline proposed, 2300 km, compatible with: . A full astro-particle program . Competitive nucleon decay measurements wrt SK and . An incremental long-baseline program with a competitive 1st stage guaranteeing high level physics performance from the beginning. (e.g. MH at 5σ CL with ~100% power in ~5years) . Stage 1 is based on a 20 kt fid. LAr detector (double phase) and a conventional beam from the CERN SPS of 700 kW.

 If the findings from stage 1 require, the detector and the beam will be upgraded to 70 kt and 2 MW.

 The location of the infrastructure is perfectly adapted to a neutrino factory, allowing the ultimate measurements in the accelerator neutrino field.

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 5 Mass Hierarchy is a fundamental measurement:

• MH is a prerequisite to study leptonic CPV • Scenarios for lepto-genesis • Hints for theory development (GUT models) • Feasibility and interpretation of 0νββ experiments • Interpretation of HDM from cosmology in terms of ν masses

LBNO strategy on MH: • To guarantee the measure MH on the > 5σ level one need to go to very long baselines > 2000 km. • MH should be settled early in the exp. to optimize the ν / ν ratio to maximize CP sensitivity. • The median 5 σ sensitivity (p = 0.5) for LBNO is reached within 2 years of running. • The guaranteed 5 σ sensitivity (p ~ 1) for LBNO is reached within 5 years of running. • Global fits of many experiments can guide and help the research but cannot replace the measurement of a dedicated experiment. • LBNO aims at exploring and resolve the mass hierarchy and the CP-phase problem by observing clear signatures and ascertaining their L/E dependence.

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 6 Mass Hierarchy with LBNO In neutrino physics we have a long history of few σ anomalies, hints and excesses…

• Sterile neutrino:  We got the reactor anomaly, the Gallium anomaly, the LSND excess, the MinibooNE excess/anomaly  All with 2 – 3 σ and the problem is still not solved  This calls for a conclusive 5 σ experiment  ν-STORM?

• Mass Hierarchy:  Many 2 σ, 3 σ maybe 4 σ experiments proposed using atm-ν’s (e.g.: INO, PINGU/ORCA) or reactor-ν’s (e.g.: JUNO,RENO-50).  Very sensitive to assumptions on the detector performance, small changes lead to vanish the projected CL reach.  The reach of future long baseline experiments depend on the knowledge of oscillation priors and systematics.  Is 1% for systematics realistic / achievable ?  No experiment is sheltered from an underestimated systematic error!  All physics potentials are evaluated using the median experiment P = 0.5, so 50% of the exp. will not reach the projected CL!  The sum of all these few sigma experiments does not give a conclusive answer to the MH problem.  This calls for a conclusive and guaranteed 5 σ experiment = LBNO!

• Today we are not in a situation that we do not know how to do such an experiment, with LBNO we have a concrete, well studied proposal on the table since 2012.

• The power of LBNO comes from the baseline of 2300 km and the possibility to switch beam polarity allowing to determine the MH at > 5 σ over the whole phase space with sufficient « margin » with respect to systematics and knowledge on oscillation priors.

In order to convince the community we need more than one experiment measuring the same parameters with a completely different method and detector technology!

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 7 MH sensitivityLBNO%poten+al%for%MH% and unique power of LBNO

• Power%vs%exposure%for%all%values%of%δCP%(shaded%bands)% β β 1 1 p = 1- 0.8 3σ p = 1- 0.8 3σ

5 0.6 σ 0.6 5σ Test power for NH Test power for IH 0.4 0.4

0.2 0.2 arXiv:1312.6520 0 0 0 1 2 3 4 5 0 1 2 3 4 5 Exposure (/1e20 POT) Exposure (/1e20 POT)

p = 0.5p%=%0.5%=>%“Median%experiment”% => “Median experiment” p%~1%=>%“Full%power%experiment”%p ~1 => “Full power experiment” 50%%chance%%to%give%the%wrong%answer% ~0%chance%%to%give%the%wrong%answer%% 50% chanceif%the%alterna+ve%hypothesis%is%true% not to achieve the if%the%alterna+ve%hypothesis%is%true% ~ 100% chance to achieve the projected %CL. % projected CL! WOULD%YOU%BET%1B€%ON%SUCH% THE%LBNO%CHOICE%TO%QUOTE% AN%EXPERIMENT%?% One should not bet on marginal THE SENSITIVITY%LBNO CHOICE TO QUOTE NuPhys%Dec.19th,2013physics% reach! A.Tonazzo% SENSITIVITY 13%

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 8 LBNO Strategy on Mass Hierarchy

Mean value of the mass hierarchy test statistic as a function of 2 20 true δCP and the value of sin Θ23 for an exposure of 4 × 10 pots (or about 5 years of running at the SPS) and LBNO 20 kton LAr double phase detector. > 2 > 0 300 0 300

2 2 2 T sin =0.40 T θ23 sin θ23=0.40 MH determination (NH assumed) sin2θ =0.44 MH determination (IH assumed) 2 23 sin θ23=0.44 2 2 50% nu+50% anu sin θ23=0.50 50% nu+50% anu < ΔΧ sin =0.50 2 θ23 250 4.0e20 pots sin θ =0.55 250 4.0e20 pots 2

23 < ΔΧ 2 sin θ =0.55 2 23 sin θ23=0.60 sin θ23=0.60 200 200

150 150

100 100

50 50

0 myfitter 0 myfitter 0 1 2 3 4 5 6 0 1 2 3 4 5 6

True δCP True δCP

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 9 CP Violation with LBNO

‣ Measure δCP by measuring the energy dependence of the neutrino spectrum (L/E behavior) and the 2nd maximum, this is highly complementary to the HK proposal which measures only the ratio at the first maximum.

‣ Measure all transitions: • Appearance: νµ→ νe and νµ→ ντ • Disappearance: νµ→ νµ • neutral currents

‣ Neutrino and anti-Neutrino beams

‣ Incremental approach of LBNO:

• Phase 1:

. Conventional beam based on 400 GeV protons from the SPS 700 kW . Total 1.5 x 1021 PoT (10 - 12 years) . 20 kt LAr double phase detector . This determines MH at > 5σ within 4 – 5 years. . The knowledge of MH allows to adjust the ν / ν ratio to maximize CP sensitivity. . Adjust beam to 2nd max to enhance CPV sensitivity

• Phase 2:

. Upgrade detector to 70 kt and / or the beam power to 2 MW . WBB to fully explore L/E and the 2nd maximum. . This allows CPV discovery at > 5σ level

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 10 5σ 2 χ CPV discovery 100% ν 0%ν

Δ 14 C2P: 15e20 pots 75% ν 25%ν δ = 0,π exclusion with NH CP 50% 50% 12 ν ν 25% ν 75%ν 0% ν 100%ν 10 3σ 8 5σ LBNO Strategy on δCP 6

4 Once MH determined run for90%C.L. 5 more years with 2

optimized sharing of neutrinos / anti-neutrinos tomyfitter 0 cover the most possible0 phase1 2space3 in4 δ 5 6 CP True δCP 2 2

χ 100% 0% CPV discovery CPV discovery 100% ν 0%ν 14 Δ 14 C2P: 15e20 pots 75% 25% C2P: 15e20 pots 75% 25% ν ν CP = 0, exclusion with IH δ = 0,π exclusion with NH 50% 50% IH CP 50% ν 50%ν NH 12 12 25% 75% 25% ν 75%ν 0% 100% 0% ν 100%ν 10 10 3 3σ 8 8

6 6

4 4 90%C.L. 90%C.L. 2 2 myfitter myfitter 0 0 0 1 2 3 4 5 6 0 1 2 3 4 5 6 True CP True δCP Figure 12: CPV sensitivity for different sharing between ⌫ :¯⌫ running modes, for 1.5 1021 protons 2 CPV discovery 100% 0% on target and LBNO 20 kton detector. The upper plot is for NH and the lower plot for IH. 14 C2P: 15e20 pots 75% 25% = 0, exclusion with IH CP 50% 50% 12 25% 75% 0% 100% 686 consider both the oscillation maxima as well as the spectral shape is a very powerful method to 10 Design value: 75 % ν - 25 % anti-ν 3 687 extract CP and to confirm the oscillatory behaviour predicted in the three neutrino oscillation 688 schema together with matter effects. This approach is the only one proposed to put in evidence ICFA Meeting, January8 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 11

6 21 4 90%C.L. 2 myfitter 0 0 1 2 3 4 5 6

True CP

Figure 12: CPV sensitivity for different sharing between ⌫ :¯⌫ running modes, for 1.5 1021 protons on target and LBNO 20 kton detector. The upper plot is for NH and the lower plot for IH.

686 consider both the oscillation maxima as well as the spectral shape is a very powerful method to 687 extract CP and to confirm the oscillatory behaviour predicted in the three neutrino oscillation 688 schema together with matter effects. This approach is the only one proposed to put in evidence

21 LBNO Strategy on δCP Use all spectral information: Rate & Shape for energy range 1st - 2nd max

689 the theoretical framework of oscillations as a whole. In fact, the spectrum shape as well as 690 the number of events strongly depend on the value of CP in particular in the energy region nd 691 corresponding to the 2 maximum. We have compared the significance of our standard 692 method to a first maximum only and a rate only analysis. The study of the significance of nd 693 the events around the 2 oscillation maximum was done by evaluating the CPV sensitivity 694 with a cut on the reconstructed energy of the e-like events placed at 2.5 GeV. This effectively nd 695 removed all information about the 2 maximum from the e-like sample. In addition we have 696 tested the importance of performing an analysis based on the e-like event distributions by a 697 rate only analysis evaluation. The rate only measurement leads to a drastic loss of sensitivity 698 of the experiment to the CPV. These studies are shown in Figure 13. The important quantity 699 in this plot is the width of the interval below the curve for a given confidence level, which 700 tells us the fraction of unknown parameter space for which we would be able to discover CP 701 violation. As can be seen in this plot, the rate only measurement leads to a drastic loss of 702 sensitivity of the experiment to the CPV. The power of measuring events over an energy st nd 703 range that covers the 1 and the 2 oscillation maxima is also evident.

2 20 χ

15E+20 POT: ν 75%, ν 25% Δ Nominal fit 15 Fit with a cut at 2.5 GeV Rate only fit

10 3σ

5 1 σ error band on δCP = 0 90%C.L. 0 0 1 2 3 4 5 6 δ (rad) Eν (GeV) CP 2nd max 1st max FigureICFA Meeting, 13: Comparison January 8 – 10, of the2014, CPV Paris, sensitivities France of a rate only analysis, anThomas analysis Patzak with - LAGUNA-LBNO a APC, Université Paris-Diderot 12 cut on a reconstructed energy at 2.5 GeV (excluding the 2nd maximum), and the nominal case where the full event spectrum is used.

704 9.3 Impact of prior uncertainties on the CP discovery potential

705 The effects of the prior uncertainties on the oscillation parameters have been studied in detail. 706 The CP phase space coverage has been evaluated setting one prior at time for each oscillation 707 parameter according to Table 5. This is shown in Figure 14 where it is evident that the priors 708 with the largest impact is that on ✓13. 709 In Figure 15 we show the effects on the expected electron neutrinos energy spectrum 710 when values of ✓ and ✓ are varied by 1 for both the appearance and the disappearance 13 23 ±

22 60 60 all electron-like events , NH, =0 all electron-like events CC beam CC beam Integrale 1e+03 Integrale 1.2e+03 , NH, =0 CC misid 50 50 CC misid 0 0 NC misid NC misid

CC + -> e CC + -> e 40 40 entries / 200 MeV entries / 200 MeV 30 30

20 20

10 10

0 0 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 Reco energy (GeV) LBNOReco Strategy energy (GeV) on δCP 2 Figure 15: Measured e-like spectrum. Left: Maximum and minimum band if sin ✓23 is varied by 2 1.Right:MaximumandminimumbandifUse best knowledgesin 2✓13 is varied and by 1realistic in the appearance assumptions channel. on systematics and oscillation ± 5 ± 2 25 with all errors parameters CPV discovery C2P: 15e20 pots 75% +25% with prior on 13 at 5% = 0, exclusion with NH CP with prior on at 2.5% 2 50 20 13 2

χ 40

no systematics χ

Δ CPV discovery CPV discovery no systematics 45 prior only on θ at 10% Δ δ = 0,π exclusion with NH C2P: 15e20 pots 75% ν+25% ν 13 CP syst only on CC_e_ev prior only on at 5% 35 C2P: 15e20 pots 15 δ = 0,π exclusion with NH θ13 syst only on CC_τ_ev 40 CP 75% ν+25% ν prior only on θ13 at 2.5% syst only on NC_e_ev 35 30 syst only on CC_e_beam_ev 10 3 30 25 5σ 25 5σ 5 20 90%C.L. 20

15 0 myfitter 0 1 2 3 4 5 6 15 10 3σ True CP 10 3 5 90%C.L. σ

Figure 17: Same as Figure 16 but with all other systematic errors included.myfitter 0 0 1 2 3 4 5 6 5 5σ 90%C.L. True δCP myfitter 737 for around 30% of the parameter space. 0 0 1 2 3 4 5 6 Figure 16: Impact of systematic errors: CPV sensitivity of LBNO phase I as a function of True 2 20 2 δCP 2 CP , with only statisticalχ and no systematic errors (black), and effect of the error on the sin 2✓13 sin θ = 0.40 CPV discovery 2 23 Δ sin θ = 0.44 parameter prior of 10%18 (green),C2P: 15e20 pots5%75%(blue), ν+25% ν 2.5% (red). 23 ± ± ± sin2θ = 0.50 δ = 0,π exclusion with NH 23 CP 2 Figure 19: Impact of systematic errors: CPV sensitivity of LBNO phase I as a function of CP , sin θ23 = 0.55 16 2 sin θ = 0.60 23 with only statistical and no systematic errors (black), and effect of the error on the normalisation 14 of the signalThe and backgrounds.most important oscillation 9.3.2 Influence of12✓23 on the CP discovery potential parameters are 23 and θ13 and Now that the value10 of ✓313σ mixing angle has been measured, the knowledge of the mixing θ 60 angles which describe8 the PMNS matrix has changed significantly. Whilst previously ✓13 all electron-like events , NH, =0 w/o oscillation , NH, =0 CC beam 500 Integrale 1.2e+03 CC misid w/ oscillation was not known, the uncertainty on it had a dominant influence on the possible50 the discovery most important systematics 0 6 NC misid CC + -> e 400 reach of long-baseline facilities, now it makes sense to investigate also the influence of ✓23 4 40 (excluding CP ) whose uncertainty90%C.L. has as well a large impact. Its true value influencesis the the knowledge of the absolute entries / 200 MeV entries / 200 MeV 300 2 30 myfitter 0 200 0 1 2 3 4 5 6 20 rate of νe CC events.

True δCP 100 26 10 ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 13 0 0 Figure 18: Impact of the prior value of ✓23: CPV sensitivity of LBNO phase I as a function0 of CP1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 ✓ for a range of values of 23. Reco energy (GeV) Reco energy (GeV)

Figure 20: Measured neutrino spectra for (left) e-like appearance (right) muon-like disappearance The dependence of the discovery reach on ✓23 can be understood analytically by following channels, when all the normalisation errors listed in Table 6 are varied by 1 in a fully correlated ± way. Statistical error are also shown.

25

10 Ultimate CPV sensitivity

We have seen that the LBNO Phase I has significant physics goals, in particular it is guaranteed to be fully conclusive for MH discovery with an expected 5 C.L. over the full

29 2 588 of the function is correctly given by 2 = . Hence, for the CPV discovery, the 2 589 significance of the result will be computed as nq= . p 590 6.5 Assumption on parameters and systematics

591 Assumptions on the oscillation parameters and uncertainties as well as on the beam line 592 characteristics are shown in Table 5. Values take into account the results from ongoing 593 experiments, in particular reactor experiments, and are based on the global analyses published 594 in the literature. Uncertainties are given at 1 and in percent. We have chosen to use 595 the values available as of today, therefore our assumptions are conservative. In order to 596 describe matter effects, we use a constant average density approximation. We have compared Best knowledge and realistic597 the analytical oscillation assumptions probability obtained with the constant on value to the one computed 598 by integration of the oscillation amplitude in 50 steps through the Earth described by the 599 Preliminary reference earth model (PREM) [31]. As can be seen in Figure 8, the assumed 3 oscillation parameters600 value ofand 3.20 g/cm systematics…describes best the probability computed with the PREM. In the figure, 601 the band corresponds to the oscillation probabilities obtained by varying the density in the 3 3 602 interval 3.2 > ⇢ > 2.8 g/cm . The upper values of the band are found for ⇢ =3.20 g/cm .

Name Value error (1) error (%) β 1 L 2300 km exact exact δCP = π/2 2 5 2 m21 7.6 10 eV exact exact 2 3 2 ⇥ δ = 3π/2 m31 10 eV 2.420 0.091 3.75 %

p = 1- CP | | ⇥2 ± ± 0.8 sin ✓12 0.31 exact exact 2 sin 2✓13 0.10 0.01 10% 2 ± ± 90% sin ✓23 0.440 0.044 10% 3 3 ± ± 0.6 σ Average density of traversed matter (⇢) 3.20 g/cm 0.13 4% ± ±

0.4 Table 5: Assumptions on the values of the oscillation parameters and their uncertainties.

0.2 Name Value error (1) Signal normalization (f )15% sig ± Beam electron contamination normalization (f )1 5% ⌫e ± Tau normalization (f )120% 50% 0 ⌫⌧ ± ± ⌫ NC and ⌫ CC background (f )110% 0 2 4 6 8 10 12 14 16 µ NC ± Exposure (/1e20 POT) Table 6: Assumptions on event normalization uncertainties (bin-to-bin correlated errors).

Figure 21: Statistical power for CPV discovery as a function of exposure for 90% and 3 CL assuming NH. The far detector of 20 kton LAr and 750 kW SPS neutrino beam are603 assumed.The assumptions on systematic errors on signal and background normalization are shown 604 in Table 6. The systematic error on the tau normalization is set to 50% for the mass hierarchy • As show before statistically LBNO Phase 1 can reach605 5determinationσ on CPV. and to 20% for the CP sensitivity studies. This reduction is due to the fact that 606 the experiment will be able to constrain ⌫⌧ cross section with the data accumulated during • Current knowledge and conservative assumptions on607 systematics allow a 3 measurement of CPV with LBNO phase 1 772 neutrino beam power in order to decrease the statistical error around thefirst 2nd few oscillation years of running performingσ specific tau neutrino appearance channel measurements 608 to constrainnd the production rate. 773 maximum.• BecauseThe baseline of the natural of 2300 cut-o kmff of allows the muon-neutrino the measurement flux spectrum of the at 2 low energy,max, less sensitive to systematic effects. 609 These errors are assumed to be fully correlated among the energy bins. 774 and the linear increase of the total neutrino cross-section with energy, the 2nd maximum is 775 more difficult to study than the 1st maximum. However, this is still possible at the LBNO 776 baseline of 2300 km since the 2nd maximum is at an accessible energy of 1.5 GeV. Since the ⇠ 16 777 CP-asymmetry at the 2nd maximum is more sensitive to CP than at the first maximum, a 778 significant gain is obtainedFor by populating all thisdetails region with oscillationsee our events. This paper: is one of arXiv:1312.6520 779 the main goals of the LBNO Phase II. The expected CPV sensitivity as a function of CP 780 is shown in Figure 22 for various upgrades of beam power with the HP-PS, and of the far ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 14 781 detector mass, from 20 kton to 70 kton. With a new powerful proton driver such as the 782 conceptual HP-PS and a 70 kton detector mass, the coverage at > 5’s C.L. will be 54% ⇠ 783 after 10 years.

784 11 Summary and Conclusions

785 The LBNO experiment is the outcome of intense and comprehensive design studies supported 786 by the European Commission since 2008. In an incremental approach, we propose LBNO 787 with a 20 kton underground detector as the first stage of a new neutrino observatory able to 788 address long-baseline neutrino physics as well as neutrino astrophysics. The programme has a 789 clear long-term vision for future stages of the experiment, including the Neutrino Factory [34], 790 for which the baseline of 2300 km is well adapted. 791 Unlike the attempts to infer MH with atmospheric neutrinos in multi-megaton low- 792 threshold detectors [35], such as the one proposed with PINGU [36] or ORCA [37], or with 793 medium-baseline reactor experiments [39], such as JUNO [38], the accelerator-based approach 794 of LBNO addresses both fundamental problems of CPV and MH in clean and straightforward

28 LBNO Strategy on δCP Go to phase II to measure 5δ CPV: Increase mass and/or beam power

High power HP-PS study! Garoby

1.5 x 1021 p.o.t.

2 588 of the function is correctly given by 2 = . Hence, for the CPV discovery, the 2 589 significance of the result will be computed as nq= . afternoon! this p 590 6.5 Assumption on parameters and systematics

591 Assumptions on the oscillation parameters and uncertainties as well as on the beam line 592 characteristics are shown in Table 5. Values take into account the results from ongoing

593

experiments, in particular reactor experiments, and are based on the global analyses published Roland by talk See 594 in the literature. Uncertainties are given at 1 and in percent. We have chosen to use 595 the values available as of today, therefore our assumptions are conservative. In order to 596 describe matter effects, we use a constant average density approximation. We have compared 597 the analytical oscillation probability obtained with the constant value to the one computed 598 by integration of the oscillation amplitude in 50 steps through the Earth described by the 599 Preliminary reference earth model (PREM) [31]. As can be seen in Figure 8, the assumed 20 kton LAr + SPS(700kW) 3 600 value of 3.20 g/cm describes best the probability computed with the PREM. In the figure, 20 kton LAr + HPPS(2MW) 601 the band corresponds to the oscillation probabilities obtained by varying the density in the 70 kton LAr + SPS(700kW) 3 3 602 interval 3.2 > ⇢ > 2.8 g/cm . The upper values of the band are found for ⇢ =3.20 g/cm . 70 kton LAr + HPPS(2MW)

Name Value error (1) error (%) L 2300 km exact exact 2 5 2 m21 7.6 10 eV exact exact 2 3 2 ⇥ m31 10 eV 2.420 0.091 3.75 % | | ⇥2 ± ± sin ✓12 0.31 exact exact sin2 2✓ 0.10 0.01 10% 13 ± ± sin2 ✓ 0.440 0.044 10% 23 ± ± Average density of traversed matter (⇢) 3.20 g/cm3 0.13 4% ± ±

Table 5: Assumptions on the values of the oscillation parameters and their uncertainties.

Name Value error (1) Signal normalization (f )15% sig ± Beam electron contamination normalization (f )1 5% ⌫e ± Tau normalization (f )120% 50% ⌫⌧ ± ± ⌫ NC and ⌫ CC background (f )110% µ NC ± Table 6: Assumptions on event normalization uncertainties (bin-to-bin correlated errors).

ICFA 603Meeting,The January assumptions 8 – 10, on 2014, systematic Paris, errors France on signal and background normalization are shownThomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 15 604 in Table 6. The systematic error on the tau normalization is set to 50% for the mass hierarchy 605 determination and to 20% for the CP sensitivity studies. This reduction is due to the fact that 606 the experiment will be able to constrain ⌫⌧ cross section with the data accumulated during 607 first few years of running performing specific tau neutrino appearance channel measurements 608 to constrain the production rate. 609 These errors are assumed to be fully correlated among the energy bins.

16 Possibility of neutrinos from Protvino!

P2P

C2P

Desired parameters for neutrino beam: Decay tunnel length:

200-300 m!

≈2000 νµ CC / 20 kton / year (no osc.) C2P+P2P sensitivity under study ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 16 LBNO with 2nd beam from Protvino

20 kt LAr 70 kt LAr

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 17 LBNO an International Collaboration

The LAGUNA/LBNO collaboration has more than 200 members and includes institutions from nine CERN member states (Bulgaria, Finland, France, Germany, Greece, Italy, Poland, Spain, Switzerland, and the UK), as well as one candidate for accession to CERN (Romania) and two CERN observers (Russia, Turkey) and (KEK). - The LBNO collaboration is nearing completion of an EU-funded design study, and the detailed engineering and full costing of LBNO Pilot, Phase 1(20kton LAr) and Phase 2 (70kton) at Pyhäsalmi will be delivered in June 2014.

Status of discussions between LBNE and LBNO: - The scientific goals and chosen detector technology of LBNE and LBNO are very similar. Leadership of both LBNE and LBNO have agreed that working together towards our common goals would be mutually beneficial. - A task force to investigate joining forces between LBNE and LBNO, which has 5 members from each Executive Committee, meets every ~2 weeks. - A Joint Physics Task Force is carrying out a careful comparison of analyses and developing a common understanding of science strategy. • As of today we agree on:  Oscillation priors and systematics have a huge impact on the scientific output  MH is a prerequisite for CP  Measuring of L/E and the 2nd max is very important • We disagree on:  What are the realistic assumptions on priors and systematics  To evaluate the potential of the experiment as median (LBNE) or full power (LBNO)  The baseline  The synergy with HK and the need for more than one experiment - LBNO plans a program of development of their far detector, in particular the two-phase readout technology, including a 6 m3 prototype in a 8 m3 cryostat that would be a proof of principle that a large scale detector can be built. This will be carried out at CERN under the newly approved WA105 project and is expected to be completed around 2017. - LBNE and LBNO are discussing initial collaboration that would be centered on detector development under WA105, including common LAr technology, comparison of single- and double-phase readout, and use of the large cryostat for prototyping LBNE as well as LBNO detectors. ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 18 Synergy of LBNO with HyperKamiokande

st • HK measures δCP from the neutrino/anti-neutrino asymmetry at the 1 max. This is not sufficient toCrucial prove the fullinput 3 neutrino for mixingHK schema. • HK δCP sensitivity is highly dependent on the knowledge of MH

• The baseline of 2300MH km knowledge for the LBNO equivalent experiment to ~10 will years provide of HKan unambiguous running determination of MH. st nd • The baseline of 2300 km + WBB allows the measurement of the L/E behavior and the10 1 and 2 max. nd • The effect of δCP is larger at the 2 max. and systematics are less critical. • The baseline of 2300 km requires higher neutrino energies where X-sections are better known. • Independent cross check with two different detector technologies.

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 19 Conclusions

• The LBNO collaboration has put a real proposal on the table since 2012. • LBNO is the only proposed experiment which can guarantee a 5σ measurement of MH

• The early determination of MH is crucial to: • Tune the beam for the CPV measurement and • Provide the long awaited input to the community

• LBNO and LBNE look similar but LBNO has clearly the better scientific potential even with conservative assumptions. • LBNO has a clear strategy with a phased program with priority on MH in the 1st phase

• LBNO has real synergy and complementary to HK by: • Providing MH • Measuring CP in a different way using L/E and the 2nd max • The deployment of a fine-grained LAr detector is sensible only if one can make complementary measurements with respect to a statistically outnumbering detector like HK.

• LBNO will provide a very detailed, reliable and competitive costing by June 2014. • The 2300 km baseline of LBNO is perfect for the ultimate neutrino factory.

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 20 ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 21 Optimistic assumptions lead to better CP coverage, but how can we know???

Curves are simulated for 20 kt and 1.5e21 POT at 2300 km - LBNO conservative assumptions - LBNO optimistic assumptions

2 50

χ 5 sigma level reached

Δ 45 CPV discovery LBNO assumptions C2P: 15e20 pots 75% ν+25% ν during the first 40 LBNOphase with LBNE assumptions20 kt δCP = 0,π exclusion with NH 35 Name conventions: 30 Values as in the SPSC paper = CONSERVATIVE VALUES 25 5σ Values with red modifications = OPTIMISTIC VALUES 20

15 The most important differences are: - The value of θ Fogli et al. arXiv:1205.5254v3 (ours: Gonzales et al. arXiv:1209.3023) 10 3σ 23 - Error on sin^22θ13 - Systematics on signal and background 5 90%C.L. 0 myfitter 0 1 2 3 4 5 6

True δCP

ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 22 Can other experiments measure MH convincingly? NO ! Reactor experiments: Atmospheric Neutrinos: 14 e.g. JUNO: 2 σ if systematics are e.g. PINGU very sensitive to oscillation priors, h Entries 1764 extremely well controlled.Mean x 2.43 the hierarchy nature has chosen 3 Mean y ×10 2.02 0.96971200 RMS x 0.006387 NH TRUE and the value of RMS y 0.5427 δCP 1.5 1000 1.01 6 4 800 0.5 5

0.0α 0 600 3

4 Nominal NH, q23=50°, d=0 -0.5 NH s 400 s N N 3 IH, q =40°, d=p 23 2 -1.0-1 et al. arXiv:1309.1638v1 al. et 200 2 -1.5 IH

90% CL 1 2013 -2.0 0 1 Capozzi 2.40 2.41 2.42 2.42 2.43 2.43 2.44 2.44 2.45 2.45 2.46 2.46 F. 2 -3 2 GLoBES 2013 GLoBES Δ mee /10 [eV ] [arXiv:1305.5539] Winter, W. 0 0 0 2 4 6 8 10 12 14 0.35 0.40 0.45 0.50 0.55 0.60 0.65 sin2q true 2 2 t yr 23 FIG. 7: Constraints in the plane (∆mee, α)at1,2and3σ (∆χ =1, 4, 9) from a fit to prospective JUNO data assuming true normal hierarchy (α =+1).Althoughtheinvertedhierarchycase(α = −1) is ∼ 3.4σ away, the hierarchy discrimination is H L already compromised at ∼ 1.7σ,wherethe“undecidable”case(• NON of the proposedα =0)isallowed. experiments or any combinationFigure 5: Left of panel: them Number@ D of is forable the hierarchy to discoveryclaim as the a function of time for two extreme cases of the true parameter values (see legend). Right panel: Number of for the hierarchy sensitivity discovery of the fundamental parameter - the mass hierarchy!2 as a function of sin ✓23 (true) for the di↵erent hierarchies. The plot range corresponds to the current 3 2 VI. RESULTS OF THE ANALYSIS allowed range for sin ✓23, and the best-fit values and 1 ranges are marked by vertical lines and shadings, • The only way to discover the MH is a very long respectivelybaseline [4]. The accelerator default setup has been usedexperiment. in both panels (three years of data taking and =0in right panel). We discuss below the results of• our statisticalIn view analysis of the of prospect high ivelevel JUNO funding data as defined needed in the for previous such Section. experiment the new facility has to We focus on the case of true NH, the results for true IH being rather symmetrical. other discussed parameters (hierarchy, , ✓ octant) will be known before the construction guarantee 2the 5 σ discovery over the whole phase space for both solutions23 (NH Figure 7 shows the results of the fit in the plane (∆mee, α) for true NH, in terms of Nσ =1, 2, 3contoursforor even analysis time [7]. one parameter (∆χ2 =1, 4, 9), all other parameters being marginalized away. The errors are rather linear on both and IH) and in depended of being lucky or unluckyWhile with the two best-fitstatistical solutions forfluctuations.✓23 in Ref. [4] favor non-maximal atmospheric mixing, parameters, and appear to be significantly• The anti-correlated. community The needs anti-correlation this input stems from therefore, the tendency such of thethe fitexperiment to 3 allowed range is has much largerto be and includesdone the within possibility a of maximal atmospheric keep constant the oscillation phase 2∆ee + αϕ in Eq. (58) for typical neutrino energies E 3–5 MeV: an increasemixing. of We therefore show in Fig. 5, right panel, the ✓23 dependence in the currently allowed 2 ! ∆mee is then compensated by a decrease in α. 3 range. Obviously, the sensitivity for the normal (inverted) hierarchy can vary between In Fig. 7, the case of wrong hierarchyreasonable (α = 1) is formally time reached scale at of3.4 σ10; however, to 15 it wouldyears be misleadingfrom now. to1. take6 and 3.9 (1.3 and 3.3) for three years of data taking within the currently allowed − ∼ 2 this “distance” as a measure of theICFA sensitivity Meeting, January to 8 the – 10, hierarchy. 2014, Paris, France Ph ysically, the discriminationThomas of NH Patzak vs - LAGUNA-LBNO IH is successful3 range, depending on the actual valueAPC, of sinUniversité✓23. TheParis-Diderot kinks in this23 plot come from the 2 if the data allow to tell the sign of a non-L/E phase, which advances or retards a hierarchy-independent L/E oscillationoctant degeneracy, i.e.,theminimal is found in the inverted hierarchy, inverted octant phase. In our adopted formalism, this requirement amounts to tell the sign of α: when the sign is undecidable (α solution,0), and may be eliminated if the octant was known at the analysis time. A definitive conclusion! on the ✓23 octant from existing equipment is, however, unlikely [7] (see Fig. 5 the discrimination is already compromised. Therefore, the sensitivity to the hierarchy is more properly measuredtherein). by 2 2 the “distance” of the true case (either α =+1orα = 1) from the null case (α =0):Nσ χ (α = 1) χ (α =0). In Fig. 7, this distance is 1.7σ, namely, about 1/2− of the 3.4 sigma which formally separate! the NH± and− IH cases. Thus, we recover independently∼ the approximate “factor of∼ two” reduction of the sensitivity! with respect to5 naive Comparison with existing beam and reactor experiments 2 expectations [30, 37, 38], as expressed by the “rule of thumb” Nσ 0.5 ∆χ (NH IH) [42]. Our approach reaches such result via a fit to a continuous parameter (α), which is conveniently! treated− as any other floating parametersCompared to PINGU, other experiments will be sensitive to the mass hierarchy at the same ! timescale, where we especially focus on existing equipment with possible upgrades. The in the statistical analysis. The case of true IH (not shown) is very similar, with only a slight enhancement ofmost the sensitive experiment is the long-baseline neutrino oscillation experiment NO⌫Ainthe sensitivity to the hierarchy ( 1.8σ instead of 1.7σ). US, with a baseline of 810 km. In addition, data from the reactor experiments (Double In conclusion, the results∼ in Fig 7 show that∼ the sign of α (i.e., the advancement or retardation of phase due to the hierarchy) can be determined at a level slightly below 2σ. This value is in the same ballpark of all recent estimates under similar assumptions, but has been derived via∼ a different approach. In particular, we have recovered 11 2 the “rule of thumb” Nσ 0.5 ∆χ (NH IH) that was found and discussed in [30, 37, 38, 42] for two alternative discrete cases, by connecting! the two cases− via a continuous variable α, whose sign tells the hierarchy. The hierarchy discrimination is successful if the! data prefer α = 1 with sufficient significance with respect to α = 0; conversely, a preference for α = 0 would compromise the experiment,| | while surprisingly large values α 1 would signal possible systematics which are artificially enhancing the hierarchy effects. If the hierarchy discrimination| | $ is successfull, then 2 2 fit results such as those in Fig. 7 provide the central value and errorof∆mee and also of ∆m via Eq. (41), namely: sign(α) ∆m2 = ∆m2 (c2 s2 )δm2 . (76) ee − 2 12 − 12 The determination of the fundamental parameter ∆m2 thus depend also on the constraints achievable on the param- 2 2 eters (δm ,s12), which we now discuss. ICFA Meeting, January 8 – 10, 2014, Paris, France Thomas Patzak - LAGUNA-LBNO APC, Université Paris-Diderot 24