X The values of all neutrino oscillation parameters in the 3ν scheme can now be extracted from global fits of available data with a precision better than 15%, the largest remaining uncertainty being on θ 23 and its octant. The relatively large value of θ 13 (which drives the ν e  ν μ oscillation) is an asset for the determination of the remaining unknown neutrino parameters: the value of the CP-violating phase δ CP, and the ordering of the neutrino masses, aka the neutrino mass hierarchy (NMH: normal or inverted, see plot). Standard strategy for the determination of NMH: probe ν e  ν μ oscillation in presence of matter effects which allow to resolve the sign of Δm 2 13 : X Measuring the neutrino mass hierarchy with atmospheric neutrinos Topical Research Meeting on Prospects in Neutrino Physics London, December 19-20, 2013 ORCA (Oscillation Research with Cosmics in the Abyss): a feasibility study for the measurement of neutrino mass hierarchy with atmospheric neutrinos in the Mediterranean Véronique Van Elewyck (Laboratoire APC & Université Paris 7 Denis Diderot) on behalf of the KM3NeT Collaboration ORCA (Oscillation Research with Cosmics in the Abyss): a feasibility study for the measurement of neutrino mass hierarchy with atmospheric neutrinos in the Mediterranean Véronique Van Elewyck (Laboratoire APC & Université Paris 7 Denis Diderot) on behalf of the KM3NeT Collaboration References Current performances with the reference detector – ν μ CC (track-like) channel Peformance studies have been focusing so far on the ν μ charged-current channel which provides a relatively clear experimental signature (muon track + hadronic shower). Semicontained events (i.e. with the reconstructed vertex inside the instrumented volume) selected with a quality criterium on the track fit provide reasonably good angular accuracy (median Δθ better than 10° for E ν ≥ 5 GeV). The corresponding effective mass of the detector reaches its plateau value of ≈ 1.8 Mton at E ν ≈ 8 GeV. The muon energy estimation is most reliable when an additional containment condition is imposed on the full muon track, ensuring that most of the Cherenkov light is deposited inside the detector. Using the median reconstructed energy as an estimator of the neutrino energy, a resolution of 35% can be reached at 10 GeV. Studies are ongoing to evaluate the energy deposited in the hadronic shower in order to improve the determination of the neutrino energy. Further studies are also conducted to evaluate the possible contribution from the cascade-like channels ( ν e and ν τ CC, all flavours NC); preliminary results suggest a significant enhancement of the mass hierarchy discrimination power (see e.g. Ge & Hagiwara, arXiv ). Δ m 2 13 NORMAL INVERTED The oscillation enhancement is maximal at resonant energy E res ≈ few GeV for Earth matter density: good prospects for atmospheric neutrinos ! atmospheric ν e & ν μ (anti-)neutrino fluxes XX ORCA: an undersea neutrino Cherenkov detector within the KM3NeT framework Simulation chain E, θ smearing (kinematics + detector resolution) First-guess vertex position ORCA is building on the expertise of the KM3NeT Collaboration, which aims at the deployment of a multi-km3 neutrino telescope in the Mediterranean to perform high-energy (TeV-PeV) neutrino astronomy: - same detection principle: Cherenkov light emitted by the secondary particles produced by the neutrino interaction with matter in and around the detector - same technology and detector design (instrumented lines anchored at the seabed and supporting multi-PMT digital optical modules) - denser array to lower the energy threshold to ~GeV neutrino energies - line spacing and length limited by deployment contingencies A reference detector is used for the first stages of the feasibility study: (full optimisation to be performed on basis of an overdense detector with different optical module masking options) - atmospheric ν e + ν μ (+anti) 1  500 GeV (Bartol flux, Agrawal et al. PRD 53, (1996) ) - neutrino interaction: QE+RES+DIS (GENHEN), both charged and neutral current; - For all particles crossing the can volume (size ~3 absorption lengths around the detector): tracking + Cherenkov light emission - down-going atmospheric muon background based on MUPAGE parameterization ( Carminati et al., Comput. Phys. Commun. 179 (2008) ) - optical background (40K decay: flat noise, 5kHz/PMT Hz in-DOM time-correlated Maximal effect at ≈ few GeV for Earth matter densities (depending on the baseline) BUT P(ν μ  ν μ, NH) = P(ν μ  ν μ, IH): other ingredients are needed to create a difference in the number of events IH vs. NH: Distinctive patterns in (E ν, cos θ ν ) oscillograms showing the difference in expected number of events for normal and inverted hierarchies MAIN CHALLENGES: -NH/IH difference intrinsically small - further limited by kinematics and intrinsic fluctuations - oscillograms blurred by limited energy and angular accuracy of the detector References Muon background rejection capabilities Impact of systematics R ν < 75 m (solid) A neutrino beam to ORCA ? NH IH Counting events related to a neutrino beam of given flavour can help increase the sensitivity of ORCA to the mass hierarchy even with a moderate capability of flavour determination in the detector. In the appearance channel ν μ  ν e and for a typical ν μ beam in the energy range 2-8 GeV, an optimal separation between NH and IH is found for cosθ = 0.2, i.e. a baseline of 2600 km. The oscillation probabilities are only marginally affected by the value of δ CP. A beam with a suitable baseline could be built from the IHEP in Protvino (near Moscow, Rus- sia) towards one of the potential sites for ORCA, with a baseline of 2588 km and a moderate (~11.7°) inclination. Assuming N pot ≈1.5*10 21 within a few years, a 7σ measurement of the hierarchy can be achieved (3σ with 3-4% systematic uncertainty on the event rate). Intrinsic detector capabilities are not the only factor that can spoil the measurement of the mass hierarchy. Other sources of uncertainties can lead to biases in the statistical distributions that may result in a degraded discrimination power or in wrong hierarchy determination (false positives). Several independent studies have addressed the impact of systematic errors on the sensitivity and identified the most critical among them.    (deg)     (deg)  m 2 large  (eV 2 )  Earth density profile (PREM model): negligible impact  Atmospheric neutrino fluxes:  moderate impact of the shape uncertainties  large overall normalization uncertainty can be mitigated by anchoring the fit at E ν > 20 GeV  Oscillation parameters: significant impact of uncertainties in atmospheric sector (Δm 2 large, θ 23, θ 13 ), in particular θ 23 octant ; solar sector and δ CP have negligible impact ORCA can improve on the precision of Δm 2 large, θ 23 through global fit procedure ! MSW effect in Earth matter (at different baselines) (anti-)neutrino cross-section: σ νN ≈ 2σ νN The main physical background in the detector comes from the abundant flux of down-going atmospheric muons which can be misreconstructed as upgoing, mimicking a ν μ CC event. Preliminary studies show that the muon contamination (Cμ) can be efficiently reduced by cuts on the reconstructed muon track parameters, while maintaining a high neutrino signal efficiency: -angular error estimate: β < 2° -track fit quality parameter: Λ > (-4)-(-4.5) - reconstructed vertex position: R ν < m (from detector centre) -muon track reconstruction: likelihood maximisation based on the hit time residuals (ANTARES-like) + length estimation on basis of first/last emission points -Identification of hits belonging to the hadronic shower and re-estimation of the vertex position (assuming spherically expanding shower) - full shower reconstruction (for NC and e/τ CC)currently under study Reconstruction strategy PRELIMINARY NH IH instrumented volume Expected number of events The performance of ORCA for the determination of the neutrino mass hierarchy is assessed by means of a likelihood ratio test with nuisance parameters: For each toy dataset generated with a given hierarchy, the mixing parameters are fit with either IH or NH and the log-likelihood ratio is computed. The separation of the two resulting distributions is used to compute the mass hierarchy significance. where Statistical method The sensitivity curves shown hereabove have been obtained from a preliminary study with simplified assumptions: - - perfect muon zenith angle reconstruction - - true neutrino vertex contained - - at least 15 hits (from GEANT4 simulation) - - no reconstruction assumed  2.5  4σ in 5 years with the reference detector  3  5σ in 3 years with 5 Mton with a significant dependence in the reconstruction performance More detailed studies ongoing… Sensitivity study 5 years with 1.75 Mton GeV (50% Probability) PRELIMINARY 3 years with 5 Mton Formaggio, Zeller, Rev. Mod. Phys. 84(3) 1307 (2012) M. Honda et al., Phys. Rev.D 75 (2007), E. Akhmedov et al., JHEP 02 (2013), 082 D. Franco et al., JHEP 04 (2013) 008 W. Winter, Phys. Rev. D88 (2013) J. Brunner, arXiv: