Very Long Baseline Oscillations

Size: px

Start display at page:

Download "Very Long Baseline Oscillations"

Reynard Garrison
6 years ago
Views:

1 Very Long Baseline Oscillations The BNL VLBL Concept Brett Viren Physics Department 7th International Workshop on Neutrino Factories and Superbeams Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 1 / 45

2 Outline 1 The Concept Oscillation Probability Illustrations 3 Example: BNL to Homestake 4 Making Very Long Baseline Neutrinos at BNL Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 / 45

3 The Concept 1 The Concept Goals and Overview Motivations for Basic Parameters Natural Benefits to Going High Energy Oscillation Probability Illustrations 3 Example: BNL to Homestake 4 Making Very Long Baseline Neutrinos at BNL Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 3 / 45

4 The Concept Goals and Overview Basics of the VLBL Concept The VLBL concept consists of three simple ideas: 1 Use a very long baseline, a wide band ν-beam, 3 at high ν-energies. This allows an ambitious but affordable experiment which is qualitatively different than previous or planned LBL experiments. It is sensitive to multiple physical effects. Allows one to break parameter degeneracies. Rough measurement of δ CP and θ 13 with only ν-running. If we get lucky with the true parameter values, these statements can be made more strongly. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 4 / 45

5 The Concept Motivations for Basic Parameters Why Long, Wide and High? 1 Wide see multiple oscillations High avoid Fermi motion 3 High Long Blue bars show flux coverage (full width at 1% peak) Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 5 / 45

6 The Concept Natural Background Reduction Q for E ν = 1-1 GeV Natural Benefits to Going High Energy Arbitrary (max bin = 1.) Arbitrary (max bin = 1.) Nuance MC ν n ν n π Q (GeV^) E π for E ν = 1-1 GeV Nuance MC ν n ν n π Putting some ν µ ν e signal at high energy, gain strong suppression of NC bkg via Q kinematic cutoff Nuance MC Q (top) and E π (bottom) for single-π NC events Each color band: mono-energetic neutrinos, 1-1 GeV in 1 GeV steps. At E ν > GeV get > 5 natural background reduction E π (GeV) Brett Viren, 3/3/18 Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 6 / 45

7 Oscillation Probability Illustrations 1 The Concept Oscillation Probability Illustrations Disappearance Appearance 3 Example: BNL to Homestake 4 Making Very Long Baseline Neutrinos at BNL Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 7 / 45

8 Oscillation Probability Illustrations Oscillation Probability Illustrations Unless noted, the values used are: m1 = 8.e-5 ev. KamLAND/SK/SNO. m3 =.5e-3 ev. Super-Kamiokande. sin (θ 3 ) = 1.. SK atmospheric maximal mixing. sin (θ 1 ) =.86. SK/SNO solar mixing. sin (θ 13 ) =.4. CHOOZ limited: use not too big, not too small. δ CP =. Totally unknown. Baseline = 54 km (other, short baselines shown for comparison) Matter effects using PREM Earth density and electron fraction profile. Probability plots: Energy range will reflect full-width-1%-max coverage. Flash short baseline examples to compare. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 8 / 45

9 Oscillation Probability Illustrations Disappearance Multiple Disappearance Oscillations Well Covered! P(µ,µ) at 54 km m 3 m 3 =.5e-3 =.8e-3 Multiple wiggles! No problem resolving the dip, will see more than one! Nodes well placed across flux coverage, robust against m change..1 Brett Viren, 5/6/ m 3 =.3e E (GeV) ν Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 9 / 45

10 Oscillation Probability Illustrations Disappearance Disappearance at Short Baselines At most one dip to see. Can slip away if m ends up low. P(µ,µ) at 3 km P(µ,µ) at 8 km m 3 =.3e-3 m 3 =.5e m 3 =.3e-3 m 3 =.5e-3.1 m =.8e E (GeV) ν Brett Viren, 5/6/.1 =.8e E (GeV) ν Brett Viren, 5/6/ m 3 Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 1 / 45

11 Oscillation Probability Illustrations Rich Appearance Effects Appearance The Very Long Baseline approach makes available a rich set of appearance effects. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 11 / 45

12 Oscillation Probability Illustrations Rich Appearance Effects Appearance The Very Long Baseline approach makes available a rich set of appearance effects. Multiple Appearance Peaks Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 11 / 45

13 Oscillation Probability Illustrations Appearance Multiple Appearance Peaks Covered P(µ,e) at 54 km m 3 =.3e-3 m m 3 3 =.5e-3 =.8e-3 Multiple peaks covered Robust against m change.1 Brett Viren, 5/6/ E (GeV) ν Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 1 / 45

14 Oscillation Probability Illustrations Appearance at Short Baselines Appearance P(µ,e) at 3 km.6 P(µ,e) at 8 km m 3 =.5e-3 m 3 m 3 =.3e-3 =.8e m 3 =.5e-3 m 3 =.3e-3 m 3 =.8e E (GeV) ν Brett Viren, 5/6/ E (GeV) ν Brett Viren, 5/6/ Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 13 / 45

15 Oscillation Probability Illustrations Appearance Rich Appearance Effects The Very Long Baseline approach makes available a rich set of appearance effects. Multiple Appearance Peaks Matter Effects Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 14 / 45

16 Oscillation Probability Illustrations Appearance Matter Effects at Very Long Baselines P(µ,e) at 54 km.6.5 m >.4.3. Vaccum Sensitivity to sign( m ) Factor of effect easy Effect mostly in first peak E ν > GeV.1 m < Brett Viren, 5/6/ E (GeV) ν Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 15 / 45

17 Oscillation Probability Illustrations Appearance Matter Effects at Short Baselines P(µ,e) at 3 km.6 P(µ,e) at 8 km m >.3 m >. Vaccum. Vaccum.1 m <.1 m < E (GeV) ν Brett Viren, 5/6/ E (GeV) ν Brett Viren, 5/6/ Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 16 / 45

18 Oscillation Probability Illustrations Appearance Rich Appearance Effects The Very Long Baseline approach makes available a rich set of appearance effects. Multiple Appearance Peaks Matter Effects CP Violation Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 17 / 45

19 Oscillation Probability Illustrations Appearance Nonzero CP Angle at Very Long Baselines P(µ,e) at 54 km Brett Viren, 5/6/ 1 CP = Deg CP = 45 deg CP = 9 deg CP = 18 deg log (E) (log (GeV)) 1 ν 1 Shift in magnitude and location Stronger effect for oscil peak n > 1 (W.Marciano) Longer baseline pulls n > 1 away from Fermi-motion region. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 18 / 45

20 Oscillation Probability Illustrations Appearance Nonzero CP Angle at Short Baselines P(µ,e) at 3 km CP = Deg CP = 45 deg CP = 9 deg CP = 18 deg P(µ,e) at 8 km CP = Deg CP = 45 deg CP = 9 deg CP = 18 deg Brett Viren, 5/6/ log 1 1 (E) (log (GeV)) ν 1 Brett Viren, 5/6/ log 1 (E) (log (GeV)) ν 1 Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 19 / 45

21 Oscillation Probability Illustrations Appearance Rich Appearance Effects The Very Long Baseline approach makes available a rich set of appearance effects. Multiple Appearance Peaks Matter Effects CP Violation ν e Appearance with θ 13 =! Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 / 45

22 Oscillation Probability Illustrations Guaranteed ν e Appearance Appearance P(µ,e) at 54 km Brett Viren, 5/6/ Effect of m sol sin (θ 3 ) =.9 sin (θ 13 ) =.4 θ 13 = E (GeV) ν What if θ 13 =? Still observe ν e via m 1! θ 13 = in black θ 13 in blue Upswing at low energies is due to m 1. Small shift if non-maximal mixing This over constrains the solar parameters - potential to see new physics! Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 1 / 45

23 Oscillation Probability Illustrations θ 13 = at Short Baselines Appearance P(µ,e) at 3 km.6 P(µ,e) at 8 km sin (θ 13 ) =.4.3 sin (θ 13 ) = = θ 13.1 = θ E (GeV) ν Brett Viren, 5/6/ E (GeV) ν Brett Viren, 5/6/ Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 / 45

24 Oscillation Probability Illustrations Appearance Rich Appearance Effects The Very Long Baseline approach makes available a rich set of appearance effects. Multiple Appearance Peaks Matter Effects CP Violation ν e Appearance with θ 13 =! Since these effects occur differently across the spectrum in a Very Long Baseline experiment, they can be disentangled with a single experiment. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 3 / 45

25 Oscillation Probability Illustrations Appearance Sensitivity to Different Parameters in Different Energy Regions E ν < 1 GeV 1 < E ν <GeV E ν > GeV sin θ 13 sign( m3 ) - - δ CP solar - It s a complex picture with many effects! But, effects have different strength at different energies. Measuring across the wide energy band makes it possible to sort them out. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 4 / 45

26 Example: BNL to Homestake 1 The Concept Oscillation Probability Illustrations 3 Example: BNL to Homestake Event Rate Disappearance Appearance 4 Making Very Long Baseline Neutrinos at BNL Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 5 / 45

27 Example: BNL to Homestake Example Baseline: 54 km Homestake, SD Henderson, CO 54 km 77 km BNL Homestake & Henderson equivalent. Assume UNO class far detector. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 6 / 45

28 Example: BNL to Homestake Neutrino Flux The following is work by Milind Diwan. BNL Wide Band. Proton Energy = 8 GeV nu/gev/m /POT at 1 km Distance = 1 km ν µ ν e νe /ν µ =.7 ν running: 1 MW, 8 GeV proton beam seconds PoT 6 cm carbon target 4 m φ m long decay tunnel ν running same but MW E ν (GeV) Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 7 / 45

29 Example: BNL to Homestake Expected Number of Events Event Rate 1 MW ν running Water-Cherenkov detector 5 kton water fiducial 54 km baseline seconds exposure Reaction Number CC ν µ + N µ + X 518 NC ν µ + N ν µ + X 1698 CC ν e + N e + X 38 QE ν µ + n µ + p QE ν e + n e + p 84 CC ν µ + N µ + π + + N NC ν µ + N ν µ + N + π 3178 NC ν µ + O 16 ν µ + O 16 + π 574 CC ν τ + N τ + X 319 (if all ν µ ν τ ) Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 8 / 45

30 Example: BNL to Homestake Disappearance Disappearance at 54 km Events/bin 5 ν µ DISAPPEARANCE BNL-HS 54 km sin θ 3 = 1. 1% energy resolution + Fermi motion nodes clearly visible 15 1 m 3 = e-3 ev 1 MW,.5 MT, 5e7 sec No oscillations: 139 evts With oscillations: 5936 evts Background: 1111 evts Black and red include background Only single-π production bkg considered here Reconstructed ν Energy (GeV) σ( m 3 ) & σ(sin (θ 3 )) 1% Resolution dominated by systematics Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 9 / 45

31 Example: BNL to Homestake Appearance Expected Appearance Results 1 Weighted SK Atm-ν MC/reconstruction study (C. Yanagisawa) Baseline requirements study (M. Diwan) 3 In progress: full study with UNO detector MC and realistic reconstruction code. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 3 / 45

32 Example: BNL to Homestake Appearance C. Yanagisawa - weighted SK Atm-ν MC study Improved π finder + likelihood cut. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 31 / 45

33 Example: BNL to Homestake Appearance Appearance - Baseline Requirements Study M. Diwan, close but stricter requirements still: ν e APPEARANCE ν e APPEARANCE Events/bin 7 6 BNL-HS 54 km sin θ ij (1,3,13) =.86/1./.4 m ij (1,3) = 7.3e-5/.e-3 ev 1 MW,.5 MT, 5e7 sec Events/bin 6 5 BNL-HS 54 km sin θ ij (1,3,13) =.86/1./.4 m ij (1,3) = 7.3e-5/e-3 ev 1 MW,.5 MT, 5 yr 5 CP 135 o : 63 evts CP 45 o : 489 evts 4 Reversed Mass Ordering CP 135 o : 398 evts 4 3 CP -45 o : 318 evts Tot Backg.: 146 evts ν e Backg.: 7 evts 3 CP 45 o : 311 evts CP -45 o : 1 evts Backg.: 146 evts ν e Backg.: 7 evts Reconstructed ν Energy (GeV) Reconstructed ν Energy (GeV) Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 3 / 45

34 Example: BNL to Homestake Sensitivity of δ CP and θ 1 3 Appearance 1-σ and 9% contours, all other parameters fixed. Resolution δ CP vs Sin θ 13 Resolution δ CP vs Sin θ 13 δ CP (deg.) 15 1 STAT+SYS ν µ and Anti-ν µ Running δ CP (deg.) 15 1 STAT+SYS ν µ and Anti-ν µ Running 5 68, 9 % C.L. Normal Hierarchy 5 68, 9 % C.L. Reversed Hierarchy BNL-HS 54 km sin θ ij (1,3,13) =.86/1./.4, δ CP =45 o m ij (1,3) = 7.3e-5/e-3 ev.5 MT, 5e7 sec for each polarity 1 MW for ν µ, MW for Anti-ν µ Sin θ BNL-HS 54 km sin θ ij (1,3,13) =.86/1./.4, δ CP =45 o m ij (1,3) = 7.3e-5/e-3 ev.5 MT, 5e7 sec for each polarity 1 MW for ν µ, MW for Anti-ν µ Sin θ 13 Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 33 / 45

35 Example: BNL to Homestake Appearance Sensitivity of δ CP and θ Combined ν/ ν 1-σ contours, all other parameters fixed. Regular hierarchy νυ and Antiνυ running Reversed hierarchy νυ and Antiνυ running δ CP (deg.) 15 1 δ CP (deg.) Sin θ Sin θ 13 Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 34 / 45

36 Example: BNL to Homestake Appearance Patrick Huber, UWisc: Both ν- and ν-running. Includes correlations and errors on all parameters, 1% background uncertainty. Will improve as other oscillation parameter measurements improve. All δ CP range covered. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 35 / 45

37 Making Very Long Baseline Neutrinos at BNL 1 The Concept Oscillation Probability Illustrations 3 Example: BNL to Homestake 4 Making Very Long Baseline Neutrinos at BNL AGS Upgrade BNL Site Development Target and Horn Price Tag Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 36 / 45

38 Making Very Long Baseline Neutrinos at BNL Upgrade to AGS AGS Upgrade To RHIC High Intensity Source plus RFQ To Target Station MeV Drift Tube Linac BOOSTER MeV 4 MeV Superconducting Linacs 8 MeV 1. GeV AGS 1. GeV 8 GeV.4 s cycle time (.5 Hz). s. s Use 1. GeV SC Linac instead of booster ppp Fill time.6 sec 1. msec Increase rep. rate.5.5 Hz New mag. PS and RF cavities Further improvements on design being worked on. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 37 / 45

39 Making Very Long Baseline Neutrinos at BNL BNL Site Development BNL Site Development Point to most points west Keep hadrons above the water table Room for a (very) near detector Hill cheaper than tunnel Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 38 / 45

40 Making Very Long Baseline Neutrinos at BNL Target and Horn Conventional pulsed hadron focussing with horns Likely carbon-carbon target, embedded in 1st horn R&D underway with material experiments Collaboration with FNAL, JPARC and others Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 39 / 45

41 Making Very Long Baseline Neutrinos at BNL Estimated Costs Price Tag October 1, 4 BNL IR The AGS-Based Super Neutrino Beam Facility Conceptual Design Report Editor: W. T. Weng, M. Diwan, and D. Raparia Contributors and Participants J. Alessi, D. Barton, D. Beavis, S. Bellavia, I. Ben-Zvi, J. Brennan, M. Diwan, P. K. Feng, J. Gallardo, D. Gassner, R. Hahn, D. Hseuh, S.Kahn, H. Kirk, Y. Y. Lee, E. Lessard, D. Lowenstein, H. Ludewig, K. Mirabella, W. Marciano, I. Marneris, T. Nehring, C. Pearson, A. Pendzick, P. Pile, D. Raparia, T. Roser, A. Ruggiero, N. P. Samios, N. Simos, J. Sandberg, N. Tsoupas, J. Tuozzolo, B. Viren, J. Beebe-Wang, J. Wei, W. T. Weng, N. Williams, P. Yamin, K. C. Wu, A. Zaltsman, S. Y. Zhang, Wu Zhang Brookhaven National Laboratory Upton, NY October 1, 4 Put through internal BNL review Bottom-up estimation with WBS Based on RHIC & SNS ring, LHC magnets. Target & Horn from BNL, KK and NuMI. $73.4M FY4 6 Years to neutrinos 3 years R&D Construction after 1 year 4.5 years to completion.5 year commissioning Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 4 / 45

42 Making Very Long Baseline Neutrinos at BNL Price Tag Summary The BNL Very Long Baseline concept provides a qualitatively different experiment than past or proposed Precision (1%) measurement of atm-ν params, systematics limited. Degeneracy-broken measurement of appearance parameters If lucky, ν-running only needed for appearance measurements, with ν-running gain precision w/out needing luck. Appearance results limited by how well background can be controlled Work on full detector simulation and reconstruction to be finished before final word. Affordable, practical beam design. No magic needed. Ongoing collaborations (C. Yanagisawa (UNO), P. Huber (GLoBES)), others most welcome! Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 41 / 45

43 Making Very Long Baseline Neutrinos at BNL Price Tag References Papers: BNL Neutrino Working Group web page: Extra Long Baseline Neutrino Oscillations and CP Violation, W. Marciano, BNL-HET-1/31, hep-ph/ Whitepaper, BNL hep-ex/111. VLBL PRD paper, Phys.Rev. D68 (3) 1, hep-ph/3381. AGS Superbeam CDR, BNL IR (avail from NWG web). The Case for a Super Neutrino Beam, M. Diwan (19 km and 54 km compared) hep-ex/4747. Publicly available software: Oscillation Probability Calculator, B. Viren, bviren/elbo/libnuosc++/ GLoBES, P. Huber, M. Lindner and W. Winter, hep-ph/47333 Nuance, D. Casper, hep-ph/83 Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 4 / 45

44 Backup Slides Making Very Long Baseline Neutrinos at BNL Price Tag Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 43 / 45

45 Making Very Long Baseline Neutrinos at BNL Price Tag CP Violation Sensitivity Independent of Baseline P(ν µ ν e ) GeV neutrino, vacuum numu cp=9 numu cp=45 numu cp= anumu cp=45 anumu cp=9 SK best fit LMA-I sin (θ) 13 = Brett Viren, bv@bnl.gov, 6/4/4 Baseline (km) Marciano, hep-ph/18181 A L Flux 1/L Statistical F.O.M. = ( A/A) = A N/(1 A) N = Nνe + N νe Linear in flux Quadratic in A Independent of BL, at least until the flux totally runs out. Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 44 / 45

46 Making Very Long Baseline Neutrinos at BNL Price Tag Resolution on θ 3 and m 3 Test point for ν µ disapp m 3 ev Small oval is for BNL-HS 9 % C.L. contours Stat. + Syst. MINOS 9 % SuperK 9 % Energy scale systematic not included (SK has.5%) O.w. σ( m 3 ) 1% Not strongly sensitive on normalization uncertainty due to multi-nodal spectrum shape Sin θ 3 Brett Viren (Brookhaven National Lab) BNL VLBL NuFact 5 45 / 45

BNL Very Long Baseline Neutrino Oscillation Expt.

Mary Bishai, BNL 1 p.1/36 BNL Very Long Baseline Neutrino Oscillation Expt. Next Generation of Nucleon Decay and Neutrino Detectors 8/04/2005 Mary Bishai mbishai@bnl.gov Brookhaven National Lab. Mary Bishai,