1. Neutrino oscillation physics with a FNAL Proton Driver
NuFact 05, Frascati, Italy
June 23, 2005
Walter Winter
Institute for Advanced Study, Princeton
Editors: Steve Brice, Debbie Harris (FNAL), Walter Winter (IAS)

2. NuFact 05 - FNAL PD - W. Winter
Contents
Introduction
Performance indicators
Proton Driver (PD) physics scenarios
Some examples for PD experiments
PD and evolution of the global program
Summary
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

3. What is a Proton Driver? (a “theorist’s perspective”)
A terribly complicated machine to obtain more protons
Machine part of the FNAL proton driver:
LINAC + Main injector modifications (Giorgio Apollinari’s talk)
Goal: 8 GeV and (up to) 120 GeV protons at 2 MW
= Factor 5-10 past current FNAL proton source at 120 GeV
Cost: M$
Proton Driver = Higher Luminosities
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

4. Why do we need a Proton Driver?
Example: sin22q13 sensitivity limit at NOvA: Appearance rate prop. to sin22q13 (1st approximation) Sensitivity limit
E. g. Factor two better limit = factor four in luminosity
Thus: Much higher luminosities required to go further!
5 x Flux = Proton Driver
5 x Running time = 25 years (unrealistic)
5 x Detector mass = 150kt (150M$ x 5 = even more exp.)
L = Flux x Detector mass x Running time
“Recyclable” for other exps
Probably not re-usable
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

5. Fermilab PD study: Facts
Two parts: How to built a more powerful proton source (W.Foster) What to do with the protons (S. Geer)
“The Fermilab Director has requested a study […], which will require documentation sufficient to establish mission need as defined by the DOE CD-0 process.”
Study launched at PD Workshop at Fermilab in Oct. 2004
Director’s review in March 2005
Missing working group reports + CD0 documentation will probably be publically available soon …
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

6. Fermilab PD study - Structure
PD study (Physics part) CD0 documentation for DOE S. Geer
Neutrino oscillations S. Brice, D. Harris, W. Winter THIS TALK! Authors (preliminary): Antusch, Bazarko, Cooper, DeJongh, Diwan, Feldman, Finley, Geer, Huber, deGouvea, Jansson, Kersten, Lindner, Mena, Michael, Parke, Ratz, Rolinec, Schwetz, Stefanski, Van de Water
Neutrino scattering J. Morfin, R. Ransome, R. Tayloe Fundamental neutrino properties (X-section measurem. for osc. Experiments, neutrino-electron scattering)
Fundamental properties of matter
Broader physics program T. Bowles, H. Cheung, D. Christian, G. Greene, P. Kasper, M. Mandelkern, P. Ratoff, R. Ray, R. Roberts, H. Nguyen, T. Yamanaka Muon physics (EDM, muon g-2, rare decays); Kaon, Pion, Neutron, and Antiproton experiments
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

7. Potential users of the FNAL PD
NOvA
MiniBOONE
Super-NOvA (2nd detector)
Others?
Proton Driver (machine)
New beamline BB superbeam FNAL- DUSEL
“multi-GeV”
Neutrino factory
“some GeV”
New beamline “use all protons” BB + n x NBB
b-Beam
120 GeV
8 GeV
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

8. Why is the neutrino oscillation part so challenging?
Many potential experiments using the proton driver
Which of these experiments should one build
Depends on physics case!
Key questions:
Does one need the FNAL proton driver in any physics case?
What does one do with it then?
Should be answered by neutrino oscillation study
PD Study has a quite “strategical” character
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

9. Pre-Proton Driver program
PD Timescale: Starts operation ~ 2014?
Ten more years to go until then
What happens until then (physics-wise)?
Pre-Proton Driver: ~ 2005 – 2014 US program: European/Japanese program: MiniBOONE KEK MINOS CNGS NOvA T2K … …
(Chapter 3, Neutrino Oscillation Document = NOD)
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

10. Proton Driver Physics cases
Main scenarios Deal with following questions: q13 discovered or not? How large is q13?
How measure the mass hierarchy and CP violation?
Special cases
LSND confirmed
q23 still consistent with maximal mixing
Something unexpected happens
(For details, see working group report)
(Chapters 4-6, NOD)
(Chapter 7, NOD)
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

11. Predictions for future experiments (or: How to simulate PD-based experiments)
Existing experiments:
Future experiments:
? But: only one set realized by nature!
Simulated data
Fit parameters to data: Precision of quantity of interest
Input parameters
Data
Fit parameters to data: Precision of quantity of interest
Simulation of future experiments = “Hypothesis testing”
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

12. Simulated versus fit parameters
Simulated/true params
Represent the values implemented by nature
Known within current limits
Change the event rates, top.
Have to be interpreted like “If the value of … is …, then the performance will be …” - Luck or not luck?
Used for risk minimization!
Determine the precision of the quantity of interest
“Unused” parameteres are usually marginalized over (projection onto axis/plane of interest)
Source of correlations!
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

13. q13 performance indicators
q13 exclusion limit (sensitivity limit):
Describes the new q13 limit for the hyopthesis of no signal (q13=0)
Correspond to new limit after the experiment has been (unsuccessfully) performed
Define as largest fit value of q13=0, which fits true q13=0
Straightfoward inclusion of correlations and degeneracies
Does not depend on the simulated dCP and mass hierarchy!
q13 discovery reach:
Describes if q13=0 can be excluded for the hyopthesis of a certain set of parameters (q13>0)
Almost no correlations+degeneracies (since fit q13=0)
Depends on simulated q13, mass hierarchy, and dCP
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

14. Discovery plots and “Fraction of dCP”
Read: For sin22q13=0.04, we expect a discovery for 20% of all values of dCP
Sensitive region as function of true q13 and dCP
New primer in PD-NOD!
dCP values now stacked for each q13
Fraction of dCP for successful discovery
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

15. “Fraction of dCP” = Measure for luck?
Remember: Only one set of simulated values actually realized by nature!
For uniform distribution in dCP: Fraction of dCP = Probability to discover dCP
dCP comes from a complex phase factor eid in the mixing matrix Thus: Distribution in sin d theoretically “unnatural”!?
No “luck” needed; works for all d, hier.
“Typical” d; chance
Best case d, hierarchy
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

16. Main PD scenarios … ... from q13 exclusion limit at PD startup
Need substantially more than existing beamline + detector But: superbeams way to go
Conceptual cases in PD study
Probably need neutrino factory
Could work on CP violation+ mass hierarchy with existing beamline + det.
NUE=“NuMI Up- graded Experiment” ~ NOvA
Scenario 3: Discovery unlikely until PD startup
Scenario 1: Certainly discovered until PD startup
Scenario 2: Discovery likely before PD startup
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

17. Scenario 1: sin22q13 larger than ~0.04
CP fraction “1” not possible!
NUE ~ NOvA
T2K*: T2K+Japenese PD
T2HK: T2K+Japanese PD+ Hyper-K-Detector
Excellent results with FNAL-PD + possibly existing equipment (NOvA)
Synergy with Japanese program (especially for mass hierarchy)
If Japanese program not funded: Other options (next slides)
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

18. Scenario 2: 0.01 < sin22q13 < 0.04
NUE ~ NOvA
2nd NUE ~ Super-NOvA
Second NUMI off-axis detector promising alternative
New long baseline experiment could “do it all”
What specific option preferred needs further study …
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

19. Example 1: BB FNAL-Homestake
Two-Peak Structure still visible at 1290km
FNAL-Homestake very competitive for q13 and also dCP
See hep-ex/ for more details
2540km
1290km
Figures from M.Diwan
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

20. NuFact 05 - FNAL PD - W. Winter
Example 2: 2 MW + 2 MW beam
~8m
30 mR maximum off axis
D.Michael,
C.White,
M.Messier
1290km
Uses all the
protons from
Booster
Main Injector
~200M
150kT Liquid Argon
~4m
Consider LAr and H2OC
90% CL
99% CL
8 GeV
protons
120 GeV
protons
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

21. Scenario 3: sin22q13 smaller than ~0.01
NuFact: 50kt magn. Iron det. 4+4 year running 4 MW target power (~1021 useful muon decays/year) 50 GeV neutrinos
For large sin22q13>0.003: New long baseline experiment interesting alternative
For smaller values: Neutrino factory (or b-Beam) can do measurements in large fraction of parameter space (see P. Huber’s talk)
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

22. Example: b-Beam at Fermilab?
“Stretched Tevatron“ aimed
at Soudan
Total circumference:
approximately 2 x Tevatron
320m 58 mrad
A. Jansson
(Huber, Lindner, Rolinec, WW, to appear)
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

23. Evolution of q13 discovery limit
Characterize dependence on dCP as bands reflecting all possible chases
Choose starting times + luminosities as close as possible to values in respective LOIs/proposals
All exps: five years running
New generation will quickly determine potential
Reactor experiments provide complementary information!
For inverted hierarchy: Beam limits shift somewhat down!
(from: FNAL proton driver study, to appear)
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

24. Examples for q13 cases (1) Assume: Actual value of sin22q13 = 0.03
~ : q13 signal likely at superbeams or reactor experiments
PD+NUE+2nd NUE very competitive Fast+cost efficient alternative to T2HK!?
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

25. Examples for q13 cases (2) Assume: Actual value of sin22q13 = 0.007
Discovery of q13 unlikely without PD and impossible for T2K
If no PD at Fermilab, probably no further superbeam program!
But: One could have done almost of all the physics with a superbeam program!
June 23, 2005
NuFact 05 - FNAL PD - W. Winter

26. NuFact 05 - FNAL PD - W. Winter
Summary
We need the FNAL proton driver in all physics cases (but not all possible discussed experiments!)
The FNAL Proton Driver is the next logical step in the FNAL neutrino oscillation program beyond NOvA
The FNAL-based neutrino oscillation program is unique, because
There are already existing experiments. Thus: No complete program from scratch
NUMI beamline is so far the only beamline with large enough matter effects to potentially resolve the mass hierarchy
Synergy with reactor program + Japanese program
For more: see
June 23, 2005
NuFact 05 - FNAL PD - W. Winter