Doug Michael, Caltech

1
A Fermilab to Homestake Experiment?

• Doug Michael, Caltech
• With Chris Smith (Caltech) and help from Mark
  Messier (Indiana)
• Mar. 3, 2004

2
Setting the Scene

• The Fermilab NuMI beam will start operating later
  this year.
• Initial operation will be at about 0.2 MW (about
  2e20 protons per year)
• The beamline is designed for 0.4 MW of 120 GeV
  protons and can probably sustain a factor of 2
  beyond that without substantial reworking.
• Up to 2 MW may be possible with the main issue
  being repetition rate and total heat in the
  target, decay pipe and dump.
• MINOS will be the first experiment to use this
  beamline.
• Will provide the best oscillation measurements
  yet But we want more!
• The Off Axis Experiment is being proposed to
  offer yet further improvement in sensitivity to
  ne appearance.
• 50 kT fine-grained calorimeter, 15 mR off axis.
• Lower energy, narrow-band beam than on-axis.
  Better rejection of NC p0.
• We will hear from Gary Feldman on this on Friday.
• Sensitive to ne appearance including matter
  effects and dCP.
• Ambiguities come from the sensitivity to the
  different terms and also from uncertainty in sin2
  q23.
• A new 8 GeV proton driver will be the central
  element of recommendations from the Future of
  Fermilab Committee. Neutrino physics is the
  initiating flagship enterprise for this machine.
• This will increase the proton intensity at 120
  GeV by a factor of 5-10.
• Reduces, but does not eliminate ambiguities for
  all parameters.
• Measurements at the second oscillation maximum
  can remove the ambiguities. This could be done
  building yet another off axis detector. But
  perhaps on this timescale it would make more
  sense to connect this effort to a new laboratory
  at Homestake (or elsewhere).

3
Measurement of Oscillations in MINOS
For Dm2 0.0020 eV2, sin2 2q 1.0
Plots on the left Oscillated/unoscillated ratio
of number of nm CC events in the far detector
vs Eobserved
Plots on the right MINOS 90 and 99 CL allowed
oscillation parameter space for the Super-K best
fit point.
4
Appearance of Electrons
Neutrinos at higher energy provide most of the
events.
Peak of oscillation prob for this Dm2
For Dm2 0.0025 eV2
For Dm2 0.0025 eV2, sin2 2q13 0.067
3 s discovery potential for three different
levels of protons on target and versus systematic
uncertainty on the background.
Observed number of events identified as coming
from ne CC interactions with and without
oscillations. 25x1020 protons on target.
5
Off Axis A possible next step

• The neutrino beam isnt limited
• just to those on axis. It is a
• broad distribution.
• Total flux goes down as one
• moves off axis, but the flux of
• low energy neutrinos can be
• greater than on axis due to pion
• decay kinematics.

Sensitivity goal for the Off Axis Experiment
Based on 20e20 P.O.T.
Note, this is the NuMI medium energy beam
Neutrino Events
The goal for the Off Axis Experiment is a factor
of 20 more sensitivity than MINOS 10 x bigger
detector, finer sampling, low Z, low energy beam,
more protons.
6
Comparison of Long Baseline and Reactor Exp'ts
Phase I
Phase I
The size of the correlation and degeneracy
bars represent sensitivity to additional physics!
Huber, Lindner, Winter
7
Assumptions for beyond Off Axis

• I dont see this as an alternative proposal to
  starting with an Off Axis experiment at Fermilab.
  Rather, it is a longer term way of building on
  investments and offering the best possible
  physics measurements.
• We assume that the following resources will be
  available
• A new 2 MW 8 GeV proton source at Fermilab
• Corresponding 2 MW of 120 GeV protons from the
  Main Injector
• A new beamline, with whatever characteristics we
  need aimed in the direction of Lead.
• An underground laboratory at Homestake (1290 km
  from Fermilab).
• A 500 kT water Cerenkov detector at Homestake
  with Super-K type capabilities. (Alternatively, a
  100 kT fine-grained calorimeter, liquid argon or
  scintillator.)
• It is clear that we are talking about an
  experiment that will start later than 2015.
• Other paths/approaches can be imagined, but this
  gives a first idea of what might be possible.

8
A Brief Side-Bar
An example 700 MeV ne CC event in a 4cm pitch
totally active scintillator detector
100 kT detector for 400M?
Courtesy Leon Mualem
9
A new proton driver at Fermilab?
Neutrino Super- Beams
SY-120 Fixed-Target
NUMI
Off- Axis
X-RAY FEL LAB
Anti- Proton
8 GeV
2 MW 8 GeV Linac
To Homestake 120 GeV and 8 Gev
Main Injector _at_2 MW
700m Active Length
Slow-Pulse Spallation Source Neutrino Target
Also includes upgrades to Main Injector for
current and cycle
10
Appearance Probabilities at 1290 km
11
The FeHo Beamline
(FErmilab to HOmestake)
8m

• A new neutrino beamline using the same 120 GeV
  extraction line as the current NuMI beam but then
  redirecting towards Homestake.
• In addition, a beamline to deliver 8 GeV protons
  from the (assumed) new proton driver.
• Dual targeting/focussing stations
• Decay region 200 m in length, 4 m tall and 8 m
  wide. (Roughly the same scale as NuMI
  construction.)
• Tunable Off Axis beams for 120 GeV protons. On
  axis beam for 8 GeV protons.
• 2 MW of proton power at both 120 GeV and 8 GeV.

30 mR maximum off axis
200M
4m
120 GeV protons
8 GeV protons
12
Unoscillated Rate of CC Events at 1290 km
The off axis beams Are NuMI beams.
The 8 GeV beam shown is exactly the
Mini-BooNE beam with more protons. Probably a
more optimal beam will be possible.
13
"Optimized" Unoscillated Rate of CC Events at
1290 km
Note that use of several narrow band beams
to form a broad energy envelope will permit
better background rejection than a single
wide-band beam
14
A better look at the HE Tails
15
Rate of QE Events Only at 1290 km
16
Oscillated rate of neutrino events at 1290 km
Unoscillated
Remaining nm
Appearing ne
17
Oscillated rate of neutrino events at 1290 km
Remaining nm
Unoscillated
Appearing ne
18
Oscillated rate of neutrino events at 1290 km
Unoscillated
Appearing ne
19
What the statistics permit with an impossible
detector

• Using all ne CC events with no
• background.
• In other words, it only gets
• worse than this.

20
What the statistics permit with a possible
detector?
Fit done using only QE events with no background.
Perhaps not so far from what might be expected
for 100 kT liquid argon or scintillator?
21
Approximating the Response in Water Cerenkov
Ala Super-K courtesy of Mark Messier
22
Measured spectrum and fit for single ring,
mu-like events
Statistics are based only on QE events but
backgrounds come from all events
23
Measured spectrum and fit for single ring, e-like
events
Statistics are based only on QE events but
backgrounds come from all events
24
Individual Reconstructed Spectra in Water Cerenkov
120 GeV, 30mR OA
8 GeV on axis
120 GeV, q0mR OA
120 GeV, 15mR OA
Statistics are based only on QE events but
backgrounds come from all events
25
Conclusions

• Use of combined 8 GeV and 120 GeV proton beams
  with a future Fermilab proton driver permits a
  very intense, wide/narrow band beam with tunable
  flux over two orders of magnitude in neutrino
  energy.
• Assuming a large detector at Homestake, an
  experiment with very high statistics will be
  possible. Even for a tracking calorimeter
  detector which may be better suited than water
  Cerenkov to the event reconstruction needs. We
  need to look at this in more detail.
• Further assuming the detector to be 500 kT water
  Cerenkov with Super-K-like features permits
• Very small errors on nm disappearance parameters.
  Systematics (not addressed here) may well
  dominate.
• OK errors on ne appearance parameters but
  disappointing compared to the statistics due to
  energy smearing and backgrounds.
• Hard to get information out of the second
  oscillation ne appearance maximum. Needs more
  work.