Bari, Bologna, Boston, Caltech, Drexel, Indiana, Frascati, Gran Sasso,L’Aquila, Lecce, Michigan, Napoli, Pisa, Roma I, Texas, Torino

with

Summary• Analysis of High and Low energy neutrino events• Sterile vs Tau neutrino oscillations (matter effects)

A. Surdo INFN - Lecce

MACRO @ Gran Sasso

• GUT Monopole and rare particle search• Cosmic ray physics• Neutrino bursts from stellar collapses• Study of atmospheric neutrinos• Neutrino astronomy• Indirect search for WIMPs

Data taking: Mar 1989 – Dec 2000(complete detector since Apr ’94)

Main features of Macro as detector

• Large acceptance (~10000 m2sr for an isotropic flux)

• Low downgoing rate (~10-6 of the surface rate )

• ~600 tons of liquid scintillator to measure T.O.F. (time resolution ~500psec)

• ~20000 m2 of streamer tubes (3cm cells) for tracking (angular resolution < 1° )

More details in Nucl. Inst. and Meth. A324 (1993) 337.

Up stop

In down

In up

Up through

Neutrino event topologies in MACRO

• Detector mass ~ 5.3 kton

(1) Up throughgoing (ToF)(2) Internal Upgoing (ToF)(3) Internal Downgoing (no ToF)(4) UpGoing Stopping (no ToF)

Absorber

Streamer

Scintillator

(1) (2)(3)

(4)

Energy spectra of events

detected in MACRO

• Emedian ~ 50 GeV for Throughgoing muons;

• Emedian ~ 4.5 GeV for Internal Upgoing (IU) ;

• Emedian ~ 3.5 GeV for Internal Downgoing (ID)

and for UpGoing Stopping (UGS) ;

Low energy events (IU, ID+UGS) allow to investigate the oscillation parameter space independently from throughgoing muons

1

T1 T2 T3 T4 c2L

+1

-1 T1

T3 T4

T2

L>2.5mStreamer tube track

Upward Throughgoing muons

Analysis and Data Set

* evaluation:

1) agreement between streamer track and scintillator counter hits (position from times at the ends)

2) -1.25 < 1/ < -0.75

3) 2m (~200 g/cm2) of absorber crossed to reduce at 1% level the bck due to ’s produced by muons (Astrop. Phys. 9 (1998) 105)

• Automated data selection procedure (no visual scan)

• 863 induced upward through-going muons during:

* Background rejection cuts:

Event selection based on time-of-flight method

Mar 89-Nov 91 (1.4y) Dec 92-Jun 93 (0.4y) Apr 94-Dec 2000 (5.5y)

1/6 lower detector Lower detector Complete detector

Phys. Lett. B357 (1995) 481 Phys. Lett. B434 (1998) 451 (until Nov ‘98)

Measured flux and comparison with Monte Carlo

• Data

number of events: 863 background (wrong ) 22.5 background (’s from muons) 14.2 Internal neutrino interactions 17

Total 809 events

• Monte Carlo prediction 1122 ± 17%

- Bartol atmospheric neutrino Flux (±14%)

- GRV-LO-94 PDF for DIS cross section (± 9%)

- Lohmann et al. muon energy loss (± 5%)

* Detector response simulation using GEANT

* Same analysis chain as the data

R = data/prediction = 0.721 ± 0.026stat ± 0.043syst ± 0.122theor

2 test on the angular distribution (10 bins) with prediction normalized to data :

• 2 = 9.6/9 d.o.f for with maximum mixing and m2 ~ 0.0025 eV2 P = 37 %

• 2 = 25.9/9 d.o.f. for no - oscillations P = 0.2 %

Upward Through-Going MuonsAngular Distribution

MACRO UPMUProbabilities for sterile

neutrino oscillations

P.Lipari and M.Lusignoli, hep-ph/ 9803440Q.Y.Liu et al, PLB440 (1998) 319,

• Sterile neutrino matter effectsreduction of angular distortion (max mixing)

Peak probability for the shape = 1.8%Peak probability for the combination = 8%

Probabilities lower than that for neutrinos

Ratio vertical / horizontalLipari -Lusignoli, Ph Rev D57 (1998)

Monte Carlo optimized:

)4.0)(cos(

)7.0)(cos(

N

NR

• The plot is for maximum mixing P(sterile) = 0.033% P() = 8.4%

• Sterile neutrino disfavored with respect to at >99% C.L. for any mixing (7% systematic on the ratio)

(Results in hep-ex/0106049, to be published on Phys. Lett.,)

(a) time-of-fligth between central / top SC layers

(b) topological criteria for vertex containment inside lower detector (~1 % bck after this cut)

Internal Upgoing ’s (IU)

Phys. Lett. B478 (2000) 5

Partially contained events

Selection criteria:

Internal Downgoing ’s (ID)+ Upgoing Stopping ’s (UGS)

(a) No T.o.F. measurement(b) topological constraints on track (bottom SC layer + track inside fiducial volume)(c) visual scan procedure for final selection(d) >100 g cm-2 of material crossed in the detector

UGS ID

IU events

From MC simulation:

• E ~ 5 GeV• ~ 87% due to -CC interactions

DATA: 5.5 live-y

DATA - Bck(1/) = 154 events

• From Apr. 1994 up to Dec. 2000• 1/ distribution after all analysis cuts:

Data vs Monte Carlo Predictions

• : Bartol flux with geomagnetic cutoffs

(error ~ 13%)• = Q.E. + 1 (Lipari et al., PRL74 (1995) 4384)

+ DIS (GRV-LO-94 PDF) (error ~ 15%)• (E,zenith): detector response and acceptance

(systematic error ~ 10 %)

DATA: 154 12stat

MC: 285 28sys 57theo

MC-oscill: 168 17sys 34theo

DATA: 262 16stat

MC: 376 38sys 76theo

MC-oscill: 284 28sys 57theo

Internal down + Upgoing stopping ’s

Internal Upward going ’s

• GEANT based program for the simulation of detector response • Simulated events processed through the same analysis chain as the data

E ,zenith

Angular Distributions

thsysstatIUMC

Data11.005.004.054.0

)(

thsysstatUGSIDMC

Data14.007.004.070.0

)(

maximal mixing , m2 = 2.5 x 10-3 eV 2

MC expectation (no oscillations)

• Data consistent with a constant deficit in all zenith angle bins (IU: 2 /d.o.f. = 3.1/4 on shape)

• Probability for NO oscillations (one side): RIU: ~ 2.3% R(ID+UGS) ~ 9.3%

InternalUpgoing

Ratio of event types at low energy

• Most of the theor. uncertainties canceled (<5%)• Systematic errors reduced (~6%)

Data: R = 0.59 ± 0.06stat

Expected (No oscillations): R = 0.76 ± 0.04sys ± 0.04th

>>> compatibility ~ 1.9%

Expected oscillations: R = 0.59 ± 0.04sys ± 0.03th

(maximal mixing and m2 = 2.5 x 10-3 eV2 )

InternalDowngoing

UpgoingStop

R =

+