Main Slides

SK-I/III combined final data sample

Cherenkov angle distribution degrees

νe CCμ/π NC elastic

low region(μ/π)

isotropic region(NC elastic)

signal region

MC (decay electron channel not shown)

low region(μ/π)

signal region isotropic region(NC elastic)

e e+p

n (invisible)

Signal region

42o

μ, π

Low angle events

25-45o

Isotropic region

N

reconstructedangle near 90o

20-38 degrees 38-50 degrees 78-90 degreesE (MeV)

SK-I/IIIdataνμ CCνe CCNC elasticμ/π > C. thr.all backgroundrelic

Combined Fit

combined 90% c.l.: < 5.1 ev / yr / 22.5 ktons interacting < 2.7 /cm2/s (>16 MeV) < 1.9 /cm2/s (scaled to >18 MeV)

combined90% c.l.

ev/yr interacting in 22.5 ktons

logL

ikel

ihoo

d

SK-I/II/IIIcombinedlikelihood

Comparison to Published /cm2/s >18 MeV

Published limit 1.2

cross section update to Strumia-Vissani

1.2 1.4

Gaussian statistics Poissonian statistics in fit

1.4 1.9

New SK-I Analysis:ETHRESH 18 16 MeVε = 52% 78 %(small statistical correlation in samples)improved fitting method takes into account NC

1.91.6

New SK-I/II/III combined fit 1.6 1.9

Secondary Slides

Final Backgrounds (after all relic cuts)

E (MeV)

These 3 can be modeled as a combination of other backgrounds, and thus aren’t considered separately

CC Backgroundsinvisible μ decay eνe CCμ > C. threshold μ/π

νμ CCνe CCμ/πNC elastic

Combined Fit

combined 90% c.l.: = 5.1 ev / yr / 22.5 ktons interacting = 2.7 /cm2/s (>16 MeV) = 1.9 /cm2/s (scaled to >18 MeV)

combined90% c.l.

ev/yr interacting in 22.5 ktons

logL

ikel

ihoo

d

SK-I/II/IIIcombinedlikelihood

SK-I (~1500 days)SK-II (~790 days)SK-III (~550 days)combined

COMPARISON TO PUBLISHED LIMIT /cm2/s >18 MeV

Published limit 1.2

cross section update to Strumia-Vissani 1.2 1.4

Gaussian statistics Poissonian statistics in fit 1.4 1.9

New SK-I Analysis:ETHRESH 18 16 MeVε = 52% 78 %(small statistical correlation in samples)improved fitting method takes into account NC

1.91.6

New SK-I/II/III combined fit 1.6 1.9

SK-I

best fit is negativefit shown is 0 relic contribution

20-38 degrees 38-50 degrees 78-90 degrees

E (MeV)

SK-Idataνμ CCνe CCNC elasticμ/π > C. thr.all backgroundrelicSK-I only 90% c.l. limit:

< 2.4 /cm2/s (>16 MeV)< 1.6 /cm2/s (scaled to >18 MeV)

SK-IIBest fit (shown):3.5 ev/yr interacting

20-38 degrees 38-50 degrees 78-90 degrees

E (MeV)

SK-II only 90% c.l. limit:< 7.4 /cm2/s (>16 MeV)< 5.2 /cm2/s (scaled to >18 MeV)

dataνμ CCνe CCNC elasticμ/π > C. thr.all backgroundrelic

SK-IIIBest fit (shown) :6.5 ev/yr interacting

20-38 degrees 38-50 degrees 78-90 degreesE (MeV)

SK-III only 90% c.l. limit:< 8.1 /cm2/s (>16 MeV)< 5.7 /cm2/s (scaled to >18 MeV)

dataνμ CCνe CCNC elasticμ/π > C. thr.all backgroundrelic

Systematics: Inefficiency

12

• Define:– r = # relic events we see in data– R = # relic events actually occurring in detector– ε = efficiency (SK-I/II/III dependent)– assume ε follows a probability distribution P(ε)– assume P(ε) is shaped like Gaussian w/ width

σineff

– then we alter likelihood:

then the 90% c.l. limit R90 is such that

dPRLRL )()()('1

0

σineff

SK-I: 3.5%SK-II: 4.5% SK-III: 3.1%

Cuts: efficiencies and sys errors

13

• SK-I:effwall : 98% (0.5%)C. angle: 95% (0.4%)pion like: 98% (0.2%)spall+solar: 88% (1%)2-peak, 2-ring: >99%Correlation cut: 99% (0.3%)1st reduction: 99% (2%)

(includes: electronic noise cuts, 50 us cut)

Total: 78 %

• SK-II95% (0.3%)88% (0.3%)97% (0.5%)87% (1.4%)>99%99% (0.3%)99% (2%)

• SK-III96% (0.3%)94% (0.3%)98% (0.5%)89% (1%)>99%99% (0.3%)99% (2%)

69% 77%