Oscillation Analysis from Spectrum and Zenith Angle Variation
Neutrino oscillations may cause three effects in Super-Kamiokande:
a distortion of the energy spectrum, a zenith-angle dependend flux
(day-night effect) and a
seasonal dependence
of the flux (seasonal
variation).
Super-Kamiokande has looked for
distortions of the
spectrum and
zenith-angle variations of the
flux.
The largest sensitivity to MSW oscillation effects is achieved by
combining the zenith-angle flux variation with the spectral distortion.
Super-Kamiokande realizes this combination by fitting oscillation
predictions to seven spectra day and a night spectrum.
Definition of Zenith Angles:
The matter density inside the earth shows a sharp increase at about
half the earth's radius (the core of the earth):
Unfortunately, Super-K's geographical location prevents collecting
solar neutrinos that pass through the earth's inner core. We divide
the data into 'Day' (neutrinos do not pass through the earth), 'Mantle'
(neutrinos pass through the earth, but not through the core) and 'Core'
(neutrinos pass through the core) according to the zenith angle of the
sun. The Mantle data set is subdivided into 5 regions as shown above
(Man 1-5). Between 5.5 and 16 MeV, the data in each energy bin
can be subdivided in zenith angle bins:
Above 16 MeV and below 5.5 MeV, the data cannot be divided in zenith
angle bins, since the statistical precision is too low. So these energy
bins are integrated in zenith angle. The top panel of the left plot shows
an 8 bin energy spectrum, the lower panels display the zenith angle
variations of the six energy bins between 5.5 and 16 MeV.
Superimposed in red is the +-1 sigma correlated
error shape due to 8B spectrum uncertainty and energy scale and resolution
systematic uncertainty. On the right hand side are the results of a chi^2
analysis of this `zenith angle spectrum'. The excluded red area results
from the shape of the zenith angle spectrum alone, the allowed
blue area arises from constraining
the total rate to the standard solar mode. The green allowed area is
a fit to only the
rates of Gallex/GNO/SAGE, Homestake, Super-Kamiokande
and SNO.
The regions inside the dotted lines are allowed at 95% using
the zenith angle spectrum and constraining the implied
neutrino-deuteron charged-current interaction rate
to the measurement by SNO. These dotted contours are INDEPENDENT of SSM
neutrino flux predictions!
Global Analysis
Starting with the SK-SNO combined fit described above, we can add the
rate information from the radio-chemical experiments. To minimize the
dependence on SSM neutrino flux predictions, we do not constrain
either 8B or Hep neutrino flux in this analysis. We show contours
at the 1 sigma (68%), 2 sigma (95%) and 3 sigma (99.73%) confidence
level. Solutions other than the LMA only appear at the three sigma
level.
Published Results using 1258 Days of Data
Exclusion and allowed areas for two-flavour oscillation:
The red area is excluded at 95% C.L. where just the zenith angle spectrum
is used. The blue area is allowed at 95% C.L. using all Super-Kamiokande
information: Super-Kamiokande flux and zenith angle spectrum. The green
areas are globally allowed at 95% C.L. using the Super-Kamiokande flux,
the flux of the Gallium experiments and the flux of the Chlorine experiment.