A Discrimination Procedure between Muon and Electron in Superkamiokande Experiment Based on the Angular Distribution Function Method

TL;DR

This paper develops an angular distribution function method to distinguish muons from electrons in Superkamiokande, achieving high discrimination accuracy and improved event reconstruction over previous geometrical methods.

Contribution

It introduces a detailed angular distribution function approach for particle discrimination and compares its effectiveness with simpler geometrical reconstruction methods.

Findings

Muon-electron discrimination error less than several percent.

Full image analysis significantly improves vertex and angular resolution.

Method reveals limitations in SK simulation assumptions.

Abstract

In the previous paper, we construct the angular distribution functions for muon and electron as well as their relative fluctuation functions to find suitable discrimination procedure between muon and electron in Superkamiokande experiment. In the present paper, we are able to discriminate muons from electrons in Fully Contained Events with a probability of error of less than several %. At the same time, our geometrical reconstruction procedure, considering only the ring-like structure of the Cherenkov image, gives an unsatisfactory resolution for 1GeV electron and muon, with a mean vertex position error, delta r, of 5-10 m and a mean directional error, delta theta, of about 6-20 degrees. In contrast, a geometrical reconstruction procedure utilizing the full image and using a detailed approximation of the event angular distribution works much better: for a 1 GeV electron, delta r is about 2 m and delta theta is about 3 degrees; for a 1GeV muon, delta r is about 3 m and delta theta is about 5 degrees. At 5 GeV, the corresponding values are about 1.4 m and about 2 degree for electron and are about 2.9m and about 4.3 degrees for muon. The numerical values depend on a single PMT contribution threshold. The values quoted above are the minima with respect to this threshold. Even the methodologically correct approach we have adopted, based on detailed simulations using closer approximations than those adopted in the SK analysis, cannot reproduce the accuracies for particle discrimination, momentum resolution, interaction vertex location, and angular resolution obtained by the SK simulations, suggesting the assumptions in these may be inadequate.