A Bayesian on-off analysis of cosmic ray data

TL;DR

This paper introduces a Bayesian method for analyzing weak cosmic ray sources in on-off measurements, effectively handling small signal counts and uncertain backgrounds to improve detection and comparison of measurements.

Contribution

It presents a Bayesian framework for on-off cosmic ray data analysis, including posterior distributions and prediction tools, addressing small sample challenges and measurement disparities.

Findings

Effective Bayesian analysis of weak cosmic ray signals.

Quantitative measures of measurement disparities.

Demonstrated approach on cosmic ray data examples.

Abstract

We deal with the analysis of on-off measurements designed for the confirmation of a weak source of events whose presence is hypothesized, based on former observations. The problem of a small number of source events that are masked by an imprecisely known background is addressed from a Bayesian point of view. We examine three closely related variables, the posterior distributions of which carry relevant information about various aspects of the investigated phenomena. This information is utilized for predictions of further observations, given actual data. Backed by details of detection, we propose how to quantify disparities between different measurements. The usefulness of the Bayesian inference is demonstrated on examples taken from cosmic ray physics.