Morphometric analysis in gamma-ray astronomy using Minkowski functionals: II. Joint structure quantification

TL;DR

This paper develops a morphometric method using Minkowski functionals to detect gamma-ray sources by analyzing structural deviations in counts maps, providing a new way to characterize background noise without prior source knowledge.

Contribution

It introduces a joint distribution approach for Minkowski functionals in gamma-ray maps, enabling unbiased background structure analysis and improved source detection sensitivity.

Findings

Accurate background structure characterization for large scan windows.

Confirmation of statistical significance of background features.

Enhanced sensitivity through local detector effect correction.

Abstract

We pursue a novel morphometric analysis to detect sources in very-high-energy gamma-ray counts maps by structural deviations from the background noise. Because the Minkowski functionals from integral geometry quantify the shape of the counts map itself, the morphometric analysis includes unbiased structure information without prior knowledge about the source. Their distribution provides access to intricate geometric information about the background. We combine techniques from stochastic geometry and statistical physics to determine the joint distribution of all Minkowski functionals. We achieve an accurate characterization of the background structure for large scan windows (with up to $15\times15$ pixels), where the number of microstates varies over up to 64 orders of magnitude. Moreover, in a detailed simulation study, we confirm the statistical significance of features in the background noise and discuss how to correct for trial effects. We also present a local correction of detector effects that can considerably enhance the sensitivity of the analysis. In the third paper of this series, we will use the here derived refined structure characterization for a more sensitive data analysis that can detect formerly undetected sources.