A background-estimation technique for the detection of extended gamma-ray structures with IACTs

¹ Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen Centre for Astroparticle Physics, Nikolaus-Fiebiger-Str. 2, 91058 Erlangen, Germany
² Max-Planck-Institut für Kernphysik, Saupfercheckweg 1, 69117 Heidelberg, Germany

^★ Corresponding author; tina.wach@fau.de

Received: 7 June 2024
Accepted: 23 August 2024

Abstract

Context. Estimation of the number of cosmic-ray-induced background events is a challenging task for Imaging Atmospheric Cherenkov Telescopes (IACTs). Most approaches rely on a model of the background signal derived from archival observations, which is then normalised to the region of interest (ROI) and respective observation conditions using emission-free regions in the observation. However, this is disadvantageous for the analysis of large, extended γ-ray structures, where no sufficient source-free region can be found.

Aims. We aim to address this issue by estimating the normalisation of a three-dimensional background model template from separate, matched observations of emission-free sky regions. As a result, the need for an emission-free region in the field of view of the observation becomes unnecessary.

Methods. To this end, we implemented an algorithm to identify observation pairs with the most closely matching observation conditions. We used the open-source analysis package Gammapy to estimate the background rate, facilitating seamless adaptation of the framework to many γ-ray detection facilities. We employed public data from the High Energy Stereoscopic System (H.E.S.S.) to validate this methodology.

Results. The analysis demonstrates that employing a background rate estimated through this run-matching approach yields results consistent with those obtained using the standard application of the background model template. Furthermore, we confirm the compatibility of the source parameters obtained through this approach with previous publications, and present an analysis employing the background model template approach, along with an estimation of the statistical and systematic uncertainties introduced by this method.