A novel approach to optimizing the image cleaning performance of Imaging Atmospheric Cherenkov Telescopes: Application to a time-based cleaning for H.E.S.S.

The Imaging Atmospheric Cherenkov Telescope (IACT) technique is essential for gamma-ray astronomy, but it suffers from performance degradation due to night sky background (NSB) noise. This degradation is mitigated through image-cleaning procedures. This study introduces a time-based cleaning method for the High Energy Stereoscopic System using CT5 in monoscopic mode, and it presents an optimization workflow for image-cleaning algorithms to enhance telescope sensitivity while minimizing systematic biases. Unlike previous optimization efforts, we do not use first-order metrics such as image size retention; instead, our pipeline focuses on the final sensitivity improvement and its systematic susceptibility to NSB. We evaluate three methods -tail-cut cleaning and two variants of time-based cleaning, <ns0:monospace>TIME3D</ns0:monospace> and <ns0:monospace>TIME4D</ns0:monospace> -and identify the best-cut configurations for two cases: optimal overall sensitivity and minimal energy threshold. The <ns0:monospace>TIME3D</ns0:monospace> method achieves a ~15% improvement compared to standard tailcut cleaning for E < 300 GeV, with a ~200% improvement for the first energy bin (36.5 GeV < E < 64.9 GeV), providing more stable performance across a wider energy range by preserving more signal. The <ns0:monospace>TIME4D</ns0:monospace> method achieves an approximately 20% improvement at low energies due to superior NSB noise suppression, allowing enhanced capability to detect sources at the lowest energies. We demonstrate that using first-order estimations of cleaning performance, such as image size retaining or NSB pixel reduction, cannot provide a complete picture of the expected result in the final sensitivity. Beyond expanding the effective area at low energies, sensitivity improvement requires precise event reconstruction, including improved energy and directional accuracy. Enhanced gamma-hadron separation and optimized pre-selection cuts further boost sensitivity. The proposed pipeline fully explores this, providing a fair and robust comparison between different cleaning methods. The method is general and can be applied to other IACT systems, such as the Very Energetic Radiation Imaging Telescope Array System and the Major Atmospheric Gamma-Ray Imaging Cherenkov Telescopes. By advancing data-driven image cleaning, this study lays the groundwork for detecting faint astrophysical sources and deepening our understanding of high-energy cosmic phenomena.