Speaker
Description
A primary advantage of Imaging Atmospheric Cherenkov Telescope (IACT) arrays over other ground-based gamma-ray detectors is their superior angular resolution. This capability is crucial for studying the morphology of gamma ray sources. Recent observations by ground-based detectors like LHAASO and HAWC have revealed a large number of extended TeV gamma-ray sources, highlighting the need for detailed studies of source structure and emission mechanisms. For IACT array, in order to fully characterize these extended sources and potentially resolve sub-structures within them, excellent angular resolution at large offset angles from the telescope pointing direction is essential. Furthermore, IACTs offer a significant advantage at large zenith angles: their effective area increases dramatically with zenith angle, making them particularly well-suited for studying PeVatrons.However, both large offset and large zenith angle observations present significant challenges to accurate shower reconstruction.
This study investigates the angular resolution performance of various reconstruction methods based on the simulation of the LACT Project. We compare the standard Hillas parameter-based intersection method, the Disp method, the ImPACT (Image Pixel-wise fit for Atmospheric Cherenkov Telescopes) reconstruction, and a machine learning approach based on neural networks. Our analysis focuses primarily on the angular resolution degradation at large offset angles and large zenith angles.
In standard Hillas direction reconstruction, we identified three critical factors affecting angular resolution: the multiplicity of the event, the intersection angle between telescopes, and the individual telescope's plane reconstruction accuracy(MISS). At large offset angles, all three factors degrade compared to on-axis observations. Preliminary results indicate that the Disp method significantly improves if the key factor is the intersection angle. However, it's not so effective if the individual telescope's plane reconstruction accuracy becomes worse. Consequently, more sophisticated reconstruction techniques, like ImPACT and neural networks, which leverage more detailed information from the shower images, are required to mitigate the degradation and achieve superior angular resolution at large offsets and zenith angle.
