Speaker
Description
Direction-time clustering methods for neutral particle searches are motivated by the possibility that ultra-high-energy (UHE) neutral particles are produced during flaring periods of astrophysical sources, thus generating clusters of events correlated in arrival direction and time. In this work, we investigate how applying different photon tags improves the sensitivity of this clustering search methods in UHE photon searches. By combining multiple observables using a Boosted Decision Tree (BDT), we optimize photon tagging and enhance photon/hadron separation, effectively establishing a new BDT-based photon tag. We validate this augmented clustering framework through a Monte Carlo study, demonstrating a reduction in the number of events required to identify a photon flare across various flare durations. Finally, we present a preliminary search for point sources of UHE photons using open data and evaluate the discovery and sensitivity thresholds of the method. This demonstrates the statistical power and, potential of the method to extend current "PeVatron" searches into the UHE regime and to open a new window in multimessenger astrophysics.