Speaker
Description
The origin of the Fermi-LAT Galactic Center Excess (GCE)—dark matter annihilation or unresolved point sources—hinges on its spatial morphology, which is challenging to disentangle from uncertainties in Galactic diffuse emission. Standard template-fitting approaches typically assume a fixed morphology for the excess, limiting their ability to discriminate between competing interpretations and to robustly characterize uncertainties. We build on Ramirez et al. (2025), who introduced a Gaussian Process (GP) framework to flexibly reconstruct the GCE morphology with uncertainty quantification, but in a single energy band. We extend this approach to multiple energy bins (2–20 GeV) within 20° of the Galactic Center, using a GP correlation structure that links morphology across energies and exploits shared spatial information. Using synthetic data generated from three diffuse emission models (O, A, F), we validate that the method recovers the injected signal when the assumed model is correct. Under cross-model fits, performance remains broadly stable, though some degradation appears for certain components in the presence of mismodeling. These results illustrate the potential of an energy-resolved GP framework for morphological inference of the GCE, while highlighting limitations relevant for application to real data.