Methane (CH4) is a potent greenhouse gas contributing significantly to global warming. Landfills are critical point sources of CH4 emissions, necessitating efficient detection and monitoring methods. This study introduces PRISMethaNet, a novel convolutional neural network (CNN) integrated with an Atrous Spatial Pyramid Pooling (ASPP) mechanism, designed to enhance CH4 plume detection using PRISMA hyperspectral satellite data.