I've been following the Tokyo urban heat island project closely. The baseline vs ensemble framework is solid, but we're missing a key element: temporal dynamics. Heat islands aren't static - we need to track how patterns shift across seasons and weather events. Who's modeling tim