From air quality forecasting to integrated and intelligent urban prediction systems

Air quality forecasting has evolved into a multi-component enterprise integrating chemical transport models, emissions, observations, data assimilation, and artificial intelligence. We argue that future progress will depend less on refinement of individual components than on integrating their interfaces across urban physics, atmospheric chemistry, machine learning, dense observation networks, human activity, and decision-making systems. We outline ten emerging directions that together constitute a transition towards integrated and intelligent urban prediction systems.