We propose to design a data-driven decision support tool that quantifies and predicts the impacts of Playbook reroutes at the strategic level. This will improve the efficiency and throughput of the NAS. Our machine learning model will predict the net effects of US national reroutes in response to severe weather. This will enable faster, more accurate, and longer horizon predictions than traditional methods. Our model will form the core of a what-if analysis tool to enable air traffic managers to rank the Playbook reroute options by suitability and then understand the intended and unintended consequences of issuing such a reroute. The tool enables air traffic managers to take a strategic data-driven approach to choosing reroutes, thereby reducing disruptions to the national airspace. Currently, traffic managers lack data-driven analysis tools and instead rely on years of experience and personal preferences. Our team includes experts in traffic management, machine learning, and airspace data processing.