In-Theatre Numerical Weather Prediction with Deep Learning-Powered Edge Computing

To continue the development of WeatherMesh and our partnership with the DAF, WindBorne will answer the needs we discovered through our Phase I work with the 16 WS and the ACC Numerical Weather Modeling Group by building a fully AI-based numerical weather prediction (NWP) system that can run many ensemble members, training WeatherMesh to downscale to high resolution or point-forecast outputs, and incorporating our live WindBorne GSB observations into data assimilation for the WeatherMesh ensemble forecasting system. The work we complete in Phase II will later enable us to build an edge computing system powered by deep learning that can enable the warfighter to run advanced NWP on a high-powered laptop anywhere in the world within minutes for the purposes of mission planning, mission readiness, go/no-go decision-making, and augmenting operational execution.This type of technical efficiency will change what is possible, because, where physics-based NWP currently requires massive machines and multiple servers, AI-based NWP could be run on a relatively low-cost server operating in the field, or even in the cockpit. This will give the warfighter access to live, dynamic weather prediction capabilities, which will enable more informed decision-making based on more accurate, up-to-the-minute environmental data. It is this type of technical advancement that will ensure that DAF remains dominant in ISR and battlespace awareness in the age of AI.