Generating a 1000 Member Ensemble of Horizontal Vapor Transport Forecasts with Diffusion

Ensemble weather predictions can be immensely beneficial towards making informed decisions for operations that require risk assessment. The traditional approach towards creating an ensemble often involves making slight perturbations to initial conditions and model physics to account for forecast uncertainty. This study instead uses diffusion, a form of generative artificial intelligence (AI), to make ensemble forecasts consisting of 1000 members from single-member deterministic dynamical West-WRF forecasts over the North Pacific. The diffusion ensemble can make predictions at high speeds and at a low computational expense while improving the skill of medium-range integrated vapor transport (IVT). It has high performance when evaluated by binned ranked histograms, Continuous Ranked Probability Score, and binned spread to Root Mean Square Error ratio. The diffusion ensemble further demonstrated utility in representing real outcomes of extreme destructive events with realistic high-quality images.