Trajectory learning for ensemble forecasts via the continuous ranked probability score: a Lorenz '96 case study
This incremental improvement addresses ensemble forecast accuracy for meteorological applications, particularly benefiting data assimilation systems.
The paper tackled the problem of generating accurate ensemble weather forecasts by developing stochastic parametrizations trained with the continuous ranked probability score (CRPS) loss function, demonstrating that this approach outperforms derivative-fitting methods in short-term predictions on the Lorenz '96 system.
This paper demonstrates the feasibility of trajectory learning for ensemble forecasts by employing the continuous ranked probability score (CRPS) as a loss function. Using the two-scale Lorenz '96 system as a case study, we develop and train both additive and multiplicative stochastic parametrizations to generate ensemble predictions. Results indicate that CRPS-based trajectory learning produces parametrizations that are both accurate and sharp. The resulting parametrizations are straightforward to calibrate and outperform derivative-fitting-based parametrizations in short-term forecasts. This approach is particularly promising for data assimilation applications due to its accuracy over short lead times.