Predicting weather forecast uncertainty with machine learning

TLDR

Weather forecasts are inherently uncertain, and their usefulness depends on uncertainty estimates, but the most accurate method—ensembles of numerical simulations—is computationally expensive. This study evaluates whether machine learning can predict forecast uncertainty from the large‑scale atmospheric state at initialization. The authors train convolutional neural networks on past forecasts to assign a scalar confidence value to medium‑range forecasts, indicating whether predictability is higher or lower than usual for that time of year. Although less skillful than ensemble models, the method is computationally efficient, outperforms other non‑forecast approaches, demonstrates that ML can estimate future forecast uncertainty, and its performance is limited by the amount of training data.

Abstract

Weather forecasts are inherently uncertain. Therefore, for many applications forecasts are only considered valuable if an uncertainty estimate can be assigned to them. Currently, the best method to provide a confidence estimate for individual forecasts is to produce an ensemble of numerical weather simulations, which is computationally very expensive. Here, we assess whether machine learning techniques can provide an alternative approach to predict the uncertainty of a weather forecast given the large‐scale atmospheric state at initialization. We propose a method based on deep learning with artificial convolutional neural networks that is trained on past weather forecasts. Given a new weather situation, it assigns a scalar value of confidence to medium‐range forecasts initialized from the said atmospheric state, indicating whether the predictability is higher or lower than usual for the time of the year. While our method has a lower skill than ensemble weather forecast models in predicting forecast uncertainty, it is computationally very efficient and outperforms a range of alternative methods that do not involve performing numerical forecasts. This shows that it is possible to use machine learning in order to estimate future forecast uncertainty from past forecasts. The main constraint in the performance of our method seems to be the number of past forecasts available for training the machine learning algorithm.