On some limitations of data-driven weather forecasting models

Massimo Bonavita

arXiv preprint · 15 September 2023 · ECMWF

In one sentence. Analysis of Pangu-Weather shows its forecasts lack the fidelity and physical consistency of physics-based models, suggesting part of its deterministic skill advantage comes from these peculiarities.

Abstract

As in many other areas of engineering and applied science, Machine Learning (ML) is having a profound impact in the domain of Weather and Climate Prediction. A very recent development in this area has been the emergence of fully data-driven ML prediction models which routinely claim superior performance to that of traditional physics-based models. In this work, we examine some aspects of the forecasts produced by an exemplar of the current generation of ML models, Pangu-Weather, with a focus on the fidelity and physical consistency of those forecasts and how these characteristics relate to perceived forecast performance. The main conclusion is that Pangu-Weather forecasts, and possibly those of similar ML models, do not have the fidelity and physical consistency of physics-based models and their advantage in accuracy on traditional deterministic metrics of forecast skill can be at least partly attributed to these peculiarities. Balancing forecast skill and physical consistency of ML-driven predictions will be an important consideration for future ML models. However, and similarly to other modern post-processing technologies, the current ML models appear to be already able to add value to standard NWP output for specific forecast applications and combined with their extremely low computational cost during deployment, are set to provide an additional, useful source of forecast information. .

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.