Skillful High-Resolution Ensemble Precipitation Forecasting with an Integrated Deep Learning Framework

TL;DR

This paper introduces a physics-inspired deep learning framework that combines deterministic and probabilistic models to improve high-resolution ensemble precipitation forecasts, effectively capturing extreme events and uncertainties.

Contribution

The novel integrated deep learning framework combines a 3D SwinTransformer and conditional diffusion models for enhanced high-resolution precipitation forecasting.

Findings

Significantly improves spatial resolution and forecast accuracy.

Ensemble system is reliable and unbiased according to rank histogram.

Outperforms ERA5 in capturing heavy precipitation events.

Abstract

High-resolution precipitation forecasts are crucial for providing accurate weather prediction and supporting effective responses to extreme weather events. Traditional numerical models struggle with stochastic subgrid-scale processes, while recent deep learning models often produce blurry results. To address these challenges, we propose a physics-inspired deep learning framework for high-resolution (0.05\textdegree{} $\times$ 0.05\textdegree{}) ensemble precipitation forecasting. Trained on ERA5 and CMPA high-resolution precipitation datasets, the framework integrates deterministic and probabilistic components. The deterministic model, based on a 3D SwinTransformer, captures average precipitation at mesoscale resolution and incorporates strategies to enhance performance, particularly for moderate to heavy rainfall. The probabilistic model employs conditional diffusion in latent space to account for uncertainties in residual precipitation at convective scales. During inference, ensemble members are generated by repeatedly sampling latent variables, enabling the model to represent precipitation uncertainty. Our model significantly enhances spatial resolution and forecast accuracy. Rank histogram shows that the ensemble system is reliable and unbiased. In a case study of heavy precipitation in southern China, the model outputs align more closely with observed precipitation distributions than ERA5, demonstrating superior capability in capturing extreme precipitation events. Additionally, 5-day real-time forecasts show good performance in terms of CSI scores.