Weather Prediction with Diffusion Guided by Realistic Forecast Processes

TL;DR

This paper introduces a diffusion model-based weather forecasting approach that integrates traditional NWP predictions, offering flexible, reliable, and improved long-term forecasts with the ability to incorporate various data sources.

Contribution

The novel diffusion model framework enables both direct and iterative weather forecasting, integrating NWP predictions during sampling, which enhances flexibility and trustworthiness.

Findings

Demonstrated feasibility and generalizability of the diffusion-based approach

Enhanced long-term forecast stability with climatology and persistence data

Allows integration of NWP predictions with varying lead times

Abstract

Weather forecasting remains a crucial yet challenging domain, where recently developed models based on deep learning (DL) have approached the performance of traditional numerical weather prediction (NWP) models. However, these DL models, often complex and resource-intensive, face limitations in flexibility post-training and in incorporating NWP predictions, leading to reliability concerns due to potential unphysical predictions. In response, we introduce a novel method that applies diffusion models (DM) for weather forecasting. In particular, our method can achieve both direct and iterative forecasting with the same modeling framework. Our model is not only capable of generating forecasts independently but also uniquely allows for the integration of NWP predictions, even with varying lead times, during its sampling process. The flexibility and controllability of our model empowers a more trustworthy DL system for the general weather community. Additionally, incorporating persistence and climatology data further enhances our model's long-term forecasting stability. Our empirical findings demonstrate the feasibility and generalizability of this approach, suggesting a promising direction for future, more sophisticated diffusion models without the need for retraining.