Universal Diffusion-Based Probabilistic Downscaling

TL;DR

This paper presents a universal diffusion-based framework for probabilistic downscaling of weather forecasts, improving resolution and uncertainty without model-specific tuning across various weather prediction systems.

Contribution

It introduces a single conditional diffusion model trained on paired data that can be applied zero-shot to diverse weather models for probabilistic high-resolution forecasts.

Findings

Ensemble mean of downscaled forecasts outperforms raw deterministic forecasts.

Significant improvements in probabilistic skill as measured by CRPS.

Effective across multiple AI and NWP weather prediction systems.

Abstract

We introduce a universal diffusion-based downscaling framework that lifts deterministic low-resolution weather forecasts into probabilistic high-resolution predictions without any model-specific fine-tuning. A single conditional diffusion model is trained on paired coarse-resolution inputs (~25 km resolution) and high-resolution regional reanalysis targets (~5 km resolution), and is applied in a fully zero-shot manner to deterministic forecasts from heterogeneous upstream weather models. Focusing on near-surface variables, we evaluate probabilistic forecasts against independent in situ station observations over lead times up to 90 h. Across a diverse set of AI-based and numerical weather prediction (NWP) systems, the ensemble mean of the downscaled forecasts consistently improves upon each model's own raw deterministic forecast, and substantially larger gains are observed in probabilistic skill as measured by CRPS. These results demonstrate that diffusion-based downscaling provides a scalable, model-agnostic probabilistic interface for enhancing spatial resolution and uncertainty representation in operational weather forecasting pipelines.