Diffusion Model-based Probabilistic Downscaling for 180-year East Asian Climate Reconstruction

TL;DR

This paper introduces a diffusion probabilistic downscaling model that enhances local climate detail accuracy and uncertainty quantification, applied to a 180-year East Asian climate dataset.

Contribution

The novel DPDM method improves downscaling accuracy and uncertainty assessment compared to traditional deterministic models.

Findings

More accurate local climate details

Enables ensemble uncertainty quantification

Provides a 180-year detailed climate dataset

Abstract

As our planet is entering into the "global boiling" era, understanding regional climate change becomes imperative. Effective downscaling methods that provide localized insights are crucial for this target. Traditional approaches, including computationally-demanding regional dynamical models or statistical downscaling frameworks, are often susceptible to the influence of downscaling uncertainty. Here, we address these limitations by introducing a diffusion probabilistic downscaling model (DPDM) into the meteorological field. This model can efficiently transform data from 1{\deg} to 0.1{\deg} resolution. Compared with deterministic downscaling schemes, it not only has more accurate local details, but also can generate a large number of ensemble members based on probability distribution sampling to evaluate the uncertainty of downscaling. Additionally, we apply the model to generate a 180-year dataset of monthly surface variables in East Asia, offering a more detailed perspective for understanding local scale climate change over the past centuries.