Can AI be enabled to dynamical downscaling? A Latent Diffusion Model to mimic km-scale COSMO5.0\_CLM9 simulations

TL;DR

This paper demonstrates that a Latent Diffusion Model can effectively downscale coarse Earth system data to high-resolution fields, matching the realism of dynamical models and outperforming traditional deep learning approaches.

Contribution

The study introduces a novel application of Latent Diffusion Models for Earth system downscaling, showing improved performance over existing deep learning methods.

Findings

LDM achieves high-resolution downscaling comparable to dynamical models.

LDM outperforms quadratic interpolation, UNET, and GAN baselines.

The approach preserves realistic fine-scale features and flow characteristics.

Abstract

Downscaling techniques are one of the most prominent applications of Deep Learning (DL) in Earth System Modeling. A robust DL downscaling model can generate high-resolution fields from coarse-scale numerical model simulations, saving the timely and resourceful applications of regional/local models. Additionally, generative DL models have the potential to provide uncertainty information, by generating ensemble-like scenario pools, a task that is computationally prohibitive for traditional numerical simulations. In this study, we apply a Latent Diffusion Model (LDM) to downscale ERA5 data over Italy up to a resolution of 2 km. The high-resolution target data consists of 2-m temperature and 10-m horizontal wind components from a dynamical downscaling performed with COSMO_CLM. Our goal is to demonstrate that recent advancements in generative modeling enable DL to deliver results comparable to those of numerical dynamical models, given the same input data, preserving the realism of fine-scale features and flow characteristics. A selection of predictors from ERA5 is used as input to the LDM, and a residual approach against a reference UNET is leveraged in applying the LDM. The performance of the generative LDM is compared with reference baselines of increasing complexity: quadratic interpolation of ERA5, a UNET, and a Generative Adversarial Network (GAN) built on the same reference UNET. Results highlight the improvements introduced by the LDM architecture and the residual approach over these baselines. The models are evaluated on a yearly test dataset, assessing the models' performance through deterministic metrics, spatial distribution of errors, and reconstruction of frequency and power spectra distributions.