Reconstructing Historical Climate Fields With Deep Learning

TL;DR

This paper introduces a deep learning method based on Fourier convolutions to reconstruct missing historical climate data, outperforming traditional statistical methods and generalizing across resolutions and climate events.

Contribution

The paper presents a novel deep-learning approach that effectively reconstructs historical climate fields, surpassing existing methods and handling higher resolutions and unseen masks.

Findings

Outperforms kriging and other machine learning methods

Successfully reconstructs large missing data areas and historical events

Generalizes to higher resolutions and unseen masks

Abstract

Historical records of climate fields are often sparse due to missing measurements, especially before the introduction of large-scale satellite missions. Several statistical and model-based methods have been introduced to fill gaps and reconstruct historical records. Here, we employ a recently introduced deep-learning approach based on Fourier convolutions, trained on numerical climate model output, to reconstruct historical climate fields. Using this approach we are able to realistically reconstruct large and irregular areas of missing data, as well as reconstruct known historical events such as strong El Ni\~no and La Ni\~na with very little given information. Our method outperforms the widely used statistical kriging method as well as other recent machine learning approaches. The model generalizes to higher resolutions than the ones it was trained on and can be used on a variety of climate fields. Moreover, it allows inpainting of masks never seen before during the model training.