Reconstructing Historical Climate Fields With Deep Learning
This work addresses the challenge of incomplete historical climate records for climate scientists, representing an incremental improvement over existing methods.
The paper tackles the problem of reconstructing historical climate fields from sparse data by using a deep-learning approach based on Fourier convolutions, achieving realistic reconstructions of large missing areas and outperforming statistical kriging and other machine learning methods.
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 Niño and La Niña 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.