Convolutional autoencoders for spatially-informed ensemble post-processing

TL;DR

This paper introduces convolutional autoencoders to capture large-scale spatial structures in ensemble weather prediction data, enhancing post-processing accuracy by preserving spatial information often lost in interpolation.

Contribution

It proposes a novel use of convolutional autoencoders to incorporate spatial information into ensemble post-processing models, improving forecast correction.

Findings

Improved temperature forecast accuracy at surface stations.

Spatial information augmentation enhances post-processing models.

Autoencoders effectively learn compact spatial representations.

Abstract

Ensemble weather predictions typically show systematic errors that have to be corrected via post-processing. Even state-of-the-art post-processing methods based on neural networks often solely rely on location-specific predictors that require an interpolation of the physical weather model's spatial forecast fields to the target locations. However, potentially useful predictability information contained in large-scale spatial structures within the input fields is potentially lost in this interpolation step. Therefore, we propose the use of convolutional autoencoders to learn compact representations of spatial input fields which can then be used to augment location-specific information as additional inputs to post-processing models. The benefits of including this spatial information is demonstrated in a case study of 2-m temperature forecasts at surface stations in Germany.