EGU24-8464, updated on 08 Mar 2024
https://doi.org/10.5194/egusphere-egu24-8464
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Machine Learning for Multivariate Downscaling: A Generative Model Inspired by Forecast Evaluation

Maybritt Schillinger1, Xinwei Shen1, Maxim Samarin2, and Nicolai Meinshausen1
Maybritt Schillinger et al.
  • 1ETH Zurich, Seminar for Statistics, Department of Mathematics, Zürich, Switzerland (maybritt.schillinger@stat.math.ethz.ch)
  • 2Swiss Data Science Center, Zurich, Switzerland

To complement computationally expensive regional climate model (RCM) simulations, machine learning methods can predict the high-resolution RCM data from low-resolution global climate model (GCM) input. Instead of merely targeting the conditional mean of the RCM field given the GCM data, more recent works are based on generative adversarial networks or diffusion models and aim to learn the full conditional distribution. In this spirit, we present a novel generative model that relies on statistical tools from forecast evaluation. The model can sample several plausible RCM realisations and enables assessing their variability. To achieve this goal, we use a simple neural network architecture that predicts Fourier coefficients of the high-resolution fields for multiple variables jointly (temperature, precipitation, solar radiation and wind). The loss function of our model is a proper scoring rule that measures the discrepancy between the model’s predictive distribution and the RCM’s true distribution. The score is minimised if both distributions agree. Our generative model is trained on multiple GCM-RCM combinations from the Euro-Cordex project. Furthermore, we show how the framework can be augmented to perform a bias-correction task: With a modified loss function, it is possible to generate data from the observational distribution, for example resembling gridded E-OBS data. To summarise, our work presents a machine learning method that allows us to generate multivariate high-resolution climate data, and can be extended flexibly to include further variables or downscale and bias-correct future projections.

How to cite: Schillinger, M., Shen, X., Samarin, M., and Meinshausen, N.: Machine Learning for Multivariate Downscaling: A Generative Model Inspired by Forecast Evaluation, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8464, https://doi.org/10.5194/egusphere-egu24-8464, 2024.

Supplementary materials

Supplementary material file

Comments on the supplementary material

AC: Author Comment | CC: Community Comment | Report abuse

supplementary materials version 1 – uploaded on 23 May 2024, no comments