Towards Physically Consistent Deep Learning For Climate Model Parameterizations

TL;DR

This paper introduces a supervised learning framework for climate model parameterizations that ensures physical consistency and interpretability by focusing on key physical features, reducing spurious correlations, and maintaining predictive accuracy.

Contribution

The authors propose a novel feature-focused training method that enhances interpretability and physical consistency of deep learning-based climate parameterizations without sacrificing performance.

Findings

Identifies relevant physical features for climate processes

Produces physically consistent neural network models

Maintains predictive accuracy comparable to unconstrained models

Abstract

Climate models play a critical role in understanding and projecting climate change. Due to their complexity, their horizontal resolution of about 40-100 km remains too coarse to resolve processes such as clouds and convection, which need to be approximated via parameterizations. These parameterizations are a major source of systematic errors and large uncertainties in climate projections. Deep learning (DL)-based parameterizations, trained on data from computationally expensive short, high-resolution simulations, have shown great promise for improving climate models in that regard. However, their lack of interpretability and tendency to learn spurious non-physical correlations result in reduced trust in the climate simulation. We propose an efficient supervised learning framework for DL-based parameterizations that leads to physically consistent models with improved interpretability and negligible computational overhead compared to standard supervised training. First, key features determining the target physical processes are uncovered. Subsequently, the neural network is fine-tuned using only those relevant features. We show empirically that our method robustly identifies a small subset of the inputs as actual physical drivers, therefore removing spurious non-physical relationships. This results in by design physically consistent and interpretable neural networks while maintaining the predictive performance of unconstrained black-box DL-based parameterizations.