Graph Convolutional Neural Networks as Surrogate Models for Climate Simulation

TL;DR

This paper demonstrates that graph convolutional neural networks can serve as efficient surrogate models for climate simulations, significantly reducing computation time while maintaining high accuracy, thus enabling faster uncertainty quantification.

Contribution

It introduces the use of GCNNs as surrogate models for climate simulations, achieving rapid results with high accuracy compared to traditional Earth-System Models.

Findings

GCNNs simulate 80 years in 310 seconds on a GPU

Mean temperature errors are below 0.1°C

Maximum errors are below 2°C

Abstract

Many climate processes are characterized using large systems of nonlinear differential equations; this, along with the immense amount of data required to parameterize complex interactions, means that Earth-System Model (ESM) simulations may take weeks to run on large clusters. Uncertainty quantification may require thousands of runs, making ESM simulations impractical for preliminary assessment. Alternatives may include simplifying the processes in the model, but recent efforts have focused on using machine learning to complement these models or even act as full surrogates. \textit{We leverage machine learning, specifically fully-connected neural networks (FCNNs) and graph convolutional neural networks (GCNNs), to enable rapid simulation and uncertainty quantification in order to inform more extensive ESM simulations.} Our surrogate simulated 80 years in approximately 310 seconds on a single A100 GPU, compared to weeks for the ESM model while having mean temperature errors below $0.1^{\circ}C$ and maximum errors below $2^{\circ}C$.