Inferring causation from time series in Earth system sciences

• Published in: Nature communications Vol. 10; no. 1; pp. 2553 - 13
• Main Authors: Runge, Jakob, Bathiany, Sebastian, Bollt, Erik, Camps-Valls, Gustau, Coumou, Dim, Deyle, Ethan, Glymour, Clark, Kretschmer, Marlene, Mahecha, Miguel D., Muñoz-Marí, Jordi, van Nes, Egbert H., Peters, Jonas, Quax, Rick, Reichstein, Markus, Scheffer, Marten, Schölkopf, Bernhard, Spirtes, Peter, Sugihara, George, Sun, Jie, Zhang, Kun, Zscheischler, Jakob
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 14.06.2019; Nature Publishing Group; Nature Portfolio
• Subjects: 639/705/1042; 639/766/530; 704/106; 704/172; 706/648/697/129; Aquatic Ecology and Water Quality Management; Aquatische Ecologie en Waterkwaliteitsbeheer; Causality; Causation; Climate change; Computer science; Computer simulation; Deep learning; Dynamical systems; Earth; Experiments; Humanities and Social Sciences; Hypothesis testing; Leerstoelgroep Aquatische ecologie en waterkwaliteitsbeheer; Machine learning; multidisciplinary; Perspective; Physics; Science; Science (multidisciplinary); Time series; WIMEK
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-019-10105-3

The heart of the scientific enterprise is a rational effort to understand the causes behind the phenomena we observe. In large-scale complex dynamical systems such as the Earth system, real experiments are rarely feasible. However, a rapidly increasing amount of observational and simulated data opens up the use of novel data-driven causal methods beyond the commonly adopted correlation techniques. Here, we give an overview of causal inference frameworks and identify promising generic application cases common in Earth system sciences and beyond. We discuss challenges and initiate the benchmark platform causeme.net to close the gap between method users and developers. Questions of causality are ubiquitous in Earth system sciences and beyond, yet correlation techniques still prevail. This Perspective provides an overview of causal inference methods, identifies promising applications and methodological challenges, and initiates a causality benchmark platform.