Accurate Attribution and Seasonal Prediction of Climatic Anomalies Using Causal Inference Theory

Abstract Using features based on correlation or noncausal dependence metrics can lead to false conclusions. However, recent research has shown that applying causal inference theory in conjunction with Bayesian networks large-sample-size data accurately attribute synoptic anomalies. Focusing the East Asian summer monsoon (EASM), this study adopts a approach model averaging investigate causation of interannual climate variability. We EASM variability five winter phenomena; our result shows eastern Pacific El Niño–Southern Oscillation largest effect. also show precursors are interpretable terms physics. linear regression, these predict one season ahead, outperforming correlation-based empirical models and three models. This even without substantial human intervention, laymen implement causes climatic anomalies construct reliable for prediction. Significance Statement use redesign attribution procedure fundamentally adjust commonly used data. Our exhaustively reveal little which is impossible studies. According attribution, better predictive performance than Therefore, will tremendously help both meteorologists (e.g., stakeholders policymakers) phenomena their causes. recommend it become standard practice studies operational