A Nonlinear Theory and Technology for Reducing the Uncertainty of High-Impact Ocean–Atmosphere Event Prediction

In this article, our nonlinear theory and technology for reducing the uncertainties of high-impact ocean‒atmosphere event predictions, with the conditional nonlinear optimal perturbation (CNOP) method as its core, are reviewed, and the “spring predictability barrier” problem for El Nino‒Southern Oscillation events and targeted observation issues for tropical cyclone forecasts are taken as two representative examples. Nonlinear theory reveals that initial errors of particular spatial structures, environmental conditions, and nonlinear processes contribute to significant prediction errors, whereas nonlinear technology provides a pioneering approach for reducing observational and forecast errors via targeted observations through the application of the CNOP method. Follow-up research further validates the scientific rigor of the theory in revealing the nonlinear mechanism of significant prediction errors, and relevant practical field campaigns for targeted observations verify the effectiveness of the technology in reducing prediction uncertainties. The CNOP method has achieved international recognition; furthermore, its applications further extend to ensemble forecasts for weather and climate and further enrich the nonlinear technology for reducing prediction uncertainties. It is expected that this nonlinear theory and technology will play a considerably important role in reducing prediction uncertainties for high-impact weather and climate events.