Answers are grounded in this paper and cite their sources.
This study extends the moment-based nonlinear inverse modeling (nLIM) framework from the stationary to the cyclostationary (CS) regime, introducing CS-Colored-nLIM. The proposed method constructs a periodic, colored-noise-forced, quadratic stochastic system from cyclostationary input data. In this formulation, the active nonlinear components can be empirically determined through a sparse identification scheme specifically developed for colored-noise-forced systems. Application to climate prediction demonstrates that CS-Colored-nLIM achieves predictive performance comparable to modern machine learning and statistical models, while offering greater flexibility with moderate computational demand, making it potentially applicable to a broader range of complex systems beyond climate dynamics.
No takes yet. Share an insight, caveat, or question.
Lien et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: