The Predictability of Tropical Pacific Decadal Variability: Insights from Attractor Reconstruction
JOURNAL OF THE ATMOSPHERIC SCIENCES(2019)
摘要
Tropical Pacific decadal variability (TPDV), though not the totality of Pacific decadal variability, has wide-ranging climatic impacts. It is currently unclear whether this phenomenon is predictable. In this study, we reconstruct the attractor of the tropical Pacific system in long, unforced simulations from an intermediate-complexity model, two general circulation models (GCMs), and the observations with the aim of assessing the predictability of TPDV in these systems. We find that in the intermediate-complexity model, positive (high variance, El Nino-like) and negative (low variance, La Nina-like) phases of TPDV emerge as a pair of regime-like states. The observed system bears resemblance to this behavior, as does one GCM, while the other GCM does not display this structure. However, these last three time series are too short to confidently characterize the full distribution of interdecadal variability. The intermediate-complexity model is shown to lie in highly predictable parts of its attractor 37% of the time, during which most transitions between TPDV regimes occur. The similarities between the observations and this system suggest that the tropical Pacific may be somewhat predictable on interdecadal time scales.
更多查看译文
关键词
Atmosphere,Ocean,Tropics,Nonlinear dynamics,Pacific decadal oscillation,Climate prediction
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络