Predictability horizon properties of chaotic dynamical systems can be related to their spectral properties. It is shown, using this relationship, that the spectral properties of the leading large-scale climate daily indices indicate a smooth predictability: i.e. their intrinsic predictability horizons can be indefinitely extended by reducing the initial error. Special properties of the Madden-Julian Oscillation and some models (including ensemble weather forecasting) have been also discussed in this context.
Multimodel ensembling has been widely used to improve climate model predictions, and the improvement strongly depends on the ensembling scheme. In this work, we propose a Bayesian neural network (BNN) ensembling method, which combines climate models within a Bayesian model averaging framework, to improve the predictive capability of model ensembles. Our proposed BNN approach calculates spatiotemporally varying model weights and biases by leveraging individual models' simulation skill, calibrates the ensemble prediction against observations by considering observation data uncertainty, and quant…
Many phenomena in the climate system lie in the gray zone between weather and climate: they are not amenable to deterministic forecast, but they still depend on the initial condition. A natural example is medium-range forecasting, which is inherently probabilistic because it lies beyond the deterministic predictability time of the atmosphere, but for which statistically significant prediction can be made which depend on the current state of the system. Similarly, one may ask the probability of occurrence of an El Niño event several months ahead of time. In this paper, we introduce a quantity w…
Long-term prediction of ENSO with physics-guided Deep Echo State Networks
School of Physical Science and Technology, Beijing University of Posts and Telecommunications（北京邮电大学物理科学与技术学院）；State Key Laboratory of Earth System Numerical Modeling and Application, Institute of Atmospheric Physics, Chinese Academy of Sciences（中国科学院大气物理研究所地球系统数值模拟与应用国家重点实验室）；School of Systems Science/Institute of Nonequilibrium Systems, Beijing Normal University（北京师范大学系统科学系/非平衡系统研究所）；State Key Laboratory of Information Photonics and Optical Communications, Beijing University of Posts and Telecommunications（北京邮电大学信息光子学与光通信国家重点实验室）；Frontier Science Center for Deep Ocean Multispheres and Earth System (FDOMES) and Physical Oceanography Laboratory, Ocean University of China（中国海洋大学深海多圈层与地球系统前沿科学中心及物理海洋学实验室）；Laoshan Laboratory（崂山实验室）
The El Niño Southern Oscillation (ENSO) is the most important driver of interannual global climate variability and can trigger extreme weather events and disasters in various parts of the globe. Depending on the region of maximal warming, El Niño events can be partitioned into 2 types, Eastern Pacific (EP) and Central Pacific (CP) events. The type of an El Niño has a major influence on its impact and can even lead to either dry or wet conditions in the same areas on the globe. Here we show that the zonal difference $ΔT_{WP-CP}$ between the sea surface temperature anomalies (SSTA) in the equato…
Using direct numerical simulation we study the behavior of the maximal Lyapunov exponent in thin-layer turbulence, where one dimension of the system is constrained geometrically. Such systems are known to exhibit transitions from fully three dimensional turbulence through a mixed two and three dimensional phenomenology state and then onto fully two dimensional dynamics. We find a discontinuous jump in the Lyapunov exponent at this second transition implying the predictability of such systems can change dramatically. Such transitions are seen in a number of different turbulent systems, for exam…
We propose a comprehensive framework able to address both the predictability of the first and of the second kind for high-dimensional chaotic models. For this purpose, we analyse the properties of a newly introduced multistable climate toy model constructed by coupling the Lorenz '96 model with a zero-dimensional energy balance model. First, the attractors of the system are identified with Monte Carlo Basin Bifurcation Analysis. Additionally, we are able to detect the Melancholia state separating the two attractors. Then, Neural Ordinary Differential Equations are applied in order to predict t…
Error Growth Dynamic and Predictability of Tropical Cyclone in Machine Learning Weather Prediction Model
Fudan University（复旦大学）；Shanghai Key Laboratory of Ocean-land-atmosphere Boundary Dynamics and Climate Change（上海陆海气边界动力学与气候变化重点实验室）；Shanghai Frontiers Science Center of Atmosphere–Ocean Interaction（上海大气-海洋相互作用前沿科学中心）；Shanghai Academy of Artificial Intelligence for Science（上海人工智能科学academy）
Deep learning the sources of MJO predictability: a spectral view of learned features
University of Chicago（芝加哥大学）；Allen Institute for Artificial Intelligence (AI2)（艾伦人工智能研究所）；University of California, Santa Cruz（加利福尼亚大学圣克鲁斯分校）；Lawrence Berkeley National Laboratory（劳伦斯伯克利国家实验室）；University of California, Berkeley（加利福尼亚大学伯克利分校）
The climate system is a complex, chaotic system with many degrees of freedom and variability on a vast range of temporal and spatial scales. Attaining a deeper level of understanding of its dynamical processes is a scientific challenge of great urgency, especially given the ongoing climate change and the evolving climate crisis. In statistical physics, complex, many-particle systems are studied successfully using the mathematical framework of Large Deviation Theory (LDT). A great potential exists for applying LDT to problems relevant for geophysical fluid dynamics and climate science. In parti…
