CMG Research: Emerging Mathematical Strategies for Stochastic Modeling and Predictability to Climate Variability

New York University, New York NY

Investigators

Abstract

The climate system involves a huge range of interacting spatial and temporal scales. Development of tools to further understanding of this very complex system and its predictability is the major focus of this project. Theoretically rigorous methods allowing for the stochastic modeling of the system have been demonstrated by the investigators in simple dynamical systems. We extend these to the more complex but practical systems underlying climate. Additionally an entirely new perspective on dynamical predictability has been derived by the investigators using information theory. This will be applied to the problems of atmospheric and climate prediction and will also enable the development of rigorous reliability measures for dynamical predictions in general chaotic systems. These investigations enable us to gain a much clearer understanding of the complex system underlying the Earth's climate. Such an understanding will be of great benefit in interpreting results from societally important projections of future climate change. Further we anticipate the development of tools which will enable us to reliably decide when a particular prediction (climate or weather) is likely to be useful and when it is not likely to have any use. This ability has important practical implications for policy makers dealing with fundamentally uncertain systems.