In this Cooperation Group, we join forces from three different disciplines in order to address issues related to physical models of climate obtained from the Earth's astronomical cycles and palaeo-climatic data, and their statistical and stochastic analysis.
Because of the long time scales on which climate varies, the phenomena to be modeled and understood to a large part refer to palaeo-climate. The fluctuations of the earth's global average temperature over the past hundreds of thousands to million years call for an understanding of the interaction of weak periodic driving forces and their stochastic perturbations, as well as of the (parametric) causes of dramatic dynamical transitions. The background knowledge is physics based climatology.
Since knowledge of past climate is intricately encoded in data series stemming from, e.g., ice cores or sediments, statistical aspects of data analysis are essential in order to understand and improve the accuracy of input data. Also the extraction of model parameters and model validation is a challenging statistical task.
Finally, the appropriate models will be stochastic. Hence, the mathematical tools of prob- ability and stochastic processes and their dynamics are indispensable for the analytical under- standing and handling of models obtained.
Our project therefore aims at a joint effort of three disciplines: non-linear physics of complex interacting systems, statistical inference and time series analysis, and of stochastic analysis and random dynamical systems. The main goal is to improve physical models, e.g., by using non-Gaussian noise processes, to calibrate and validate models by adequate and partially new statistical methods, and to understand and explore these models from the mathematical point of view.
Please direct any inquiries about the scientific program of the Cooperation Group to Prof. Dr. Ilya Pavlyukevich (firstname.lastname@example.org).