Workshop organized by: V. Livina, S. Blesic and J. Ludescher
This session is inviting submissions of work of modelling and forecasting of dynamical systems. It aims to showcase how data availability shapes modelling and prediction and how the issues of lack of data or temporal or spatial inconsistency and inherent biases of data sources may be overcome.
Real-world records originate from diverse sources and are collected with different objectives, instruments, methodologies, and levels of precision. This variability represents a challenge in construction and maintenance of continuous and harmonized time series of dynamical systems, which is essential for their modelling and forecasting. In addition, there often exists a scarcity of historical information (metadata) in such systems that undermines understanding and modelling of the long-term patterns and behavior. Furthermore, some dynamical systems depend on variables with such intermittent dynamics that physical understanding, more than a prediction, may be a more plausible research goal.
The session is interested in examples of different modelling and forecast solutions that consider and solve problems of uncertainty, consistency with physical laws, and long-time scale in dynamical systems. Contributions are welcomed from, but not limited to, various areas of geosciences, like climate or environmental sciences, engineering, socio- and econophysics, biophysics, and related sensory systems. We encourage submissions on surrogate data simulations, model validation using ground truth measurements, and measurement uncertainties.