Workshop - Computer Simulation: Prediction, Risk, Uncertainty

Society & HLRS
Workshop - Computer Simulation: Prediction, Risk, Uncertainty


Hildrun Lampe (Philophy / HLRS):
"Introduction: Computer simulation - prediction, risk, uncertainty"

Jens Kouros (Sociology / HLRS):
"Stochastic modelling of systemic risk with probabilistic programming"

Systemic Risk is regarded as a new topic in risk research. Even though there is no consensual definition, the term generally refers to complex systems integrating social, economical, technical and ecological aspects. Because of their complexity it is a major difficulty to get a grip on such systems and the processes that take place in them, with respect to scientific advancement and policy advice. An attempt is made to tackle these problems using generative models in a probabilistic programming language. This allows for more expressive models than would be possible using more standard statistical models.

Prof. Dr. Holger Lange (Environmental Science / Norwegian Institute of Bioeconomy Research):
"Time Series, models and uncertainty: process-based and data-based approaches to environmental problems"

A few difficult problems of environmental science are presented. Two different approaches to their solution are discussed: the process-based and the data-based approach. The data-based approach allows for a classification of measured time series according to their complexity, and their stochastic and deterministic parts, respectively. These classifications can be used to evaluate process-based models qualitatively and quantitatively. Examples from hydrology, the carbon cycle and vegetation productivity are given.