Session

12.09.2017     16:30–18:30

Title:
MS39
Dynamics and data in stochastic systems, far from equilibrium
Type:
Atmosphere / climate / weather models
Room:

The challenges in complex systems are familiar: coupling of physics at different scales, high dimensions, model uncertainty, unresolved scales, etc.  Further complications arise when equilibrium ideas arising from statistical physics do not apply. The purpose of this mini-symposium is to foster a cross-fertilization  of computational, statistical, and theoretical techniques and strategies, applicable to complex systems, far from equilibrium.  We will focus on strategies that combine dynamics, stochasticity, and data. The talks will expand on data-driven, computational as well as theoretical approaches.

12.09.2017
16:30–18:30

Title:
MS39
Dynamics and data in stochastic systems, far from equilibrium
Type:
Atmosphere / climate / weather models
Room:

The challenges in complex systems are familiar: coupling of physics at different scales, high dimensions, model uncertainty, unresolved scales, etc.  Further complications arise when equilibrium ideas arising from statistical physics do not apply. The purpose of this mini-symposium is to foster a cross-fertilization  of computational, statistical, and theoretical techniques and strategies, applicable to complex systems, far from equilibrium.  We will focus on strategies that combine dynamics, stochasticity, and data. The talks will expand on data-driven, computational as well as theoretical approaches.


16:30–16:50
O202 Stochastic modelling of transport noise
Dan Crisan (London/GB)


16:50–17:10
O203 A dynamically driven paradigm for data assimilation in geophysical models
Colin Grudzien (Bergen/NO)


17:10–17:30
O204 Assessing the reliability of ensemble forecasting systems
Jochen Broecker (Reading/GB)


17:30–17:50
O205 Iterative updating of model error in Bayesian inversion for efficient posterior sampling
Matthew Dunlop (Pasadena/US)


17:50–18:10
O206 Tensor product decomposition methods for compact representations of large data sets
Thomas von Larcher (Berlin/DE)