29 September 2017 from 11:40 to 13:10

CLUe lunch meeting #9

The bimonthly CLUe meetings aim to provide an interacting platform for topic-oriented presentations on Complexity Science and real-time discussions among researchers from different fields.

Short Bio about the Keynote Speaker

Prof. Sandra Chapman is Professor of Physics and Director of the Centre for Fusion, Space and Astrophysics at the University of Warwick, UK. This year she received the Fulbright Lloyds of London Scholar Award - one of the best-regarded and impactful scholarship programmes in the world. She has pioneered the development of nonlinear and complex systems approaches to solar system and laboratory plasmas and more widely, to problems outside plasma physics including climate and neuroscience. She develops and applies data analysis techniques to plasma turbulence, 'space weather' and more widely in ‘real world’ non-linear and complex systems, most recently in observations of our changing climate.




11:45 – 12:00              Lunch


12:00 – 12:10             CLUe Developments

                                   Prof. dr. Henk Dijkstra, Head of the Department of Physics

Keynote Talk:

12:10 – 13:10             Dynamical networks characterization of natural data: space weather events and the human brain response to surprise

                                   Prof. Sandra Chapman, University of Warwick


Complexity topics: Networks

Content at a glance: Prof. Sandra Chapman will show you dynamical networks characterization of natural data, from the human brain to space weather.


Dynamical networks are in widespread use in quantifying societal data. Here, we consider the possibilities of this approach in real world observations of natural systems where multipoint (in space) observations are available of multiscale systems that exhibit an emergent nonlinear response to an impulse or change in driving. Two examples will be discussed in detail. Magnetoencephalography (MEG) is a noninvasive human brain diagnostic in which the magnetic field at the scalp surface is measured at typically 100 equally spaced points. It can be used to infer how the functional network of the human brain dynamically emerges in response to stimuli, here, repeated and unexpected or surprising auditory tones. Differences in this response between healthy subjects and those with a neurological condition such as schizophrenia may offer the potential for early diagnosis. The plasma and magnetic field of earths near-space environment is highly dynamic, with its own space weather which is an emergent response to the highly variable sun’s expanding atmosphere, the solar wind. Space weather effects on the ground are monitored by 100+ magnetometer stations in the auroral region as well as by a constellation of satellites.  Spatio-temporal patterns of correlation between the magnetometer time series can be used to form a dynamical network. Space weather impacts include power loss, aviation disruption, communication loss, and disturbance to or loss of satellite systems, on some of which a range of technologies depend for navigation or timing.

A dynamical networks approach to these datasets offers the possibility of capturing the time variation in emergent spatial pattern with a few time-varying network parameters. It could in principle be used to compare across many subjects or events and to obtain an averaged or typical system response if this is meaningful, and its dependence on system parameters. Whilst networks are in widespread use in the data analytics of societal and commercial data, there are additional challenges in their application to physical timeseries. Determining whether two nodes (here, ground based magnetometer stations or MEG sensors) are connected in a network (seeing the same dynamics) requires normalization w.r.t. the detailed sensitivities and dynamical responses of specific observations and instrumentation. The spatial sampling points may be signal integrating and may not be uniformly spatially distributed. In the case of ground based observations of space weather, the stations are moving w.r.t. the plasma-current system under observation. In the case of MEG human brain measurements, the background is highly variable both within and across subjects. In both the human brain and in space weather the observed current system itself is non-linear and highly dynamic. As well as presenting some of the first dynamical network analysis of these natural systems, this talk will focus on the challenges of natural data and some potential approaches.


Room 7.12, Buys Ballotgebouw, Princetonplein 5, De Uithof, 3584 CC Utrecht


The lunch is FREE for all the participants, please register before Tuesday 26 September


Start date and time
29 September 2017 11:40
End date and time
29 September 2017 13:10