Prof Myles Allen

Chaos and climate change

Video Podcast Presentation (PDF)

In this lecture, Myles Allen addressed how computers have transformed our understanding of the role of chaos and exponential error growth in weather forecasting; and our understanding of how climate change is impacting regional weather. He showed how research in Oxford Physics, made possible by high-end computing, is demonstrating the crucial role of eddies in controlling ocean climate; and how the probability of extreme weather events may respond to rising greenhouse gas concentrations. He concluded by throwing out a more controversial suggestion that super-computers haven’t really contributed very much to the problem of predicting century-timescale changes in global average temperature, however much they may have contributed to understanding the regional implications of large-scale warming.

Prof Mark Newman

Structure in complex systems

Video Podcast Presentation (PDF)

In physics, "complex systems" are systems of many similar interacting parts, such as the interacting atoms that make up a solid or liquid, but also interacting organisms in an ecosystem, or interacting traders in the stock market. This lecture will discuss how recent advances in modeling and computer simulation have allowed us to apply physics-style approaches to these previously challenging real-world systems to learn about such things as the spread of diseases, the flow of traffic or the structure of entire human societies.

Dr Thorsten Wahl

Breaking through the quantum barrier

Video Podcast Presentation (PDF)

Many exciting phenomena observed in condensed matter systems, such as superconductivity and the quantum Hall effect, emerge due to the quantum mechanical interplay of many electrons. The laws of quantum physics are governed by the Schrödinger equation, whose complexity grows exponentially with the number of particles it describes. Hence, even an approximate numerical solution of the Schrödinger equation is impossible for only just a few particles, not to mention for the millions of particles that are present in real materials. This talk focuses on a new approximation scheme in terms of so-called Tensor Network States, which allow for an arbitrarily accurate description of realistic quantum solid state systems at merely a polynomial overhead in the particle number, thus enabling efficient simulations of such systems on today's computers.