Dr Francesco Miniati

Abstract:

We present a machine learning based approach to address the study of transport processes, ubiquitous in continuous mechanics, with particular attention to those phenomena ruled by complex micro-physics, impractical to theoretical investigation, yet exhibiting emergent behavior describable by a closed mathematical expression. Our machine learning model, built using simple components and following a few well established practices, is capable of learning latent representations of the transport process substantially closer to the ground truth than expected from the nominal error characterising the data, leading to sound generalisation properties. This is demonstrated through an idealized study of the long standing problem of heat flux suppression under conditions relevant for fusion and cosmic plasmas. A simple analysis shows that the result applies beyond those case specific assumptions and that, in particular, the accuracy of the learned representation is controllable through knowledge of the data quality (error properties) and a suitable choice of the dataset size. While the learned representation can be used as a plug-in for numerical modeling purposes, it can also be leveraged with the above error analysis to obtain reliable mathematical expressions describing the transport mechanism and of great theoretical value.

Abstract:

The interplay between charged particles and turbulent magnetic fields is crucial to understanding how cosmic rays propagate through space. A key parameter which controls this interplay is the ratio of the particle gyroradius to the correlation length of the magnetic turbulence. For the vast majority of cosmic rays detected at the Earth, this parameter is small, and the particles are well confined by the Galactic magnetic field. But for cosmic rays more energetic than about 30 EeV, this parameter is large. These highest energy particles are not confined to the Milky Way and are presumed to be extragalactic in origin. Identifying their sources requires understanding how they are deflected by the intergalactic magnetic field, which appears to be weak, turbulent with an unknown correlation length, and possibly spatially intermittent. This is particularly relevant given the recent detection by the Pierre Auger Observatory of a significant dipole anisotropy in the arrival directions of cosmic rays of energy above 8 EeV. Here we report measurements of energetic-particle propagation through a random magnetic field in a laser-produced plasma. We characterize the diffusive transport of these particles and recover experimentally pitch-angle scattering measurements and extrapolate to find their mean free path and the associated diffusion coefficient, which show scaling-relations consistent with theoretical studies. This experiment validates these theoretical tools for analyzing the propagation of ultra-high energy cosmic rays through the intergalactic medium.

Abstract:

Galaxy clusters are filled with hot, diffuse X-ray emitting plasma, with a stochastically tangled magnetic field whose energy is close to equipartition with the energy of the turbulent motions \cite{zweibel1997, Vacca}. In the cluster cores, the temperatures remain anomalously high compared to what might be expected considering that the radiative cooling time is short relative to the Hubble time \cite{cowie1977,fabian1994}. While feedback from the central active galactic nuclei (AGN) \cite{fabian2012,birzan2012,churazov2000} is believed to provide most of the heating, there has been a long debate as to whether conduction of heat from the bulk to the core can help the core to reach the observed temperatures \cite{narayan2001,ruszkowski2002,kunz2011}, given the presence of tangled magnetic fields. Interestingly, evidence of very sharp temperature gradients in structures like cold fronts implies a high degree of suppression of thermal conduction \cite{markevitch2007}. To address the problem of thermal conduction in a magnetized and turbulent plasma, we have created a replica of such a system in a laser laboratory experiment. Our data show a reduction of local heat transport by two orders of magnitude or more, leading to strong temperature variations on small spatial scales, as is seen in cluster plasmas \cite{markevitch2003}.