Sebastian von Alfthan

Ion distributions in the Earth's foreshock: Hybrid-Vlasov simulation and THEMIS observations

We present the ion distribution functions in the ion foreshock upstream of the terrestrial bow shock obtained with Vlasiator, a new hybrid-Vlasov simulation geared toward large-scale simulations of the Earths magnetosphere (http://vlasiator.fmi.fi). They are compared with the distribution functions measured by the multispacecraft Time History of Events and Macroscale Interactions during Substorms (THEMIS) mission. The known types of ion distributions in the foreshock are well reproduced by the hybrid-Vlasov model. We show that Vlasiator reproduces the decrease of the backstreaming beam speed with increasing distance from the foreshock edge, as well as the beam speed increase and density decrease with increasing radial distance from the bow shock, which have been reported before and are visible in the THEMIS data presented here. We also discuss the process by which wave-particle interactions cause intermediate foreshock distributions to lose their gyrotropy. This paper demonstrates the strength of the hybrid-Vlasov approach which lies in producing uniformly sampled ion distribution functions with good resolution in velocity space, at every spatial grid point of the simulation and at any instant. The limitations of the hybrid-Vlasov approach are also discussed.

Reconnection rates and X line motion at the magnetopause: Global 2D‐3V hybrid‐Vlasov simulation results

We present results from a first study of the local reconnection rate and reconnection site motion in a 2D-3V global magnetospheric self-consistent hybrid-Vlasov simulation with due southward interplanetary magnetic field. We observe magnetic reconnection at multiple locations at the dayside magnetopause and the existence of magnetic islands, which are the 2-D representations of flux transfer events. The reconnection locations (the X lines) propagate over significant distances along the magnetopause, and reconnection does not reach a steady state. We calculate the reconnection rate at the location of the X lines and find a good correlation with an analytical model of local 2-D asymmetric reconnection. We find that despite the solar wind conditions being constant, the reconnection rate and location of the X lines are highly variable. These variations are caused by magnetosheath fluctuations, the effects of neighboring X lines, and the motion of passing magnetic islands.

Mirror modes in the Earth's magnetosheath: Results from a global hybrid‐Vlasov simulation

We investigate mirror mode structures in the Earths magnetosheath using our global kinetic Vlasiator simulation, which models ion behavior through their distribution function and treats electrons as a charge-neutralizing fluid. We follow the evolution of waves as they advect along velocity streamlines through the magnetosheath. We find that mirror mode waves are observed preferentially in the quasi-perpendicular magnetosheath along velocity streamlines that enter the sheath in the vicinity of the foreshock ULF wave boundary where there are enough initial perturbations in the plasma for the mirror modes to grow, and the plasma properties fulfill the mirror instability condition better than elsewhere in the magnetosheath. We test selection criteria defined by previous studies and show that the spatial extent over which mirror modes occur ranges from much of the magnetosheath on the quasi-perpendicular side of the subsolar point to very small isolated regions depending on the criteria.

Wave dispersion in the hybrid-Vlasov model: Verification of Vlasiator

Vlasiator is a new hybrid-Vlasov plasma simulation code aimed at simulating the entire magnetosphere of the Earth. The code treats ions (protons) kinetically through Vlasovs equation in the six-dimensional phase space while electrons are a massless charge-neutralizing fluid [M. Palmroth et al., J. Atmos. Sol.-Terr. Phys. 99, 41 (2013); A. Sandroos et al., Parallel Comput. 39, 306 (2013)]. For first global simulations of the magnetosphere, it is critical to verify and validate the model by established methods. Here, as part of the verification of Vlasiator, we characterize the low-β plasma wave modes described by this model and compare with the solution computed by the Waves in Homogeneous, Anisotropic Multicomponent Plasmas (WHAMP) code [K. Ronnmark, Kiruna Geophysical Institute Reports No. 179, 1982], using dispersion curves and surfaces produced with both programs. The match between the two fundamentally different approaches is excellent in the low-frequency, long wavelength range which is of interest ...

On the importance of spatial and velocity resolution in the hybrid-Vlasov modeling of collisionless shocks

In hybrid-Vlasov plasma modeling, the ion velocity distribution function is propagated using the Vlasov equation while electrons are considered a charge-neutralizing fluid. It is an alternative to particle-in-cell methods, one advantage being the absence of sampling noise in the moments of the distribution. However, the discretization requirements in up to six dimensions (3D position, 3V velocity) make the computational cost of hybrid-Vlasov models higher. This is why hybrid-Vlasov modeling has only recently become more popular and available to model large-scale systems. The hybrid-Vlasov model Vlasiator is the first to have been successfully applied to model the solar-terrestrial interaction. It includes in particular the bow shock and magnetosheath regions, albeit in 2D-3V configurations so far. The purpose of this study is to investigate how Vlasiator parameters affect the modeling of a plasma shock in a 1D-3V simulation. The setup is similar to the Earths bow shock in previous simulations, so that the present results can be related to existing and future magnetospheric simulations. The parameters investigated are the spatial and velocity resolution, as well as the phase space density threshold, which is the key parameter of the so-called sparse velocity space. The role of the Hall term in Ohms law is also studied. The evaluation metrics used are the convergence of the final state, the complexity of spatial profiles and ion distributions as well as the position of the shock front. In agreement with previous Vlasiator studies it is not necessary to resolve the ion inertial length and gyroradius in order to obtain kinetic phenomena. While the code remains numerically stable with all combinations of resolutions, it is shown that significantly increasing the resolution in one space but not the other leads to unphysical results. Past a certain level, decreasing the phase space density threshold bears a large computational weight without clear physical improvement in the setup used here. Finally, the inclusion of the Hall term shows only minor effects in this study, mostly because of the 1D configuration and the scales studied, at which the Hall term is not expected to play a major role.

Topology aware process mapping

A parallel program based on the Message Passing Interface (MPI) commonly uses point-to-point communication for updating data between processes, and its scalability is ultimately limited by communication costs. To minimize these costs we have developed a library that reduces network congestion, and thus improves performance, by optimizing the placement of processes onto nodes allocated to the parallel job. Our approach is useful on production machines, as irregular communication patterns can at run-time be optimally placed on non-contiguous node allocations. It is also portable as it supports multiple architectures: Cray XT, IBM BlueGene/P and regular SMP clusters. We demonstrate on a Cray XT5m and an Infiniband cluster that good placement of processes doubles the total bandwidth compared to random placement and, furthermore, by up to a factor of 1.4 compared to to the original placement. It is not only important to place processes well on individual nodes, minimizing the number of link traversals on the Cray XT5m provides up to 20 % of additional performance. The scalability of a real-world application, Vlasiator, is also investigated and the scalability is shown to improve by up to 35 %. For communication limited applications the approach provides an avenue to improve performance, and is useful even with dynamic load balancing as the placement is optimized at run-time.

Myosin motor mediated contraction is enough to produce cytokinesis in the absence of polymerisation

Actin filament networks play an active role in cytokinesis of eukaryotic cells. These networks, linked mainly by myosin, are concentrated below the cell membrane forming a spherical supporting shell. During cytokinesis, this network is modified such that a contractile ring is formed along the diameter of the shell. We present a realistic three-dimensional simulation model to study the dynamics of this spherical shell of elastic actin filaments and myosin motors. The results show compelling evidence that this fibre-spring model, with the motors activated in a narrow region around the division plane, is sufficient to reproduce most of the essential mechanics of cytokinesis: A spontaneous formation of a contractile ring, a characteristic filament orientation structure, and realistic cleavage furrow dynamics. These results demonstrate that, though cytokinesis is a highly complex process with large variation in intricate details, the fundamental dynamics are largely generic. In particular, motor mediated contraction of an unstructured filament mesh is sufficient to undergo division without concentrated and directed polymerization in the cleavage zone.

Vlasov methods in space physics and astrophysics

This paper reviews Vlasov-based numerical methods used to model plasma in space physics and astrophysics. Plasma consists of collectively behaving charged particles that form the major part of baryonic matter in the Universe. Many concepts ranging from our own planetary environment to the Solar system and beyond can be understood in terms of kinetic plasma physics, represented by the Vlasov equation. We introduce the physical basis for the Vlasov system, and then outline the associated numerical methods that are typically used. A particular application of the Vlasov system is Vlasiator, the world’s first global hybrid-Vlasov simulation for the Earth’s magnetic domain, the magnetosphere. We introduce the design strategies for Vlasiator and outline its numerical concepts ranging from solvers to coupling schemes. We review Vlasiator’s parallelisation methods and introduce the used high-performance computing (HPC) techniques. A short review of verification, validation and physical results is included. The purpose of the paper is to present the Vlasov system and introduce an example implementation, and to illustrate that even with massive computational challenges, an accurate description of physics can be rewarding in itself and significantly advance our understanding. Upcoming supercomputing resources are making similar efforts feasible in other fields as well, making our design options relevant for others facing similar challenges.

Investigation of load balancing scalability in space plasma simulations

In this study we report the load-balancing performance issues that are observed during the petascaling of a space plasma simulation code developed at the Finnish Meteorological Institute (FMI). The code models the communication pattern as a hypergraph, and partitions the computational grid using the parallel hypergraph partitioning scheme (PHG) of the Zoltan partitioning framework. The result of partitioning determines the distribution of grid cells to processors. It is observed that the initial partitioning and data distribution phases take a substantial percentage of the overall computation time. Alternative (graph-partitioning-based) schemes that provide better balance are investigated. Comparisons in terms of effect on running time and load-balancing quality are presented. Test results on Juelich BlueGene/P cluster are reported.

Analyzing the power consumption behavior of a large scale data center

The aim of this paper is to illustrate the use of application and system level logs to better understand scientific data center behavior and energy-spending. Analyzing a data center log of 900 nodes (Sandy Bridge and Haswell), we study node power consumption and describe approaches to estimate and forecast it. Our results include methods to cluster nodes based on different vmstat and RAPL measurements as well as Gaussian and GAM models for estimating the plug power consumption. We also analyze failed jobs and find that non-successfully terminated jobs consume around 40% of computing time. While the actual numbers are likely to vary in different data centers at different times, the purpose of the paper is to share ideas of what can be found by statistical and machine learning analysis of large amount of log data.