Synergy in Computational Astrophysics

A common hardware platform for solar physics and cosmology was built by purchasing computing equipment for about 2 million NOK in 2013 and 2014. This was integrated with computing infrastructure purchased both with own funding and funded from other externally funded projects to become a High Performance Computing cluster which has been instrumental for all computational research at the Institute of Theoretical Astrophysics, in cosmology as well as in solar physics. This common hardware platform has been a great success. As this hardware platform has been utilized in basically all research in solar physics and cosmology from 2013, we can here only single out a few of huge number of results: The solar physics group has used the hardware platform to develop the necessary tools and codes to treat 3D radiation-magneto-hydrodynamics problems with a realistic treatment of the radiation. It has worked intensively with IRIS data and the comparison between observations and synthetic observations. Five papers with co-authors from the solar physics group were published in the October 17 2015 edition of Science, including the front cover. A series of eight papers on the formation of IRIS diagnostics were published in ApJ 2013 - 2015. An important utilization of the hardware infrastructure by the cosmology group has been development of an independent N-body code which is the first which can treat a dynamic and inhomogeneous scalar field that couples to matter. Until now, the community has made use of numerical simulations in which the quasistatic limit was assumed when solving the equation of motion of the scalar field. We developed a method to solve the full equations of motion for scalar degrees of freedom coupled to matter and ran cosmological simulations which track the full time and space evolution of the scalar field, finding striking differences with respect to the commonly used quasistatic approximation. The results so far have been published in two papers in Physical Review Letters and several papers in Physical Review D, Astronomy & Astrophysics and MNRAS.

The application is for funding a project that through a strengthening of computational astrophysics will give closer cooperation between the Norwegian solar physics and cosmology communities, from which new research fields will emerge in the future. By bu ilding up special expertise on numerical simulations and analysis of huge data sets, cutting across the research groups, both the solar physics and cosmology communities will be able to develop further with major synergy between them. We will form a comm on platform for computational methodology at the Institute of Theoretical Astrophysics, University of Oslo. This platform will consist of one postdoctoral research fellow or researcher and one PhD student, and necessary computational hardware for analysis and visualisation of large datacubes from numerical simulations and analyses performed on national High Perofrmance Computing clusters. This platform will be part of both the solar physics and cosmology research groups. A major goal of this proposal is to develop computational algorithms for general use in astrophysical research, but most importantly to aid the specific work of the two groups of ITA. Development of new software will be the major part of the work. Examples of relevant algorithms are mul tigrid PDE solvers, efficient spherical harmonics transforms, development of accurate Bayesian MCMC techniques and visualization algorithms for large datacubes.