Comoving-frame radiative transfer in arbitrary velocity fields -- II. Large scale applications

TL;DR

This paper presents a scalable, memory-efficient, and parallelizable algorithm for large-scale radiative transfer problems in arbitrary velocity fields, enabling multi-dimensional applications with improved speed.

Contribution

It introduces a domain decomposition and iterative solvers to handle large problems, along with a quasi-analytic solution that accelerates computations.

Findings

Memory footprint is significantly reduced.

Algorithm achieves faster computation times.

Applicable to multi-dimensional, chaotic velocity fields.

Abstract

A solution of the radiative-transfer problem in arbitrary velocity fields introduced in a previous paper, has limitations in its applicability. For large-scale applications, the methods described also require large memory sets that are commonly not available to state-of-the-art computing hardware. In this work, we modify the algorithm to allow the computation of large-scale problems. We reduce the memory footprint via a domain decomposition. By introducing iterative Gauss-Seidel type solvers, we improve the speed of the overall computation. Because of the domain decomposition, the new algorithm requires the use of parallel-computing systems. The algorithm that we present permits large-scale solutions of radiative-transfer problems that include arbitrary wavelength couplings. In addition, we discover a quasi-analytic formal solution of the radiative transfer that significantly improves the overall computation speed. More importantly, this method ensures that our algorithm can be applied to multi-dimensional Lagrangian radiative-transfer calculations. In multi-dimensional atmospheres, velocity fields are in general chaotic ensuring that the inclusion of arbitrary wavelength couplings are mandatory.