Ten organising principles for coupling in multiphysics and multiscale models

Abstract

Computational science faces new challenges posed by multiphysics and multiscale, or more generally put, coupled models. These systems are composites formed from separate subsystem models that interact via data exchanges. These data dependencies pose a coupling problem, and on distributed memory computers, a parallel coupling problem. This article presents a definition of terms and a set of organising principles for the coupling and parallel coupling problems. It is meant as a first step towards creating a theory of coupled models. These principles are then employed in a case study of a coupled climate model and offer remarkable insight into its structure.

PyCCSM: Prototyping a python-based community climate system model

“…Numerous Python interfaces to mpi exist, including pympi, 7 pypar, 8 Scientific Python, 9 and mympi. 10 For ccsm, we must support a multiexecutable application that bridges both Fortran and Python.…”

“…Message passing introduces a new obstacle called the parallel coupling problem (pcp) [9]: Given N multiple, separately developed, message passing parallel models, combine them into an efficient parallel coupled model. The pcp is the central computational science challenge faced by developers of coupled climate, multiphysics, and multiscale models.…”

“…mct is parallel coupling middleware that eases the programming of parallel couplings in mpi-based applications. mct addresses the parallel data processing part of the pcp [9], maximising developer flexibility in choices regarding parallel coupled model architecture. mct provides classes and methods and a library of routines to effect distributed data description and parallel data transfer and transformation.…”

Graphical Notation for Diagramming Coupled Systems

“…Multiscale models capture spatiotemporal-scale interactions by coupling separate models that operate on disparate time and length scales. Multiphysics and multiscale models share a common requirement for data exchange mechanisms that allow their constituent subsystems to compute their respective states-the coupling problem (cp) [1]. In many cases, coupled systems contain constituent subsystems possessing high levels of computational complexity that warrant parallel processing.…”

“…In many cases, coupled systems contain constituent subsystems possessing high levels of computational complexity that warrant parallel processing. Coupling in distributed-memory parallel environments is particularly difficult-the parallel coupling problem (pcp) [1].…”