Cloud processes are appreciated to be of increasing importance to the comprehension of the atmosphere. Therefore, weather models have recently been extended by detailed spectral descriptions of cloud processes. However, the high computational costs hinder their practical application. This paper introduces the novel framework FD4 (Four-Dimensional Distributed Dynamic Data structures), which is developed to parallelize and couple cloud models to atmospheric models in an efficient way and to enable a higher scalability on HPC systems. Results of first tests with the regional forecast model COSMO are presented.
|Title of host publication||Parallel Computing: From Multicores and GPU's to Petascale|
|Editors||Barbara Chapman, Frédéric Desprez, Gerhard R. Joubert, Alain Lichnewsky, Frans J. Peters, Thierry Priol|
|Publisher||IOS Press, Amsterdam [u. a.]|
|Number of pages||8|
|Publication status||Published - 2010|
|Series||Advances in Parallel Computing|
- atmospheric modeling, dynamic load balancing, high performance computing, model coupling