Premium
Some numerical properties of approaches to physics–dynamics coupling for NWP
Author(s) -
Dubal Mark,
Wood Nigel,
Staniforth Andrew
Publication year - 2006
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1256/qj.05.49
Subject(s) - coupling (piping) , numerical weather prediction , stability (learning theory) , mathematics , statistical physics , computer science , numerical stability , steady state (chemistry) , numerical analysis , mathematical optimization , meteorology , physics , mathematical analysis , engineering , mechanical engineering , chemistry , machine learning
At the present time there exist a number of different approaches to the problem of coupling parametrized physical processes to the dynamical core in operational numerical weather‐prediction (NWP) and climate models. Motivated by the various strategies in use, some idealized representative coupling schemes are constructed and subsequently analysed using a methodology in which the physics and dynamics terms are represented in a simplified way. Particular numerical properties of the idealized schemes which are of interest are the ability to capture correct steady‐state solutions and to be second‐order accurate in time. In general, the schemes require specific choices for the time‐differencing of certain coupled processes if correct steady‐state solutions are to be obtained. This has implications for the overall numerical stability of a coupling strategy. An alternative physics–dynamics coupling approach is then described and analysed. A multiple‐sweep predictor–corrector coupling scheme is shown to capture the correct steady‐state solution and to allow for second‐order accuracy, provided that the convective process is coupled explicitly. This approach has a number of advantages over those currently used in operational NWP models. Copyright © 2006 Crown copyright