Main challenges of to-day's numerical weather prediction (NWP) and their numerical implications
Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique42, Avenue Gaspard Coriolis
31057 Toulouse Cedex
Tel : 05 61 19 31 31 - Fax : 05 61 19 30 00
MAIN CHALLENGES OF TO-DAY'S NUMERICAL WEATHER PREDICTION (NWP) AND THEIR NUMERICAL IMPLICATIONS
Jean-Francois Geleyn
(Meteo-France, CNRM/GMAP, Toulouse)
Tuesday April 28, 11.00 a.m. CERFACS Conference Room
While never being in the forefront of research in numerics in any individual topic, NWP covers a large spectrum of methods at the limit between physics and numerical analysis. One can indeed distinguish, inside any modern NWP operational suite, several independent ways to treat the scientific chain of actions: "observe, understand, model, forecast". The subject of this conference will be a rapid review of these methods topic by topic for the areas where some progress is foreseen for the few coming years and using mainly as examples the two operational applications in Toulouse: ARPEGE-(IFS) and ALADIN. The focus points will be:
variational data assimilation (building of the cost function, data selection, filtering properties, ...);
predicability studies and their link with the previous topic (singular vectors, Kalman filtering, ...);
model dynamics (semi-Lagrangian methods, non-hydrostatic algorithms, revisit of some simplifying hypothesis, ...);
parameterisation of physical processes (problems in the interaction with the adiabatic part of the time-step, fibrillations, ...).
This review will offer the occasion to identify the specificities that make NWP a technical endeavour without too many common points with other neighbouring scientific applications.
Cerfacs' Conferences 1998 Home Page
