4D-Var for numerical weather prediction
Bonavita, Massimo ; Lean, Peter
At the European Centre for Medium-Range Weather Forecasts (ECMWF) and in most other global forecasting centres, the methodology used to produce atmospheric analyses is based on a technique called 4-dimensional variational data assimilation (4D-Var). 4D-Var aims to estimate a model trajectory that is as consistent as possible with two sources of information: a prior estimate obtained from a short-range forecast from the previous analysis (the 'background') and all the observations available over the length of the assimilation window (typically 6-12h). 4D-Var was introduced in the mid-1980s in the meteorological community by Dimet and Talagrand (1986) but had its origins in earlier work in the mathematical community, developing optimal control theory (Lions, 1971; Marchuk, 1974).</p>
Accès à la notice sur le site du portail documentaire de Météo-France