Abstract

Data assimilation is a very complicated cycle in numerical weather prediction and influences the quality of forecast. In this technique, the observered meteorological data are combined with the results of previous short-range forecast of the model to make an initial condition for a new prediction. This paper presents: four-dimensional variational assimilation (4D-Var), the most advanced technique for the data assimilation; the application of data assimilation in numerical weather prediction in Vietnam at present and the development of this technique in future.