Numerical weather prediction (NWP) models use meteorological data from various
sources. Especially remote-sensing data are required because of their high spatial and time
resolution as well as their availability. The problem is with assimilation of the data to
operational NWP models. Increasing resolution of NWP models require correspondingly high
resolution wind and moisture data for initialisation. NWP data assimilation techniques have
advanced considerably in 1990s and 2000s, with the arrival of techniques capable of
extracting information from time sequences of observations (see the reviews of MacPherson
et al. 2004; Sun 2005). Development of radar networking and processing around Europe
makes feasible the routine operational delivery of radar information.

More (in pdf)