engleski

Assimilation system at DHMZ

Numerical weather prediction (NWP) can be seen as an initial problem in mathematics where, if the initial state of the atmosphere at a given time is known, geophysical systems equations can be solved to obtain values for variables at future time points. Sensitivity to initial conditions can be even greater if the nonlinearity of a geophysical system is taken into account, which requires “best possible” initial conditions for the NWP model. Data assimilation is technique of approximating “true” state of atmosphere at given time taking into account dynamical properties of the physical system. A description of the setup for a local assimilation system for a limited area model, ALADIN (Aire Limiteé Adaptation Dynamique dévelopement InterNational) is given. The assimilation system at DHMZ (Meteorological and Hydrological Service of Croatia) consists of two parts: the surface assimilation, which is used to change the state of a model land surface variables, and the upper air assimilation, which changes the upper air model fields. The surface assimilation is performed by the optimal interpolation (OI) technique, while the upper air assimilation is conducted using the 3D variational technique (3DVAR). A basic verification was performed for a forecast starting from the initial state given by the assimilation system and the operational forecast. The verification results showed a positive impact of assimilation on forecast for the upper airfields and for screen- level variables.

data assimilation; surface parameters; operational forecast

nije evidentirano