Variational data assimilation

In variational assmilation (VAR) techniques, the best estimate of the atmospheric state - called the analysis - is determined by variationally minimising the mismatch between the observations, on the one hand, and a first guess - the background - on the other.

There are three main flavours of VAR that are used for weather forecasting. 4d-VAR minimises the mismatch using the constraint of the atmosphere's equations of motion. 3d-VAR is a simplification which is not directly constrained by the equations of motion, but is otherwise similar to 4d-VAR. 1d-VAR is a further simplification that deals with individual vertical columns of the atmosphere. It is often used to help interpret measurements from satellites.

The variational technique is extremely powerful and it is the technique of choice in most European weather prediction centres.


4D-Var in high-resolution versions of the UM M. Dixon
Forecast errors for convective scale variational data assimilation R. Bannister, M. Dixon
Adaptive boundary layer data assimilation A. Fowler, R. Bannister,
S. Migliorini

You are visitor number [an error occurred while processing this directive] to this page.


Page navigation