On four‐dimensional variational assimilation of ozone data in weather‐prediction models

Abstract

AbstractRoutinely made satellite observations of total ozone contain information about the dynamical state of the atmosphere in the regions where the ozone gradients are strong. It is argued that four‐dimensional variational assimilation (4D‐VAR) of these data with a meteorological model may be a way of extracting this information in order to use it for determining the initial condition of the model. In the first part of this paper the problem of assimilating ozone data in a variational context is discussed. In the second part a series of simple experiments is presented in which ozone pseudo‐observations are assimilated with a barotropic vorticity‐equation model. The aim of these experiments is to examine whether the atmospheric flow at a given level near the tropopause can be reconstructed by a 4D‐VAR scheme using a series of simulated observations of passive tracer mixing ratios. Both data produced by the assimilating model itself and data generated by a more realistic three‐dimensional model have been tested. It is found that when started from a purely zonal flow field, 4D‐VAR can to a large extent reconstruct the flow field, using only observations of the mixing ratio of the tracer. The quality of the result depends on the resolution of the model and on the choice of the length of the assimilation window.