Ozone Distributions Derived from Satellite Data Assimilation and Their Impact on the Stratospheric Temperatures

Using a linear version of the ECMWF radiative transfer code it is possible to evaluate the impact of a prescribed ozone climatology on the stratospheric temperatures produced by climate models. So far referenced ozone climatologies have been derived by compilations and simple spatial interpolations of satellite and soundings data with rather coarse vertical resolutions and sometimes inhomogeneous sampling time coverage. It is therefore interesting to evaluate the effects of using homogeneous fields obtained from ozone data assimilation of recently available satellite missions. In this study we use a 3D-VAR data assimilation technique to construct 4-dimensional ozone fields. The methodology combines a chemistry transport model forced by the ECMWF temperature and wind field analyses, and the PALM software to drive the overall assimilation suite. Two different sources of ozone data, from the ENVISAT/MIPAS and from the SMR/ODIN instruments, are used for data assimilation during the period July to November 2003. Each dataset is assimilated separately and monthly mean ozone fields are used for the radiative calculations. Differences in the ozone distributions are discussed and bias between the instruments evaluated. The influence of using the different ozone fields within the LRM model are analysed in terms of sensitivity of the stratospheric temperatures to the spatial distribution of the ozone changes. The relative role of IR and visible radiative transfer is evaluated. It is concluded that the ozone distribution plays an important role in modelling the middle atmosphere temperatures, especially in the lower stratosphere where the radiative relaxation times are long.