Dynamical impact of total‐ozone observations in a four‐dimensional variational assimilation

Observing-system simulation experiments (OSSEs) are performed to investigate the dynamical impact of total-ozone data used in a four-dimensional variational analysis in a numerical weather prediction (NWP) model. Ozone columns are generated by the model at the location of data retrieved by the National Environmental Satellite, Data, and Information Service, from the measurements of the TIROS (Television Infra-Red Observation Satellite) Operational Vertical Sounder (TOVS). They are assimilated over a 12-hour period. In the present experiments, ozone mixing ratio is considered to be a passive tracer, simply advected by winds. The initial three-dimensional ozone field is not analysed. An important positive impact is found in a fully idealized experiment, on the flow analysis as well as on the temperature analysis. The ability to derive thermodynamic information from ozone-column observations is confirmed by an OSSE performed in a more realistic context. Moreover, the information on wind fields supplied by ozone data may be in addition to the information already provided by observations currently assimilated in NWP models. Some conditions, needed to make the information provided by ozone columns effective, are discussed, such as the accuracy of the ozone measurements and the necessity of analysing ozone in addition to the usual model variables.