Sensitivity Studies for Assimilated Ozone Profiles

An ozone data assimilation system at the NASA/Goddard Data Assimilation Office (DAO) produces three-dimensional global ozone fields. They are obtained by assimilating ozone retrieved from the Solar Backscatter UltraViolet/2 (SBUV/2) instrument and the Earth Probe Total Ozone Mapping Spectrometer (EP TOMS) measurements into an off-line parameterized chemistry and transport model. In this talk we focus on the quality of lower stratospheric assimilated ozone profiles. Ozone in the lower stratosphere plays a key role in the forcing of climate. A biased ozone field in this region will adversely impact calculations of the stratosphere-troposphere exchange and, when used as a first guess in retrievals, the values determined from satellite observations. The SBUV/2 ozone data have a coarse vertical resolution with increased uncertainty below the ozone maximum, and TOMS provides only total ozone columns. Thus, the assimilated ozone profiles in the lower stratosphere are only weakly constrained by the incoming SBUV and TOMS data. Consequently, the assimilated ozone distribution should be sensitive to changes in inputs to the statistical analysis scheme. We investigate the sensitivity of assimilated ozone profiles to changes in a variety of system inputs: TOMS and SBUV/2 data selection, forecast and observations error covariance models, inclusion or omission of a parameterized chemistry model, and different versions of DAO assimilated wind fields used to drive the transport model. Comparisons of assimilated ozone fields with independent observations, primarily ozone sondes, are used to determine the impact of each of these changes.