The GEOS ozone data assimilation system: Specification of error statistics

A global three-dimensional ozone data assimilation system has been develope d at the Data Assimilation Office of the NASA Goddard Space Flight Center. The Total Ozone Mapping Spectrometer (TOMS) total ozone data and the Solar Backscatter UItraviolet/2 (SBUV/2) partial ozone profile observations are a ssimilated. The assimilation. into an off-lint ozone transport model, is do ne using the global Physical-space statistical Analysis Scheme. This system became operational in December 1999. A detailed description of the statistical analysis scheme and, in particula r, of the forecast- and observation-error covariance models is given. A new global anisotropic horizontal forecast-error correlation model accounts fo r a varying distribution of observations with latitude. Correlations are la rgest in the zonal direction in the tropics where data are sparse. Forecast -error variance is assumed to be proportional to the ozone field. The forec ast-error covariance parameters were determined by maximum-likelihood estim ation. The error covariance models are validated using chi (2) statistics. The analysed ozone fields in the winter 1992 are validated against independ ent observations from ozone sondes and the Halogen Occultation Experiment ( HALOE). The difference between the mean HALOE observations and the analysis fields is less than 10% at pressure levels between 70 and 0.2 hPa. The glo bal root-mean-square difference between TOMS observed and forecast values i s less than 4%. The global root-mean-square difference between SBUV observe d and analysed ozone between 50 and 3 hPa is less than 15%.