TOTAL OZONE VARIATIONS IN THE TROPICAL BELT - AN APPLICATION FOR QUALITY OF GROUND-BASED MEASUREMENTS

The study of the regime of ozone variations in the huge tropical belt( 25 degrees S to 25 degrees N), which are, in general, very small and z onally nearly symmetric, permits to establish a statistical model for estimating the ozone deviations using Total Ozone Mapping Spectrometer (TOMS) data. ? he equatorial stratospheric winds at 25 and 50 hPa and the solar flux at 10.7 cm are used as major predictors and the linear trend was also estimated. The 10 m/sec stratospheric wind change is r elated to similar to 1.2% ozone change at the equator, to practically no change in the 8-15 degrees belts and up to 1.4% change with opposit e phase over the tropics in spring but nearly zero change in fall. The solar cycle related amplitude is about 1.4% per 100 units of 10.7 cm solar flux. The ozone trends are negative: not significant over the eq uator and about -2% per decade (significant at 95% level) over the tro pics. The latter could have been enforced by the 2 to 4% lower ozone v alues during 1991-1993, part of which might be related to the effects of the Mt. Pinatubo eruption, but might also be due to the strong QBO. The estimated deviations are verified versus reliable observations an d the very good agreement permits applying the model for quantitative quality control of the reported ozone data from previous years. The st andard deviation of the difference between observed ozone deviations a nd those estimated from the model is only 0.9-1.6% for yearly mean, th at means instruments used for total ozone observations in the tropical belt should have systematic error of less than 1%. Cases when the dis crepancies between the model and reported observations at a given stat ion exceed 2-3% for time interval of 2 or more years should be verifie d.