A PRELIMINARY-STUDY OF THE IMPACT OF THE ERS-1 C-BAND SCATTEROMETER WIND DATA ON THE EUROPEAN-CENTER-FOR-MEDIUM-RANGE-WEATHER-FORECASTS GLOBAL DATA ASSIMILATION SYSTEM

A preliminary assessment of the impact of the ERS 1 scatterometer wind data on the current European Centre for Medium-Range Weather Forecast s analysis and forecast system has been carried out. Although the scat terometer data results in changes to the analyses and forecasts, there is no consistent improvement or degradation. In this sense, the impac t of the data is neutral. Our results are based on comparing analyses and forecasts from assimilation cycles which are identical in all resp ects except that the control experiment uses no scatterometer data. Th e experiment period runs from 0000 UT, December 26, 1991, through Janu ary 2, 1992. The two sets of analyses are very similar except for the low level wind fields over the ocean. Impacts on the analyzed wind fie lds are greater over the southern ocean, where other data are scarce. For the most part the mass field increments are too small to balance t he wind increments. The effect of the nonlinear normal mode initializa tion on the analysis differences is quite small, but we observe that t he differences tend to wash out in the subsequent 6-hour forecast. In the northern hemisphere, analysis differences are very small, except d irectly at the scatterometer locations. Forecast comparisons reveal la rge differences in the southern hemisphere after 72 hours. Notable dif ferences in the northern hemisphere do not appear until late in the fo recast. Overall, however, the southern hemisphere impacts are neutral. The experiments described are preliminary in several respects. In lig ht of expected refinements to the scatterometer model function, the am biguity removal algorithm and the analysis procedures, we expect these data to ultimately prove useful for global data assimilation.