An Assessment of the FGGE Satellite Observing System during SOP-1

This study investigates the degree to which data from the space-borne FGGE observing systems are able to determine the complete state of the atmosphere when incorporated into a global objective analysis cycle. Three data assimilation experiments are performed with the Goddard Laboratory for Atmospheric Sciences (GLAS) analysis/forecast system, using different combinations of the FGGE level II–b data collected during the first Special Observing Period (SOP-1), 5 January through 5 March 1979. The control experiment is an assimilation cycle with the complete FGGE II–b data. The other two assimilation/forecast experiments consist of i) the conventional system without the satellite data and special FGGE data sets; and ii) the FGGE II–b surface and satellite temperature soundings and cloud-track winds, aircraft data, and special FGGE data sets, but without the conventional rawinsonde/pilot balloon network. From these experiments, we attempt to assess the accuracy of the inferred mass and motion fields over data-sparse regions, by examining their influence on analyses and forecasts over data-rich regions. The sensitivity of the analysis to the FGGE satellite data is shown by comparisons of the 6 h forecast error of the 300 mb geopotential height fields for these three experiments. It is found that large 6 h forecast errors downstream of data-sparse regions are reduced when the satellite observations are incorporated in the analysis. Forecast impact results from the initial states of these assimilation cycles show the geographical influence of the FGGE satellite observing system on short- to medium-range (two to five days) weather forecasting. Over North America and Europe, there is a small improvement in forecast skill from the use of the FGGE II–b data. Over Australia, as expected, the positive impact of satellite data is much larger. The number of skillful four- and five-day forecasts over North America and Europe has been increased substantially by the addition of the FGGE II–b data. Examples of useful eight-day forecasts, which occurred in periods of atmospheric blocking situations also are presented.