SSM/I ----- Data from the "Special Sensor Microwave Imager" is only being used in determining the sea ice concentration. We expect to use more SSM/I data in future NCEP (re)analyses. The SSM/I data we expect to use in the future includes surface wind speeds over the ocean, precipitable water over the ocean, and snow/ice cover. Problems that we had in using SSM/I data: 1) The SSM/I data starts in mid-1987. Unfortunately if you make a 15-year climatology, only half of the reanalysis would use SSM/I data. 2) NCEP's current method of assimilating SSM/I wind-speed data is to estimate the surface-wind direction from the first-guess winds and observations and to use SSM/I to provide the wind speed. For this method to improve the analysis, the estimated wind direction must already be reasonable (at least in direction). The preferred method, to directly assimilate the wind speeds, may be ready in the near future. 3) The current SSM/I snow cover is of higher spatial and temporal resolution than snow-cover maps produced operationally by NESDIS. However, operational snow-cover maps have a much longer history, and are a more tested product. 4) There were other arguments against using SSM/I of both a financial and practical natures. Various analyses use varying amount of SSM/I data. The current NCEP operational analyses use both SSM/I wind speed and sea-ice concentration. Reanalysis only uses SSM/I sea ice concentration (1987) and the ECMWF does not use SSM/I.