The sounder data and imager data from GOES platform offer their unique usability for improving initial condition of numerical weather forecasting. The multi-channel sounder data are appropriate for describing vertical distribution of temperature and humidity profiles, while imager data are appropriate to detail horizontal structures, particularly cloud-top distribution. Hence the combination of sounder and imager will give best result in cloudy sky.

NOAA's Forecast Systems Laboratory developed a method to combine cloudy radiances of sounder and imager within one dimensional variational framework. The examples at the Southern Great Plain ARM/CART site using MAPS/RUC forecast will be shown. One dimensional versional is planned to be ported to real-time 20km MAPS/RUC.