A Case Study Assessing the Impact of GOES Imager Data Assimilation on Precipitation Forecast

Assimilation of infrared radiance observations from Geostationary Operational Environmental Satellite (GOES-11/12) for a heavy rainfall event is carried out using the Gridpoint Statistical Interpolation (GSI) of National Centers for Environmental Prediction (NCEP). Although all four infrared channels from both GOES-11 (3.91 , 6.75 , 10.69 and 11.97 ) and GOES-12 (3.90 , 6.48 , 10.71 and 13.31 ) are assimilated, temperature and specific humidity analyses are most sensitive to water vapor channels at 6.75 and 6.48 . Numerical results show that the 3-hourly accumulative quantitative precipitation forecast up to 24 forecasting hours can be significantly improved if both conventional data and GOES data are assimilated. It is also found that impact of GOES imager data assimilation on quantitative precipitation forecast of the selected case depends strongly on the starting time of data assimilation cycling. A detailed analysis of numerical results reveals it relates to the UTC-time dependence of the accuracy of background field.