Case studies and applications of the Empirical Orthogonal Function quality control in variational data assimilation systems for surface observation data

Surface data assimilation can provide a wealth of ground information, which is particularly important for the accurate simulation of atmospheric boundary layer. Surface data assimilation has always been impacted by the poor quality of observations, and an improving data quality control (QC) method for surface observations is an essential way to advance the effect of data assimilation. In order to improve the impact of surface data assimilation and discuss the influence of this method for precipitation forecast, the EOF (Empirical Orthogonal Function)-QC is introduced into the WRF three-dimensional variational assimilation system. Two severe precipitation events during the periods of 28-29 January 2008 and 14-15 July 2008 have been used to investigate the difference between the EOF-QC and the OMB (observation minus background)-QC. The results show that the EOF-QC can retain weather oscillations in the observations more effectively and reflect the true state of the atmosphere more objectively and accurately than the OMB-QC. The temperature is cooling when assimilating the data after EOF-QC, which yield to a cyclonic circulation, leading the cold air from the north to moving toward the east and weakening the southwest flow with abundant water vapor. The EOF-QC can make the model forecasts of the range and intensity of precipitation become more reasonable. The spatial distribution of precipitation also shows that EOF-QC can improve the forecasting capability of precipitation area and precipitation intensity, the forecasting skill scores are also significantly enhanced. The spatial distribution of precipitation is more close to the real situation than other experiments. The numerical simulations indicate that the EOF-QC method has good potential application in WRF-3DVAR.