On the assimilation of near surface observations with 3Dvar and EnKF

Effective incorporation of single-level observations, especially those of air temperature and wind data near the earth's surface, to accurately determine the three-dimensional initial atmospheric conditions represents a challenge in numerical weather prediction. In this study we evaluated the ability of both 3-dimensional variational (3DVAR) data assimilation method and ensemble Kalman filter (EnKF) technique in assimilating surface observations with the weather research and forecasting (WRF) model. A series of data assimilation experiments is performed to examine the impact of surface observations on accurate representation of the atmospheric boundary layer (ABL) structure. Various components of surface observations (2-meter temperature or 10-meter winds) are assimilated and their influence on reconstructing the ABL structural features is evaluated. Results from two different data assimilation methods (3DVAR and EnKF) are compared. Specifically, the ability of each method in dealing with the data over complex terrains is examined. Challenges and potential improvements are addressed.