The Ensemble Kalman Filter (EnKF) is considered to be a promising data assimilation approach in land surface data assimilation, as near-real-time land observations such as MODIS and AMSR-E land satellite products have become available, and the high-performance uncoupled Land Information System (LIS) infrastructure has become available as a test bed at NCEP. Currently the Kalman Filter data assimilation technique has been implemented in LIS and experiments of assimilation of AMSR-E soil moisture retrievals have been performed. However, we found that the standard AMSR-E and modeled soil moisture temporal and spatial variability compare poorly with in-situ observations. Subsequently, assimilating the AMSR-E soil moisture product leads to an analysis with low variability. To overcome this problem, bias-correction strategies are used. For example, one can perform a Cumulative Distribution Function (CDF) matching to scale the AMSR-E soil moisture to fit to the in-situ or modeled climatology. In this study, in order to use AMSR-E to provide an improved analysis of land fields that can be directly used as initial conditions for weather and climate prediction, our focus is to reduce the possible biases existing in observations and model forecasts, and investigate the efficiency and benefits of assimilating the AMSR-E data products into the Noah Land Surface Model (LSM). To this end, over the North American domain at 1/8 degree resolution, we first conduct bias correction via 3-year CDF matching with AMSE-E soil moisture products scaled to match Noah LSM climatology; next, we perform the assimilation of the scaled AMSR-E data using 1-D EnKF scheme; next we employ the bias-correction algorithm introduced by Dee (1998) in the data assimilation. Lastly, the assimilation results are compared with available in-situ observations such as SMEX02, or SMEX03 and SCAN data, and the performance of each bias correction option are compared.