Future Work and Challenges:
§ Ensemble Kalman Filter will be implemented in LDAS to further
compare the relative advantages of EnKF and EKF.
§ Other land surface models (Catchment, CLM, NOAH, VIC) will be used
with Kalman Filters for assimilating land data products from satellite
observations.
§ The most efficient and reliable data assimilation algorithm and land
surface model will be selected for operational procedures of AMSR-E
land product validation.
§ Soil moisture assimilation results will also be analyzed with model
output of fluxes.
§ How model error variance and observation error variance can be
known a priori?
§ How to implement the assimilation algorithm to the global scale?
§ How the spatial heteorogeneity of land surafce state variables can be
considered or used in the data assimilation and/or validation?