Assimilation of Leaf Area Index in a Multi-Land Surface Model System to Improve Water Flux and Storage Estimations

High-quality estimates of water fluxes and storage are crucial for understanding the complex land-atmosphere interactions. To improve the accuracy of water variables in land surface models (LSMs), this study assimilates leaf area index (LAI) in a multi-model ensemble data assimilation (DA) system. Specifically, we i) verify the compatibility of integrating two LSMs (Noah-MP and CLM) together; ii) investigate the efficiency of each LSM in assimilating LAI observations; and eventually iii) evaluate the skill of the multi-model assimilation technique where both LSMs are used simultaneously.

The study is conducted by observing system simulation experiments (OSSEs) at global scale. The multi-model assimilation technique is developed based on ensemble Kalman Filter (EnKF) algorithm. A nature run (NR) is conducted first to represent the “truth” in the OSSE. Then the Open Loop (OL) runs are conducted by adding error to the meteorological forcing data. Finally the DA runs apply the assimilation procedure with the synthetic LAI observations extracted from the NR. The multi-model DA system is evaluated by water flux and storage outputs.