The Effect of Joint Assimilation of Soil Moisture and Leaf Area Index to Improve Seven-day Land Surface Forecasts over Grassland

The Noah-MP land surface model contains a simple carbon allocation model as a dynamic vegetation option. Vegetation growth is determined by simulated photosynthesis and subsequent allocation of the assimilated carbon into several carbon pools. These pools then feedback to the surface fluxes simulated by the Noah-MP model through the leaf area index and its control on latent heat flux and canopy absorbed radiation. Open loop simulations over a Kansas Ameriflux grassland site show that the grass exhibits a double peak in growth that is not present in observations. This results in a drying of the soil and negative effect on late season prediction. By assimilating either observed 8-day leaf area index or daily soil moisture, the simulations are improved but exhibit significant divergence by the next assimilation period. Assimilating both soil moisture and leaf area index and including one vegetation growth parameter, Vcmax, in the EnKF assimilation procedure improves the simulation of surface energy fluxes and near surface meteorological fields like 2-m temperature and humidity.