Terrestrial ecosystems influence the Earth's climate through complex feedbacks. While biogeochemical pools and fluxes can be estimated from a range of measurement techniques, the trajectory of those pools and fluxes can only be assessed by long term monitoring or dynamic modeling. There is substantial disagreement between the projections made by different Land Surface Models (LSMs), which contributes to substantial uncertainty in future climate projections. To improve predictions of carbon, water and energy exchange between the land surface and the atmosphere, we have implemented an Ensemble Kalman Filter data assimilation scheme for the Community Land Model (CLM) using the Data Assimilation Research Testbed (DART). This system can be used to improve the coherence of CLM with observations and accurately characterize uncertainty in terrestrial ecosystem processes to more appropriately estimate biogeochemical feedbacks to the climate system. Here we describe the structure and initial testing of the CLM-DART coupling using data from forest ecosystem research sites which have approximately a decade of eddy co-variance records and ancillary observations. The system reduces discrepancies between the model and observations.