Surface-atmosphere exchanges estimated by current land surface models may not be representative of inhomogeneous surface types that are manifest at fine spatial scales. A recent study aimed at characterizing the partitioning of the available energy at the Howard University Beltsville research site, a rural site dominated by typical temperate mixed deciduous and coniferous forests but abutted by urban settings, showed a modulation of the microclimate at the site via microscale heat advection from the urban surfaces during the winter months as opposed to the dominance of local convection during the warm season of 2006. This study seeks to contrast the land surface atmosphere interactions with an emphasis on turbulent fluxes and near-surface thermodynamic quantities simulated by the fully coupled WRF/Noah model with those observed at the Howard University Beltsville site under selected cloud free conditions during the winter, spring, and summer seasons of 2006. Results showed an underestimation of sensible heat flux during the winter and spring with the largest biases during the winter as opposed to a slight overestimation during the summer of 2006, suggesting that the model was unable to account for the microscale heat advection associated with the land surface discontinuity during the winter. Latent heat flux, consistent with past findings, was under predicted throughout the seasons, in particular during the summer of 2006 with net effects of large cool and dry biases in the boundary layer. Potential sources of model errors remained and will be discussed in this presentation.