Wind farm-atmosphere interaction is complicated by the effect of turbine array configuration on momentum, scalar and kinetic energy fluxes. Wind turbine arrays are often arranged in rectilinear grids and, depending on the prevailing wind direction, may be perfectly aligned or perfectly staggered. The two extreme configurations are end members with a spectrum of infinite possible layouts. A wind farm of finite length may be modeled as an added roughness or as a canopy in large-scale weather and climate models. However, it is not clear which analogy is physically more appropriate. Also, surface scalar flux, including heat, moisture and trace gas (e.g. CO2), are affected by wind farms, and need to be properly parameterized in large-scale models. Experiments involving model wind farms, in aligned and staggered configurations, were conducted in a thermally controlled boundary-layer wind tunnel. Measurements of the turbulent flow were made using a custom x-wire/cold-wire probe. Particular focus was placed on studying the effect of wind farm layout on flow adjustment, momentum and scalar fluxes, and turbulent kinetic energy distribution. The flow statistics exhibit similar turbulent transport properties to those of canopy flows, but retain some characteristic surface-layer properties in a limited region above the wind farms as well. The initial wake growth over columns of turbines is faster in the aligned wind farm. However, the overall wake adjusts within and grows more rapidly over the staggered farm. The flow equilibrates faster and the overall momentum absorption is higher for the staggered compared to the aligned farm, which is consistent with canopy scaling and leads to a larger effective roughness. Surface heat flux is found to be altered by the wind farms compared to the boundary-layer flow without turbines, with lower flux measured for the staggered wind farm.