Natural and anthropogenic fragmented landscapes are pervasive and this complexity significantly affects the structure of the atmospheric boundary layer, causing classic similarity theories to break down. This is especially true in areas affected by wake turbulence. Steep topography and canopy patches can lead to separation of the boundary layer and delay in the adjustment of turbulence to an adjacent underlying surface. Canopy wakes have been shown, in controlled wind tunnel experiments, to significantly affect the mean and turbulence profiles compared to classic rough to smooth transitions (Markfort et al. 2014, Env. Fluid Mech.). The added turbulence due to wakes delay the development of a new boundary layer and turbulent flux measurements and models that rely on similarity theory to determine surface fluxes exhibit significant errors. Here we compare lab-scale experimental measurements using PIV to field-scale measurements using scanning Doppler wind LiDARs. The measurements provide information on how the wake evolves in space and varies over time. Results from the lab and field show a time-varying recirculation zone downwind of the canopy, enhanced turbulence extending far downwind of the transition and reduced surface fluxes in the wake region. The field measurements show that the open trunk space near the base of the canopy results in a surface jet that can be detected just downwind of the canopy and farther downwind dissipates as it mixes with the wake flow above. The implications of canopy wakes for measurements and modeling of surface fluxes will be discussed.