To investigate the impact of decadal-scale changes in vegetative cover on wind resources over realistic forested terrain, two large-eddy simulations were conducted. Two temporally separated airborne laser scanning datasets were processed to construct realistic forest representations, which were used to define drag sources in LES under neutral atmospheric conditions. The results are analyzed using spatial and temporal averages, which are compared to measurements at the met-mast location when applicable. Temporal changes in forest structure have been shown to induce differences in the horizontally averaged mean wind speed of up to 0.7%, while the horizontally averaged turbulence intensity varies by up to 3.5% below 200 m above ground height. Localized wind speed-up associated with newly formed clearings, flow deceleration due to forest growth, and changes in overall forest heterogeneity significantly affect rotor-averaged wind speeds, with differences ranging from 5% reductions to increments of up to 9% between the two forest realizations. Finally, changes in the effective aerodynamic surface roughness are found to cancel out to a large degree when spatially averaged over different segments of the simulated domain, with maximum increases of up to 0.24 m.