Wind signal transmission in the shallow subsurface is a challenging process due to attenuation, geometrical spreading, reverberation etc. apart from the contribution of acoustic signals propagating through the air. High-frequency ambient noise in seismic recordings is of two types: anthropogenic (noise generated from human interaction such as cars, buildings, wind turbines etc.) and natural (rain, wind, river, animals, etc.). It is common that both seismic networks and noise sources such as wind farms will share locations across the United States in regions such as Oklahoma. Since wind energy exists in all places all the time and using wind turbines for renewable wind energy is increasing; wind effects play a major role and should not be overlooked. To investigate the noise characteristics associated wind power, we deployed a temporary array with 8 Fairfield nodes with varying distances (10 m - 2000 m) to wind turbine towers located in Grant County, Oklahoma in 2017. Spectral analysis reveals a linear relationship between wind speed (1m/s to 15m/s) and noise levels for the entire frequency range. The spectral amplitudes and peak frequencies of the power spectrum show spatiotemporal variations in noise levels in respect to the location of the turbine towers. To quantify the relationship, we characterize directivity using cross-correlation function where time lag indicates signal propagation, and observe the wind turbines to be a clear source of seismic noise with noise power degradation at distance. We compare the 8-node array with three other nodal arrays in Oklahoma. The first array is a concurrent 64 Fairfield node array deployed 70 km west of the Grant County array in Alfalfa County lacking proximity to any wind turbines and does not show strong correlation between noise amplitude and wind speed. The second is a larger aperture, 2016 IRIS wavefield nodal array located nearby the 8-node array in the Grant County. We observe correlation between noise amplitude and wind speed at selected frequency bands. The final array we include is along state Hwy 412 in Enid, where the nodes are located at various distances from several large-scale wind farms. Updated results on the comparison will be further discussed. Preliminary comparisons suggest different noise characteristics with different spatial proximity to wind turbines.