A regional wind network with complex surface conditions must be designed with sufficient space and time resolution to resolve the local circulations. In this study, the spatial variations of the wind field observed in the Seoul and Jeju regional networks were evaluated in terms of annual, seasons, and months to assess the spatial homogeneity of wind fields within the regional networks. The coherency of the wind field as a function of separation distance between stations indicated that significant coherency was sometimes not captured by the network, as inferred by low correlations between adjacent stations. A meso-velocity scale was defined in terms of the spatial variability of the wind within the network. This problem is predictably most significant with weak winds, dull prevailing wind, clear skies and significant topography. The relatively small correlations between stations imply that the wind at a given point cannot be estimated by interpolating winds from the nearest stations. For the Seoul and Jeju regional network, the meso-velocity scale has typically a same order of magnitude as the speed of the network averaged wind, revealing the large spatial variability of the Jeju network station imply topography and weather. Significant scatter in the relationship between spatial variability of the wind field and the wind speed is thought to be related to thermally-generated flows. The magnitude of the mesovelocity scale was significantly different along separation distance between stations, wind speed, intensity of prevailing wind, clear and cloudy conditions, topography. Resultant wind vectors indicate much different flow patterns along condition of contributing factors. As a result, the careful considerations on contributing factors such as prevailing wind in season, weather, and complex surface conditions with topography and land/sea contrast are required to assess the spatial variations of wind field on a regional network. The results in the spatial variation from the mesovelocity scale are useful to represent the characteristics of regional wind speed including lower surface conditions over the grid scale of large scale atmospheric model.