식생지수 기법을 이용한 풍력자원지도 작성에 관한 연구

Abstract

Recently, many companies in the wind energy industry have been planning for overseas expansion because of the slump in their own country. To do so, these companies need to evaluate the feasibility of wind resources and wind plants. We need baseline data like roughness length data, contour data and wind data for the wind resource data. In this study, land cover maps were created by remote sensing technique with satellite images. Roughness length maps were created using land cover maps. They are used as basic data to evaluate wind resources with wind data and elevation data. First, the Landsat TM images of 30m resolution were used to create land cover maps. The K-means clustering technique and the NDVI technique were applied to classify land-cover. The overall accuracy using K-means clustering technique was 62.31%. The overall accuracy using NDVI technique was 72.44%. When comparing the K-means clustering technique and the NDVI technique, the overall accuracy was increased about 10%. Next, wind resource maps was created. Roughness length data, contour data and the wind data for the wind resource map were required. To do this, the land cover maps were applied to a majority filter by a leading study for roughness length data. If the majority filter is used, the analysis gets faster and an error happens decreases. Then, the contour data were created by digital maps and wind data from met-masts and AWSs. By using data calculated this way, the wind resource map was made. The energy density data from Han-dong and Gu-jwa were compared based on the wind resource map. As a result, we found that the NDVI technique was very useful to calculate wind energy density in jeju.