The assessment of wind resources must be carried out to choose wind farm sites adequately. Additionally, input data on surface roughness maps and topographic maps are required to evaluate wind resources, where input data accuracy determines the accuracy of their overall analysis. To estimate this accuracy, we used met-mast data in Jeju and produced the ground roughness value for the Jeju region. To determine these values, an unsupervised classification method using SPOT-5 images was carried out for image classification. The wind resources of the northeastern part of Jeju were predicted, and the ground roughness map of the region was calculated by the WindPRO software. The wind speed of the Pyeongdae region of Jeju from the ground roughness map was calculated using WindPRO as 8.51 m/s. The wind speed calculated using the remote sensing technology presented in this study was 8.69 m/s. To assess the accuracy of the measured WindPro and the remote sensing technology values, we compared these results to the observed values in the Pyeongdae region using met-mast. This comparison shows that remote sensing data are more accurate than the WindPro data. We also found that the ground roughness map calculated in this study is useful for generating an accurate wind resource map of Jeju Island.

Abstract
1. 서 론
2. 재료 및 방법
3. 결과 및 고찰
    3.1. 제주지역에 적합한 지면조도계수 계산
    3.2. Remote Sensing 기술을 이용한 영상 분류
    3.3. 영상 분류지도를 이용한 지면조도지도생성
4. 결 론
감사의 글
REFERENCES

• 이병걸(제주대학교 토목공학과) | Byung-Gul Lee (Department of Civil Engineering, Jeju National University) Corresponding author
• 변지선(제주대학교 토목공학과) | Ji-Seon Byun (Department of Civil Engineering, Jeju National University)
• 고정우(제주대학교 토목공학과) | Ko, Jung-Woo (Department of Civil Engineering, Jeju National University)