http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
개별검색 DB통합검색이 안되는 DB는 DB아이콘을 클릭하여 이용하실 수 있습니다.
통계정보 및 조사
예술 / 패션
<해외전자자료 이용권한 안내>
- 이용 대상 : RISS의 모든 해외전자자료는 교수, 강사, 대학(원)생, 연구원, 대학직원에 한하여(로그인 필수) 이용 가능
- 구독대학 소속 이용자: RISS 해외전자자료 통합검색 및 등록된 대학IP 대역 내에서 24시간 무료 이용
- 미구독대학 소속 이용자: RISS 해외전자자료 통합검색을 통한 오후 4시~익일 오전 9시 무료 이용
※ 단, EBSCO ASC/BSC(오후 5시~익일 오전 9시 무료 이용)
The objective of this study was to analyze during summer season short-wave and long-wave radiation characteristics between downtown area and suburban area in Daegu. This study was confirmed the regional and monthly radiation environments and it can utilize as basic data for the analysis of the urban radiation environment and the effects of urbanization.
The purpose of this study is to understand relation of meteorological elements of air temperature, relative humidity and vapor pressure of four cities with Daegu. The followings are main results from this study. 1) There is very high correlation of meteorological elements according to distance between city and city. 2) In case of seaside town at Pohang, there were little changes than other cities for temperature, humidity and vapor pressure. 3) It was analysed stable and similar diurnal variation in water vapor pressure than air temperature and relative humidity at all observation site.
The purpose of this study is to understand the characteristics of urban climate in several cities, from observing radiation according to wavelength band(UV, short and long wave radiation). Observation start from 5 May to 31 August 2013. The followings are the main results from this study. 1) In every observation area, greater amounts of short-wave radiation have been recorded in May compared to June. Even though the highest solar elevation occurs in June, May sees clearer days, which has attributed to the outcome. 2) The analysis concerning the correlation between ultraviolet radiation and shortwave radiation have revealed that regions closer to the Daegu area have stronger correspondence. 3) The time series of daily long-wave radiation shares a similar tendency with the time series of air temperature, and the maximum value was recorded at 14:00 and 15:00.
The purpose of this study is to understand daily variation of short-wave radiation trends according to the state of surface and observation of atmosphere conditions in downtown and suburban observation area. The followings are main results from this study. 1) We found out daily air temperature variation of downtown is less than that of suburban area because of bigger heat capacity of artificial elements such as massive buildings and pavements. 2) It is more effective to estimate of air condition by water vapor pressure than relative humidity in the atmosphere. 3) The difference of solar radiation ratio between downtown and suburban area is dependant on different atmosphere conditions at two observation stations.
기존의 역학모델은 오랜 연구로 인한 이론적 바탕이 누적되어 높은 수준의 기상예측이 가능하나, 이러한 모델을 수행하기 위해서는 과도한 전산자원을 필요로 한다. 전산자원의 감축 및 성능개선을 위해서 최근 기계학습 관련 연구가 대두되고 있다. 본 논문은 슬라이딩 윈도우 기법과 기계학습 (심층 신경망, 서포트 벡터 머신, 랜덤 포레스트)을 활용하여 단기 풍속예측모델을 생성하고, 이를 평가하여 최적의 예측모델을 제안하고자 한다. 본 연구에 사용된 자료는 기상청에서 제공하는 남한 전역의 ASOS 95지점의 2017년 08월부터 2018년 8월까지의 관측자료이며, 제안된 풍속 예측모델의 입력 변수는 기온, 풍향, 습도, 강수와 관측지점의 위경도, 시간 변수를 활용하였다. 그 결과, 기계학습 기법은 저풍속대와 육지지형에서 뛰어난 풍속 예측상능을 보였으며, 기계학습 기법 중에서는 랜덤 포레스트 기법을 이용하였을 때 가장 우수한 성능을 나타내었다. In this paper, we propose a method that utilizes machine learning (deep neural network, support vector machine, random forest) learned from weather observation data to increase the accuracy of short-term prediction of wind speed. The proposed method selects an optimal model after a sensitivity experiment, and then performs verification with the observed wind speed. The sensitivity experiment targets the data extracted by applying the sliding window method and the set factor values of machine learning. The elements used in the learning materials are time, spatial and meteorological factors (temperature, wind direction, humidity, precipitation). The meteorological data used was the value of ASOS 95 sites provided by the Korea Meteorological Administration from August 2017 to August 2018. As a result, the optimal machine learning method showed excellent predictive performance in the low wind speed section and the land terrain. In particular, it can be seen that the Random forest is the best in performance and time resources compared to Supporter vector machines and Deep neural networks.
Characteristics of wind resources of offshore and coastal regions were compared using wind data obtained from HeMOSU-1 (Herald of Meteorological and Oceanographic Special Unit-1) meteorological mast located at Southwestern Sea, and ground-based LiDAR (Light Detection And Ranging) at Gochang observation site near it. The analysis includes comparison of basic wind statistics such as mean wind speed, wind direction, power law exponent and their temporal variability as well as site assessment items for the wind power plant such as turbulence intensity and wind power density at the two observation sites. It was found that the wind at HeMOSU-1 site has lower diurnal and seasonal variability than that at Gochang site, which lead to smaller turbulence intensity. Overall, the results of the comparative analysis show that the wind resource at HeMOSU-1 site located offshore has more favorable condition for wind power generation than the wind resource at Gochang which shows nature of coastal area.