• Journal of the Korean earth science society
• /
• v.40 no.6
• /
• pp.606-612
• /
• 2019

In this study, to estimate the occurrence of frost that has a negative effect on th growth of crops, we constructed to the statistical model. We factored such various meteorological elements as the minimum temperature, temperature at 18:00, temperature at 21:00, temperature at 24:00, average wind speed, wind speed at 18:00, wind speed at 21:00, amount of cloud, amount of precipitation within 5 days, amount of precipitation within 3 days, relative humidity, dew point temperature, minimum grass temperature and ground temperature. Among the diverse variables, the several weather factors were selected for frost occurrence estimation model using statistical methods: T-test, Variable importance plot of Random Forest, Multicollinearity test, Akaike Informaiton Criteria, and Wilk's Lambda values. As a result, the selected meteorological factors were the amount of cloud, temperature at 24:00, dew point temperature, wind speed at 21:00. The accuracy of the frost occurrence estimation model using Random Forest was 70.6%. When it applied to the main production areas of vegetables, a estimation accuracy of the model was 65.2 and 78.6%.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.540-540
• /
• 2015

유역의 하천관리 및 홍수관리를 위하여 강우량을 정확하게 예측하고자 많은 수문학자들에 의해 강우량을 예측하는 연구를 진행하였다. 강우를 예측하기 위한 여러 가지 방법 중 인공신경망을 이용하여 강우를 예측하는 선행연구들을 살펴볼 수 있었다. 그러나 기존에 강우량을 예측하는 사례들을 살펴보게 되면, 강우사상이 발생된 후 강우량 예측은 비교적 높은 정확도를 가지고 있으나, 강우가 발생하기 시작하는 시점에 대한 강우량 예측은 그 정확성이 떨어지는 것을 확인할 수 있었다. 이에 본 연구에서는 무강우 기간에도 보다 정확하게 강우량을 예측할 수 있는 인공신경망 모델을 제안하고자 한다. 이를 위해 강우량 이외에도 기온, 풍속, 습도, 증기압, 전운량을 인공신경망의 입력자료로 활용하고자 하였다. 입력자료을 구성을 여러 가지 CASE로 구분하여 부산지점의 강우량을 예측하고 그 정확성을 평가하고자 하였다. 이 때, 사용되는 자료는 기상청 부산지점에서 제공하고 있는 1시간 간격자료를 적용하였다. 본 연구를 통해 개발된 인공신경망 모형을 이용하여 예측된 강우량은 부산 내에 위치한 하천관리 뿐 만 아니라 하천의 홍수 예 경보에 필요한 기초적인 자료로 활용될 수 있을 것으로 판단된다.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.26 no.1
• /
• pp.16-24
• /
• 2014

Long-term measured wind data are absolutely necessary to estimate extreme offshore wind speed. However, it is almost impossible to collect offshore wind measured data. Therefore, typhoon simulation is widely used to analyze offshore wind conditions. In this paper, 74 typhoons which affected the western sea of Korea during 1978-2012(35 years) were simulated using Holland(1980) model. The results showed that 49.02 m/s maximum wind speed affected by BOLAVEN(1215) at 100 m heights of HeMOSU-1 (Herald of Meteorological and Oceanographic Special Unit - 1) was the biggest wind speed for 35 years. Meanwhile, estimated wind speeds were compared with observed data for MUIFA, BOLAVEN, SANBA at HeMOSU-1. And to estimate extreme wind speed having return periods, extreme analysis was conducted by assuming 35 annual maximum wind speed at four site(HeMOSU-1, Gunsan, Mokpo and Jeju) in western sea of the Korean Peninsular to be Gumbel distribution. As a results, extreme wind speed having 50-year return period was 50 m/s, that of 100-year was 54.92 m/s at 100 m heights, respectively. The maximum wind speed by BOLAVEN could be considered as a extreme winds having 50-year return period.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.14 no.4
• /
• pp.277-282
• /
• 2012

When wind speed exceeds a certain threshold, daily minimum temperature does not drop as predicted by the geospatial model in a cold pooling catchment. A linear regression equation was derived to explain the warming effect of wind speed on daily minimum temperature by analyzing observations at a low lying location within an enclosed catchment. The equation, Y=2X+0.4 ($R^2$=0.76) where Y stands for the warming ($^{\circ}C$) and X for the mean horizontal wind speed (m/s) at 2m height, was combined to an existing model to predict daily minimum temperature across an enclosed catchment on cold pooling days. The adjusted model was applied to 3 locations submerged in a cold air pool to predict daily minimum temperature on 25 cold pooling days with the input of simulated wind speed at each location. Results showed that bias (mean error) was reduced from -1.33 to -0.37 and estimation error (RMSE) from 1.72 to 1.20, respectively, in comparison with those from the unadjusted model.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2008.05a
• /
• pp.1013-1017
• /
• 2008

이 연구는 기상수치예보 모델 중 지역수치예보모델인 RDAPS 모델을 이용하여 강우자료를 예측한 값과 실제 강우관측지점에서의 강우량을 비교해 보고 RDAPS 예측량의 정확도를 높이기 위한 연구이다. RDAPS 모델의 자료는 00UTC와 12UTC에 3시간 누적 자료를 48시간에 대해서 생성하고, 30km 격자망에 대한 정보를 담고 있기 때문에 1시간 간격으로 측정된 지점 강우량과의 비교를 위해서는 관측지점과 근거리 정보를 찾고 1시간 간격의 관측 자료를 3시간 누적강우량으로 바꾸는 전처리 과정이 필요하다. 실제 강우예측이 어려움을 겪는 것처럼 RDAPS의 예측 강우량과 관측 강우량은 큰 차이를 보이는 것으로 나타났다. 예측 강우량의 정확도를 높이고자 인공신경망을 적용하였다. 인공신경망이란 뇌기능의 특성 몇가지를 컴퓨터 시뮬레이션으로 표현하는 것을 목표로 하는 수학 모델이다. 강우수치예측 자료 외에도 RDAPS 모델에서 얻을 수 있는 풍향, 풍속, 상대습도, 기압, 온도 등의 다른 수치자료들을 이용하여 인공신경망을 이용하여 자료들의 패턴을 시뮬레이션 하여 정확도가 높은 예측값을 얻을 수 있었다.

• Journal of Bio-Environment Control
• /
• v.8 no.1
• /
• pp.9-18
• /
• 1999

Natural ventilation in a four and one-half span, double polyethylene commercial greenhouse was investigated with actual data collected at Quailcrest Farm near Wooster, Ohio. Moreover, a computational fluid dynamics (CFD) numerical technique, FLUENT V4.3, was used to predict natural ventilation rates, thermal conditions, and airflow distributions in the greenhouse. The collected climate data showed that the multi-span greenhouse was well ventilated by the natural ventilation system during the typical summer weather conditions. The maximum recorded air temperature difference between inside and outside the greenhouse was 3.5$^{\circ}C$ during the hottest (34.7$^{\circ}C$) recorded sunny day; the air temperatures in the greenhouse were very uniform with the maximum temperature difference between six widely dispersed locations being only 1.7$^{\circ}C$. The CFD models predicted that air exchange rates were as high as 0.9 volume per minute (A.C. .min$^{-1}$ ) with 2.5m.s$^{-1}$ winds from the west as designed.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.445-445
• /
• 2018

기후변화에 따른 환경오염은 21세기 인류에게 가장 심각한 문제 중의 하나로 대두되고 있다. 환경적인 측면에서 하천오염은 경제적으로 많은 문제를 발생시키고 있다. 이러한 하천오염 문제를 해결하기 위해서는 오염물질의 농도 측적 및 데이터 축적이 필수적이라 할 수 있다. 그러나 일반적으로 오염물질 부하량에 대한 직접적인 측정은 비용 측면에서 쉽지 않은 것이 사실이다. 또한 실시간으로 BOD, COD, TN, TP 등의 자료를 이용하여 예측하는 것에는 자료의 부족성으로 인해 한계가 있다. 본 연구에서는 구글의 딥러닝 오픈소스 라이브러리인 텐서플로우를 활용하여 기계학습을 통한 하천오염 예측을 목적으로 하고 있다. 기계학습을 위하여 텐서플로우를 활용하여 RNN, LSTM 인공신경망 모형을 구축하였다. 하천오염의 학습과 예측을 위해 결과치 분석을 위한 자료로는 금강 유역에 위치한 옥천 관측소 충청북도 옥천군 이원면 이원대교에 위치한 $36^{\circ}14'31.0''N$ $127^{\circ}40'02.6''E$의 관측소에서 BOD, COD, DO, 부유물질의 자료를 사용하였다. 모형의 학습을 위해서 입력자료는 수위, 유량, 평균기온, 평균풍속 자료를 2004년 ~ 2017년까지의 14년간의 자료를 사용하였다. 연구를 위해 BOD, COD, DO 부유물질 자료는 물환경정보시스템(http://water.nier.go.kr/)의 자료를 활용하고 수위, 유량등의 자료는 국가수자원관리종합정보시스템 (http://www.wamis.go.kr/)의 자료를 사용하였다. 그러나 수온, 수위, 풍속등의 자료는 일 자료가 있는가 반면 BOD, COD, TN, TP등의 자료는 일 자료가 있지 않아 이를 원활히 활용할 수 있도록 예측을 위한 결과치의 선형보간법을 통해 일 자료를 획득한 후 연구를 하였다. RNN, LSTM의 분석 시 학습속도, 반복시행횟수 sequence length의 길이 등의 값을 조절 하면서 결과치를 분석하였다.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.23 no.4
• /
• pp.222-234
• /
• 2021

Airborne-pests can be introduced into Korea from overseas areas by wind, which can cause considerable damage to major crops. Meteorological models have been used to estimate the wind trajectories of airborne insects. The objective of this study is to analyze the effect of input settings on the prediction of areas where airborne pests arrive by wind. The wind trajectories were predicted using the HYbrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model. The HYSPLIT model was used to track the wind dispersal path of particles under the assumption that brown plant hopper (Nilaparvata lugens) was introduced into Korea from sites where the pest was reported in China. Meteorological input data including instantaneous and average wind speed were generated using meso-scale numerical weather model outputs for the domain where China, Korea, and Japan were included. In addition, the calculation time intervals were set to 1, 30, and 60 minutes for the wind trajectory calculation during early June in 2019 and 2020. It was found that the use of instantaneous and average wind speed data resulted in a considerably large difference between the arrival areas of airborne pests. In contrast, the spatial distribution of arrival areas had a relatively high degree of similarity when the time intervals were set to be 1 minute. Furthermore, these dispersal patterns predicted using the instantaneous wind speed were similar to the regions where the given pest was observed in Korea. These results suggest that the impact assessment of input settings on wind trajectory prediction would be needed to improve the reliability of an approach to predict regions where airborne-pest could be introduced.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.5 no.4
• /
• pp.213-217
• /
• 2003

Forest fire spread and intensity were modeled as a function of wind and fuel. Spread rate and intensity of forest fire were related to weight and thickness of forest fuel beds and to wind speed. Forest fire spread rate and fire intensity were differentiated according to wind speed. Rapid wind speed causes a faster forest fire spread rate and greater fire intensity than does slow wind speed. Relative burning time of the fire from beginning to end in the model was 161 sec at a wind speed of 0.5 m/sec and 146 sec at 1m/sec on the model. Average forest lire spread rate was 0.014 m/sec at a wind speed of 0.5 m/sec and 0.020 m/sec at 1m/sec. Average fire intensity was 0.183 ㎾/m at a wind speed of 0.5 m/sec, 0.259 ㎾/m at 1m/sec. Fire intensity was greater when forest fire spread rate was rapid.

• Journal of Advanced Marine Engineering and Technology
• /
• v.33 no.8
• /
• pp.1170-1179
• /
• 2009

The purpose of this work is to predict the low frequency aero-acoustic noise generated from the horizontal axis wind turbine, NREL Phase VI for the whole operating conditions of various wind speeds using large eddy simulation and Ffowcs-Williams and Hawkings model provided in the commercial code, FLUENT. Because there is no experimental data about wind turbine noise, we first of all compared aerodynamic performance such as shaft torque and power with experimentally measured value. Performance results show a good agreement with experimental data within about 0.8%. As the wind speed increases, the overall sound pressure level and the sound pressure level by the quadrupole and dipole source show a increasing tendency. Also, sound pressure level is proportional to $r^{-2}$ in the near field and $r^{-1}$ in the far field according to the increase of distance from the center of hub of wind turbine. According to 2 times increase of distance, sound pressure level is reduced by about 6dB.