• Atmosphere
• /
• v.16 no.4
• /
• pp.259-267
• /
• 2006

In the present study, the neural network (NN) model with cluster analysis method was developed to predict storm surge in the whole Korean coastal regions with special focuses on the regional extension. The model used in this study is NN model for each cluster (CL-NN) with the cluster analysis. In order to find the optimal clustering of the stations, agglomerative method among hierarchical clustering methods was used. Various stations were clustered each other according to the centroid-linkage criterion and the cluster analysis should stop when the distances between merged groups exceed any criterion. Finally the CL-NN can be constructed for predicting storm surge in the cluster regions. To validate model results, predicted sea level value from CL-NN model was compared with that of conventional harmonic analysis (HA) and of the NN model in each region. The forecast values from NN and CL-NN models show more accuracy with observed data than that of HA. Especially the statistics analysis such as RMSE and correlation coefficient shows little differences between CL-NN and NN model results. These results show that cluster analysis and CL-NN model can be applied in the regional storm surge prediction and developed forecast system.

• Korean Journal of Culture and Social Issue
• /
• v.30 no.3
• /
• pp.239-265
• /
• 2024

The purpose of this study was to compare the predictive accuracy of traditional prediction models (methods) and machine learning algorithms in predicting suicidal behaviors. The research aimed to go beyond a systematic review level and scientifically examine the predictive capabilities of these two techniques through meta-analysis, analyzing variables identified through domestic research, particularly at the regional level. In order to achieve this, a total of 124 studies, including 50 studies utilizing machine learning and 74 studies employing traditional methods, were included in the meta-analysis. The results of the study revealed that the integrated area under the curve (AUC) for studies using traditional methods was .770, which was lower than the integrated AUC value of .853 for studies using machine learning. Particularly, studies conducted in Asia (AUC = .944) demonstrated higher accuracy compared to studies in Western countries (AUC = .820) and Korea (AUC = .864). Additional analysis of the moderating effects in domestic research indicated that a higher proportion of males and the prediction of suicide attempts were associated with higher prediction accuracy. On the other hand, prediction accuracy was lower when the prediction target was suicide deaths and when studies utilized neural network analysis. This study synthesized various research findings on the prediction of suicidal behaviors, verified the effectiveness of prediction using machine learning, and holds significance in exploring variables applicable in the context of South Korea.

• Journal of the Society of Naval Architects of Korea
• /
• v.48 no.6
• /
• pp.569-574
• /
• 2011

In research vessels or naval ships, airborne noise from machineries such as diesel engine is the major source of underwater noise at low speed. In this paper, effect of engine noise on underwater noise is studied by considering two paths; sound radiation from hull plate and direct airborne noise transmission through hull plate. SEA (Statistical energy analysis) is used to predict hull plate vibration induced by engine noise, where SEA model consists of only two subsystems; engine room air space and hull plate. The pressure level in water is calculated from sound radiation by plate. Engine noise transmission through hull plate is obtained by assuming plane wave propagation in air-limp plate-water system. Two effects are combined and compared to the measurement, where speaker is used as a source in engine room and sound pressure levels in engine room and water are measured. The hydrophone is located 1 m away from the hull plate. It is found below 1000 Hz, prediction overestimates underwater sound pressure level by 5 to 12 dB.

• Atmosphere
• /
• v.27 no.3
• /
• pp.345-358
• /
• 2017

The decadal change in rainfall for Changma period over the South Korea in early-2000s is detected in this study. The Changma rainfall in P1 (1992~2002) decade is remarkably less than in P2 (2003~2013) decade. The much rainfall in P2 decade is associated with the increase of rainy day frequency during Changma period, including the frequent occurrences of rainy day with a intensity of 30 mm/day or more in P2 decade. This decadal change in the Changma rainfall is due to the decadal change of atmospheric circulation around the Korean Peninsula which affects the intensity and location of Changma rainfall. During P2 decade, the anomalous anti-cyclone over the south of the Korean Peninsula, which represents the expansion of the North Pacific high with warm and wet air mass toward East Asia, is stronger than in P1 decade. In addition, the upper level zonal wind and meridional gradient of low-level equivalent potential temperature in P2 decade is relatively strengthened over the northern part of the Korean Peninsula than in P1 decade, which corresponds with the intensification of meridional gradient between air mass related to the East Asian summer monsoon nearby the Korean Peninsula in P2 decade. The enhanced meridional gradient of atir mass during P2 decade is favorable condition for the intensification of Changma rainfall band and more Changma rainfall. The atmospheric conditions related to enhanced Changma rainfall during P2 decade is likely to be influenced by the teleconnection linked to the suppressed convection anomaly over the southern part of China and South China Sea in P2 decade.

• Journal of the Korean Solar Energy Society
• /
• v.36 no.1
• /
• pp.27-37
• /
• 2016

A simple but practical Ensemble Prediction System(EPS) for wind power forecasting was developed and evaluated using the measurement of the offshore meteorological tower, HeMOSU-1(Herald of Meteorological and Oceanographic Special Unite-1) installed at the Southwest Offshore in South Korea. The EPS developed by the Korea Institute of Energy Research is based on a simple ensemble mean of two Numerical Weather Prediction(NWP) models, WRF-NMM and WRF-ARW. In addition, the Kalman Filter is applied for real-time quality improvement of wind ensembles. All forecasts with EPS were analyzed in comparison with the HeMOSU-1 measurements at 97 m above sea level during Typhoon Bolaven episode in August 2012. The results indicate that EPS was in the best agreement with the in-situ measurement regarding (peak) wind speed and cut-out speed incidence. The RMSE of wind speed was 1.44 m/s while the incidence time lag of cut-out wind speed was 0 hour, which means that the EPS properly predicted a development and its movement. The duration of cut-out wind speed period by the EPS was also acceptable. This study is anticipated to provide a useful quantitative guide and information for a large-scale offshore wind farm operation in the decision making of wind turbine control especially during a typhoon episode.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.21 no.3
• /
• pp.241-253
• /
• 2009

This study performs the investigation of a long-term variation of annual maximum surge heights(AMSH) and main characteristics of high surge events, and the statistical evaluation of the AMSH using sea level data at Yeosu and Tongyeong tidal stations over more than 30 years. It is found that the long-term uptrends based on the linear regression in the AMSH are 34.5 cm/34 yr at Yeosu and 33.6 cm/31 yr at Tongyeong, which are relatively much higher than those at Sokcho and Mukho in the Eastern Coast. 71% and 68% of the AMSH occur during typhoon's event in Yeosu and Tongyeong tidal stations, respectively, and the highest surge records are mostly produced by the typhoon. The generalized extreme value distribution taking into account of the time variable is applied to detect time trend in annual maximum surge heights. In addition, Gumbel distribution is checked to find which one is best fitted to the data using likelihood ratio test. The return level and its 90% confidence interval are obtained for the statistical prediction of the future trend. The prevention of the growing storm surge damage by the intensified typhoon requires the steady analysis and prediction of the surge events associated with the climate change.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.28 no.6
• /
• pp.874-881
• /
• 2022

Until 2016, before discussions on the restoration of brackish water of the Nakdong River Estuary started in earnest, the downstream water level was predicted using the data of existing tide level observatories (Busan and Gadeokdo) several kilometers away from the estuary. However, it was not easy to carry out the prediction due to the dif erence in tide level and phase. Therefore, this study was conducted to estimate tide prediction more accurately through tidal harmonic analysis using the measured water level affected by the tides in the offshore waters adjacent to the Nakdong River Estuary. As a research method, the storage status of observation data according to the period and abnormal data were checked at 10-minute intervals in the offshore sea area near the Nakdong River Estuary bank, and the observed and predicted tides were measured using TASK2000 (Tidal Analysis Software Kit) Package, a tidal harmonic analysis program. Regression analysis based on one-to-one comparison showed that the correlation between the two components was high correlation coef icient 0.9334. In predicting the tides for the current year, if possible, more accurate data can be obtained by harmonically analyzing one-year tide observation data from the previous year and performing tide prediction using the obtained harmonic constant. Based on this method, the predicted tide for 2022 was generated and it is being used in the calculation of seawater inflow for the restoration of brackish water of the Nakdong River Estuary.

• Journal of Navigation and Port Research
• /
• v.27 no.4
• /
• pp.359-366
• /
• 2003

The paper describes on the construction and analysis of marine casualty numerical D/B (N-D/B) to implement Korean MArine Casualty Forecasting System (K-MACFOS). The main target of K-MACFOS is to broadcast the prediction number and risk level of marine casualties as like daily weather forecasting. The data relating to a total of 724 ship casualties in the west-southern sea area (33oN∼35oN, 124oE∼127oE) of Korean peninsula for 11 years (1990∼2000) have been compiled and converted into quantitative data with 14 numeric conversion scales. Through the statistical analysis using contour-map visualization, the usability of N-D/B and the casualty features of the target sea areas are discussed. In addition, the optimum year-band selection method is also proposed to provide correct N-D/B analysis and precise prediction of the number of marine casualties.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.21 no.4
• /
• pp.422-428
• /
• 2018

Increase in use of lightweight structures, coupled with the increased acoustic loads resulting from larger and longer range guided missiles, has made missile more susceptible to failures caused by acoustic loads. Thus, accurate prediction of acoustic environment and the response is becoming ever more important for mission success. In this paper, the acoustic response of a sandwich composite skin structure to diffuse acoustic excitation is predicted over a broad frequency range. For the low frequency acoustic analysis, coupled FE-BEM method is used where the structure is modeled using FEM and the interior and exterior fluid is modeled using BEM. For the high frequency region, statistical energy analysis is applied. The predicted acoustic level inside the structure is compared with the result from acoustic test conducted in reverberation chamber, which shows very good agreement.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.28 no.5
• /
• pp.505-513
• /
• 2010

This study investigate the characteristics of tidal constituents, and long-term mean sea level oscillations at Incheon bay. For this, the conditions of three tide stations around Incheon bay have examined, and carried out harmonic analysis on water level data for periods of about 40 years(1960~2007). Four major tidal constituents($M_2$, $S_2$, $K_1$, $O_1$) of each tide station showed tendency that change over the 18.61year lunar node cycle, and the type of tide at three stations is mainly semi-diurnal tides. And also, the past monthly tidal modulations are especially sensitive to the cumulative year of water level data in accuracy of tidal prediction. In case that regard the detached data at three tide stations as a single time series data of 40 years, the results of analysis on a single time series, long-term mean sea level oscillations and modulations of tidal datum at tide stations appears with a range of about 10cm, respectively. In addition, the predicted tides at the Inchcon harbor by global and regional tide models of OSU(Oregon State University) based on various satellite altimetric(Topex Poseidon, Topex Tandem, ERS, GFO) data are compared with the observed tides by KHOA(the Korea Hydrographic and Oceanographic Administration). The results show that the high resolution regional model is a quite good agreement at coastal shallow water region.