• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.1
• /
• pp.83-88
• /
• 2024

In this research, we propose a Semi-Supervised learning based railroad surface defect detection method. The Resnet50 model, pretrained on ImageNet, was employed for the training. Data without labels are randomly selected, and then labeled to train the ResNet50 model. The trained model is used to predict the results of the remaining unlabeled training data. The predicted values exceeding a certain threshold are selected, sorted in descending order, and added to the training data. Pseudo-labeling is performed based on the class with the highest probability during this process. An experiment was conducted to assess the overall class classification performance based on the initial number of labeled data. The results showed an accuracy of 98% at best with less than 10% labeled training data compared to the overall training data.

• Journal of Korean Society for Atmospheric Environment
• /
• v.29 no.3
• /
• pp.251-263
• /
• 2013

The surface ozone concentrations changes were investigated in response to climate change over the Korean peninsula for summertime using the global-regional one way coupled Integrated Climate and Air quality Modeling System (ICAMS). The future simulations were conducted under the Intergovernmental Panel on Climate Change (IPCC) Special Report on Emissions Scenarios (SRES) A2 and B1 scenarios. The modeling system was applied for four 10-year simulations: 1996~2005 as a present-day case, 2016~2025, 2046~2055, and 2091~2100 as future cases. The results in this study showed that the mean surface ozone concentrations increased up to 0.5~3.3 ppb under the A2, but decreased by 0.1~10.9 ppb under the B1 for the future, respectively. However, its increases were lower than an increase of the average daily maximum 8-hour (DM8H) surface ozone concentrations which was projected to increase by 2.8~6.5 ppb under the A2. The DM8H surface ozone concentrations seem to be therefore far more affected by the climate and emissions changes than mean values. The probability of exceeding 60 ppb was projected to increase by 6~19% under the A2. In the case of B1, its changes were presented with an increase of 2.9% in the 2020s but no occurrence in the 2100s due to the effect of the reduced emissions. Future projection on surface ozone concentrations was generally shown to have almost the similar trend as the emissions of $NO_x$ and NMVOC.

• Journal of the Korean earth science society
• /
• v.32 no.3
• /
• pp.249-264
• /
• 2011

A multi-Gaussian kriging approach extended to space-time domain is presented for uncertainty modeling as well as time-series mapping of environmental variables. Within a multi-Gaussian framework, normal score transformed environmental variables are first decomposed into deterministic trend and stochastic residual components. After local temporal trend models are constructed, the parameters of the models are estimated and interpolated in space. Space-time correlation structures of stationary residual components are quantified using a product-sum space-time variogram model. The ccdf is modeled at all grid locations using this space-time variogram model and space-time kriging. Finally, e-type estimates and conditional variances are computed from the ccdf models for spatial mapping and uncertainty analysis, respectively. The proposed approach is illustrated through a case of time-series Particulate Matter 10 ($PM_{10}$) concentration mapping in Incheon Metropolitan city using monthly $PM_{10}$ concentrations at 13 stations for 3 years. It is shown that the proposed approach would generate reliable time-series $PM_{10}$ concentration maps with less mean bias and better prediction capability, compared to conventional spatial-only ordinary kriging. It is also demonstrated that the conditional variances and the probability exceeding a certain thresholding value would be useful information sources for interpretation.

• Korean Journal of Remote Sensing
• /
• v.35 no.2
• /
• pp.299-316
• /
• 2019

The purpose of this study was to analyze landslide susceptibility in the Pyeongchang area using Weight of Evidence (WOE) and Evidential Belief Function (EBF) as probability models and Artificial Neural Networks (ANN) as a machine learning model in a geographic information system (GIS). This study examined the widespread shallow landslides triggered by heavy rainfall during Typhoon Ewiniar in 2006, which caused serious property damage and significant loss of life. For the landslide susceptibility mapping, 3,955 landslide occurrences were detected using aerial photographs, and environmental spatial data such as terrain, geology, soil, forest, and land use were collected and constructed in a spatial database. Seventeen factors that could affect landsliding were extracted from the spatial database. All landslides were randomly separated into two datasets, a training set (50%) and validation set (50%), to establish and validate the EBF, WOE, and ANN models. According to the validation results of the area under the curve (AUC) method, the accuracy was 74.73%, 75.03%, and 70.87% for WOE, EBF, and ANN, respectively. The EBF model had the highest accuracy. However, all models had predictive accuracy exceeding 70%, the level that is effective for landslide susceptibility mapping. These models can be applied to predict landslide susceptibility in an area where landslides have not occurred previously based on the relationships between landslide and environmental factors. This susceptibility map can help reduce landslide risk, provide guidance for policy and land use development, and save time and expense for landslide hazard prevention. In the future, more generalized models should be developed by applying landslide susceptibility mapping in various areas.

• Journal of Environmental Health Sciences
• /
• v.47 no.4
• /
• pp.366-377
• /
• 2021

Background: Polycyclic aromatic hydrocarbons (PAHs) are generated in petrochemical complexes, can spread to residential areas and affect the health of residents. Although harmful PAHs are mainly present in particle phase, gas phase PAHs can generate stronger toxic substances through photochemical reaction. Therefore, the risk assessment for PAHs around the petrochemical complex should consider both particle and gas phase concentrations. Objectives: This study aimed to investigate the concentration characteristics of particle and gas phase PAHs by season in residential areas around petrochemical complexes, and to assess the risk of PAHs. Methods: Samples were collected for 7 days by seasons in 2014~2015 using a high volume air sampler. Particle and gas phase PAHs were sampled using quartz filter and polyurethane foam, respectively, analyzed by GC-MS. Chronic toxicity and probabilistic risk assessment were performed on 14 PAHs. For chronic toxicity risk assessment, inhalation unit risk was used. Monte-Carlo simulation was performed for probabilistic risk assessment using the mean and standard deviation of measured PAHs. Results: The concentration of particle total PAHs was highest in autumn. The gas phase concentration was highest in autumn. The average gas phase distribution ratio of low molecular weight PAHs composed of 2~3 benzene rings was 85%. The average of the medium molecular weight composed of 4 benzene rings was 53%, and the average of the high molecular weight composed of 5 or more benzene rings was 9%. In the chronic toxicity risk assessment, 7 of the 14 PAHs exceeded the excess carcinogenic risk of 1.00×10^-6. In the Monte-Carlo simulation, Benzo[a]pyrene had the highest probability of exceeding 1.00×10^-6, which was 100%. Conclusions: The concentration of PAHs in the residential area around the petrochemical complex exceeded the standard, and the excess carcinogenic risk was evaluated to be high. Therefore, it is necessary to manage the air environment around the petrochemical complex.

• Journal of Korea Water Resources Association
• /
• v.55 no.4
• /
• pp.257-266
• /
• 2022

Inundation damage is increasing every year due to localized heavy rain and an increase of rainfall exceeding the design frequency. Accordingly, the importance of hydraulic structures for flood control and defense is also increasing. The hydraulic structures are designed according to its purpose and performance, and the amount of flood is an important calculation factor. However, in Korea, design rainfall is used as input data for hydrological analysis for the design of hydraulic structures due to the lack of sufficient data and the lack of reliability of observation data. Accurate probability rainfall and its temporal distribution are important factors to estimate the design rainfall. In practice, the regression equation of temporal distribution for the design rainfall is calculated using the cumulative rainfall percentage of Huff's quartile method. In addition, the 6^th order polynomial regression equation which shows high overall accuracy, is uniformly used. In this study, the optimized regression equation of temporal distribution is derived using the variable selection method according to the principle of parsimony in statistical modeling. The derived regression equation of temporal distribution is verified through the significance test. As a result of this study, it is most appropriate to derive the regression equation of temporal distribution using the stepwise selection method, which has the advantages of both forward selection and backward elimination.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.2A
• /
• pp.119-130
• /
• 2009

Seismic fragility curves of a structure represent the probability of exceeding the prescribed structural damage state for a given various levels of ground motion intensity such as peak ground acceleration (PGA), spectral acceleration ($S_a$) and spectral displacement ($S_d$). So those are very essential to evaluate the structural seismic performance and seismic risk. The purpose of this paper is to develop seismic fragility curves for PSC box girder bridges. In order to construct numerical fragility curve of bridge structure using nonlinear time history analysis, a set of ground motions corresponding to design spectrum are artificially generated. Assuming a lognormal distribution, the fragility curve is estimated by using the methodology proposed by Shinozuka et al. PGA is simple and generally used parameter in fragility curve as ground motion intensity. However, the PGA has not good relationship with the inelastic structural behavior. So, $S_a$ and $S_d$ with more direct relationship for structural damage are used in fragility analysis as more useful intensity measures instead of PGA. The numerical fragility curves based on nonlinear time history analysis are compared with those obtained from simple method suggested in HAZUS program.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.43 no.6
• /
• pp.699-708
• /
• 2023

Nuclear power plants (NPPs) consider possible external events, including natural hazards, during the design phase to ensure safe operation. However, in recent years, due to the increasing probability of natural hazards exceeding the design, a careful review of extreme natural hazards and unforeseen external events during the design phase has become necessary. In this study, the objective was to screen potential extreme natural hazards at NPP sites in Korea. Initially, we investigated and analyzed the characteristics of NPP sites and the events caused by external hazards. Furthermore, we analyzed existing literature and research data to establish screening procedures and criteria that suit the actual conditions of domestic NPPs. Based on these criteria and data, we conducted qualitative screening for each NPP site and identified potential extreme natural hazards through quantitative screening and walkdown. As a result of the screening, in addition to internal flooding caused by heavy rain, wind pressure and extreme air pressure caused by extreme winds were screened as potential extreme natural hazards common to all sites. Additionally, at the Kori site, storm surge was selected as the most significant potential extreme natural hazard.

• Journal of Environmental Impact Assessment
• /
• v.28 no.2
• /
• pp.152-168
• /
• 2019

The element arsenic, which is abundant in the Earth's crust, is used for various industrial purposes including materials for disease treatment and household goods. Various human activities, such as the disposal of soil waste, metal mining and smelting, and combustion of fossil fuels, have caused the pollution of the environment with arsenic. Recently, guidelines for arsenic in rice have been adopted by the Korean ministry of food and drug safety to prevent health risks based on rice consumption. Because of the exposure to arsenic and its accumulation in the human body through various channels, such as air inhalation, skin contact, ingestion of drinking water, and food consumption, integrated multimedia risk assessment is required to adopt appropriate risk management policies. Therefore, integrated human health risk assessment was carried out in this study using integrated exposure assessment based on multimedia (e.g., air, water, and soil) and multi-route (e.g., oral, inhalation, and dermal) scenarios. The results show that oral uptake via drinking water is the most common pathway of arsenic into the human body, accounting for 57%-96% of the total arsenic exposure. Among various age groups, the highest exposures to arsenic were observed in infants because the body weight of infants is low and the surface areas of infant bodies are large. Based on the results of the exposure assessment, the cancer and non-cancer risks were calculated. The cancer risk for CTE and RME is in the range of 2.3E-05 to 6.7E-05 and thus is negligible because it does not exceed the cancer probability of 1.0E-04 for all age groups. On the other hand, the cancer risk for RME varies from 6.4E-05 to 1.8E-04 and from 1.3E-04 to 1.8E-04 for infants and preschool children, exceeding the excess cancer risk of 1.0E-04. The non-cancer risks range from 5.4E-02 to 1.9E-01 and from 1.5E-01 to 6.8E-01, respectively. They do not exceed the hazard index 1 for all scenarios and all ages.