• Journal of Korean Society for Quality Management
• /
• v.50 no.4
• /
• pp.701-716
• /
• 2022

Purpose: This study aims to predict the dry cargo transportation market economy. The subject of this study is the BDI (Baltic Dry Index) time-series, an index representing the dry cargo transport market. Methods: In order to increase the accuracy of the BDI time-series, we have pre-processed the original time-series via time-series decomposition and data augmentation techniques and have used them for ANN learning. The ANN algorithms used are Multi-Layer Perceptron (MLP), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM) to compare and analyze the case of learning and predicting by applying time-series decomposition and data augmentation techniques. The forecast period aims to make short-term predictions at the time of t+1. The period to be studied is from '22. 01. 07 to '22. 08. 26. Results: Only for the case of the MAPE (Mean Absolute Percentage Error) indicator, all ANN models used in the research has resulted in higher accuracy (1.422% on average) in multivariate prediction. Although it is not a remarkable improvement in prediction accuracy compared to uni-variate prediction results, it can be said that the improvement in ANN prediction performance has been achieved by utilizing time-series decomposition and data augmentation techniques that were significant and targeted throughout this study. Conclusion: Nevertheless, due to the nature of ANN, additional performance improvements can be expected according to the adjustment of the hyper-parameter. Therefore, it is necessary to try various applications of multiple learning algorithms and ANN optimization techniques. Such an approach would help solve problems with a small number of available data, such as the rapidly changing business environment or the current shipping market.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.33 no.6
• /
• pp.265-274
• /
• 2021

Outlier detection research in ocean data has traditionally been performed using statistical and distance-based machine learning algorithms. Recently, AI-based methods have received a lot of attention and so-called supervised learning methods that require classification information for data are mainly used. This supervised learning method requires a lot of time and costs because classification information (label) must be manually designated for all data required for learning. In this study, an autoencoder based on unsupervised learning was applied as an outlier detection to overcome this problem. For the experiment, two experiments were designed: one is univariate learning, in which only SST data was used among the observation data of Deokjeok Island and the other is multivariate learning, in which SST, air temperature, wind direction, wind speed, air pressure, and humidity were used. Period of data is 25 years from 1996 to 2020, and a pre-processing considering the characteristics of ocean data was applied to the data. An outlier detection of actual SST data was tried with a learned univariate and multivariate autoencoder. We tried to detect outliers in real SST data using trained univariate and multivariate autoencoders. To compare model performance, various outlier detection methods were applied to synthetic data with artificially inserted errors. As a result of quantitatively evaluating the performance of these methods, the multivariate/univariate accuracy was about 96%/91%, respectively, indicating that the multivariate autoencoder had better outlier detection performance. Outlier detection using an unsupervised learning-based autoencoder is expected to be used in various ways in that it can reduce subjective classification errors and cost and time required for data labeling.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.32 no.3
• /
• pp.235-240
• /
• 1996

Forecasts of monthly landings of walleye pollock, Theragra chalcogramma, in Korea were carried out by the seasonal Autoregressive Integrated Moving Average(ARlMA) model. The Box - Cox transformation on the walleye pollock catch data handles nonstationary variance. The equation of Box - Cox transformation was Y'=($Y^0.31$_ 1)/0.31. The model identification was determined by minimum AIC(Akaike Information Criteria). And the seasonal ARlMA model is presented (1- O.583B)(1- $B^1$)(l- $B^12$)$Z_t$ =(l- O.912B)(1- O.732$B^12$)et where: $Z_t$=value at month t ; $B^p$ is a backward shift operator, that is, $B^p$$Z_t$=$Z_t$-P; and et= error term at month t, which is to forecast 24 months ahead the walleye pollock landings in Korea. Monthly forecasts of the walleye pollock landings for 1993~ 1994, which were compared with the actual landings, had an absolute percentage error(APE) range of 20.2-226.1 %. Thtal observed annual landings in 1993 and 1994 were 16, 61OM/T and 1O, 748M/T respectively, while the model predicted 10, 7 48M/T and 8, 203M/T(APE 37.0% and 23.7%, respectively).

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.34 no.6
• /
• pp.222-232
• /
• 2022

In this study we investigated the characteristics of long-term tidal constituents based on 30 tidal gauge data along the coasts of Korea and its the effects on total water level (TWL) forecasts. The results show that the solar annual (Sa) and semiannual (Ssa) tides were dominant among long-term tidal constituents, and they are relatively large in western coast of Korea peninsula. To investigate the effect of long-term tidal constituents on TWL forecasts, we produced predicted tides in 2021 with and without long-term tidal constituents. The TWL forecasts with and without long-term tidal constituents are then calculated by adding surge forecasts into predicted tides. Comparing with the TWL without long-term tidal constituents, the results with long-term tidal constituents reveals small bias in summer and relatively large negative bias in winter. It is concluded that the large error found in winter generally caused by double-counting of meteorological factors in predicted tides and surge forecasts. The predicted surge for 2021 based on the harmonic analysis shows seasonality, and it reduces the large negative bias shown in winter when it subtracted from the TWL forecasts with long-term tidal constituents.

• Journal of Korean Society for Atmospheric Environment
• /
• v.29 no.4
• /
• pp.407-438
• /
• 2013

Current status and future direction of air quality modeling for monitoring and forecasting air quality in East Asia were discussed in this paper. An integrated air quality modeling system, combining (1) emission processing and modeling, (2) meteorological model simulation, (3) chemistry-transport model (CTM) simulation, (4) ground-based and satellite-retrieved observations, and (5) data assimilation, was introduced. Also, the strategies for future development of the integrated air quality modeling system in East Asia was discussed in this paper. In particular, it was emphasized that the successful use and development of the air quality modeling system should depend on the active applications of the data sets from incumbent and upcoming LEO/GEO (Low Earth Orbit/Geostationary Earth Orbit) satellites. This is particularly true, since Korea government successfully launched Geostationary Ocean Color Imager (GOCI) in June, 2010 and has another plan to launch Geostationary Environmental Monitoring Spectrometer (GEMS) in 2018, in order to monitor the air quality and emissions in/around the Korean peninsula as well as over East Asia.

• Ocean Science Journal
• /
• v.42 no.4
• /
• pp.211-222
• /
• 2007

We compare insolation results calculated from two well-known empirical formulas (Socket and Beaudry's SB73 formula and the original Smithsonian (SMS) formula) and a radiative transfer model using input data predicted from meteorological weather-forecast models, and review the accuracy of each method. Comparison of annual mean daily irradiance values for clear-sky conditions between the two formulas shows that, relative to the SMS, the SB73 underestimates spring values by 9 W $m^{-2}$ in the northern Adriatic Sea, although overall there is a good agreement between the annual results calculated with the two formulas. We also elucidate the effect on SMS of changing the 'Sun-Earth distance factor (f)', a parameter which is commonly assumed to be constant in the oceanographic context. Results show that the mean daily solar radiation for clear-sky conditions in the northern Adriatic Sea can be reduced as much as 12 W $m^{-2}$ during summer due to a decrease in the f value. Lastly, surface irradiance values calculated from a simple radiative transfer model (GM02) for clear-sky conditions are compared to those from SB73 and SMS. Comparison with iu situ data in the northern Adriatic Sea shows that the GM02 estimate gives more realistic surface irradiance values than SMS, particularly during summer. Additionally, irradiance values calculated by GM02 using the buoy meteorological fields and ECMWF (The European Centre for Medium Range Weather Forecasts) meteorological data show the suitability of the ECMWF data usage. Through tests of GM02 sensitivity to key regional meteorological factors, we explore the main factors contributing significantly to a reduction in summertime solar irradiance in the Adriatic Sea.

• Proceedings of the Korean Society of Fisheries Technology Conference
• /
• 2009.05a
• /
• pp.38-41
• /
• 2009

The jellyfish Nemopilema nomurai has occurred in large numbers in Japan and Korea and has had on a negative effect on coastal fisheries in this region. Data on the abundance and distribution of jellyfish are needed to forecast when and where they will occur in coastal areas. Acoustic techniques are commonly used to study the distribution and abundance of fish and zooplankton. However before such surveys can be conducted, the acoustic characteristics of the target species must be known. In this study, the density of and speed-of-sound in jellyfish were measured to clarify their acoustic characteristics using a theoretical scattering model. The acoustic characteristics were estimated with the distorted-wave Born approximation (DWBA) model using these material properties and the shape of free-swimming jellyfish.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.1203-1205
• /
• 2003

To detect dust-loaded air-mass over land and ocean, we propose an index, which is essentially the difference in Rayleigh-corrected reflectance between 412 and 443 nm bands of SeaWiFS. Radiative transfer simulations are conducted to show that the index is linearly related to the optical thickness of modeled dust-contaminated aerosol while showing insensitivity against non-absorbing model aerosols. Asian SeaWiFS data set of 2001 spring is used to produce daily composite imagery of the index, which compares well with TOMS Aerosol Index and with predicted aerosol optical thickness predicted by CFORS chemical weather forecast.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.27 no.2
• /
• pp.49-70
• /
• 2022

Sea Surface Temperature (SST), one of the ocean features, has a significant impact on climate, marine ecosystem and human activities. Therefore, SST prediction has been always an important issue. Recently, deep learning has drawn much attentions, since it can predict SST by training past SST patterns. Compared to the numerical simulations, deep learning model is highly efficient, since it can estimate nonlinear relationships between input data. With the recent development of Graphics Processing Unit (GPU) in computer, large amounts of data can be calculated repeatedly and rapidly. In this study, Short-term SST will be predicted through Convolutional Neural Network (CNN)-based U-Net that can handle spatiotemporal data concurrently and overcome the drawbacks of previously existing deep learning-based models. The SST prediction performance depends on the seasonal and interannual SST variabilities around the southern coast of Korea. The predicted SST has a wide range of variance during spring and summer, while it has small range of variance during fall and winter. A wide range of variance also has a significant correlation with the change of the Pacific Decadal Oscillation (PDO) index. These results are found to be affected by the intensity of the seasonal and PDO-related interannual SST fronts and their intensity variations along the southern Korean seas. This study implies that the SST prediction performance using the developed deep learning model can be significantly varied by seasonal and interannual variabilities in SST.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.29 no.4
• /
• pp.180-188
• /
• 2017

Recently, as the strength of winds and waves increases due to the climate change, abnormal waves such as swells have been also increased, which results in the increase of downtime events of loading/unloading in a harbour. To reduce the downtime events, breakwaters were constructed in a harbour to improve the tranquility. However, it is also important and useful for efficient port operation by predicting accurately and also quickly the downtime events when the harbour operation is in a limiting condition. In this study, numerical simulations were carried out to calculate the wave conditions based on the forecasted wind data in offshore area/outside harbour and also the long-term observation was carried out to obtain the wave data in a harbour. A forecasting method was designed using an auto-regressive (AR) and artificial neural networks (ANN) models in order to establish the relationship between the wave conditions calculated by wave model (SWAN) in offshore area and observed ones in a harbour. To evaluate the applicability of the proposed method, this method was applied to predict wave heights in a harbour and to forecast the downtime events in Pohang New Harbour with highly complex topography were compared. From the verification study, it was observed that the ANN model was more accurate than the AR model.