• Korean Journal of Remote Sensing
• /
• v.38 no.6_3
• /
• pp.1749-1760
• /
• 2022

In modern society, human and social damages caused by natural disasters and frequent disaster accidents have been increased year by year. Prompt access to dangerous disaster sites that are inaccessible or inaccessible using state-of-the-art Earth observation equipment such as satellites, drones, and survey robots, and timely collection and analysis of meaningful disaster information. It can play an important role in protecting people's property and life throughout the entire disaster management cycle, such as responding to disaster sites and establishing mid-to long-term recovery plans. This special issue introduces the National Disaster Management Research Institute (NDMI)'s disaster management technology that utilizes various Earth observation platforms, such as mobile survey vehicles equipped with close-range disaster site survey sensors, drones, and survey robots, as well as satellite technology, which is a tool of remote earth observation. Major research achievements include detection of damage from water disasters using Google Earth Engine, mid- and long-term time series observation, detection of reservoir water bodies using Sentinel-1 Synthetic Aperture Radar (SAR) images and artificial intelligence, analysis of resident movement patterns in case of forest fire disasters, and data analysis of disaster safety research. Efficient integrated management and utilization plan research results are summarized. In addition, research results on scientific investigation activities on the causes of disasters using drones and survey robots during the investigation of inaccessible and dangerous disaster sites were described.

• Journal of Korea Water Resources Association
• /
• v.54 no.spc1
• /
• pp.1195-1204
• /
• 2021

Recently, local torrential rain have become more frequent and severe due to abnormal climate conditions, causing a surge in human and properties damage including infrastructures along the river. In this study, water surface elevation prediction algorithm was developed using the LSTM (Long Short-term Memory) technique specialized for time series data among Machine Learning to estimate and prevent flooding of the facilities. The study area is Jamsu Bridge, the study period is 6 years (2015~2020) of June, July and August and the water surface elevation of the Jamsu Bridge after 3 hours was predicted. Input data set is composed of the water surface elevation of Jamsu Bridge (EL.m), the amount of discharge from Paldang Dam (m³/s), the tide level of Ganghwa Bridge (cm) and the number of tweets in Seoul. Complementary data were constructed by using not only structured data mainly used in precedent research but also unstructured data constructed through wordcloud, and the role of unstructured data was presented through comparison and analysis of whether or not unstructured data was used. When predicting the water surface elevation of the Jamsu Bridge, the accuracy of prediction was improved and realized that complementary data could be conservative alerts to reduce casualties. In this study, it was concluded that the use of complementary data was relatively effective in providing the user's safety and convenience of riverside infrastructure. In the future, more accurate water surface elevation prediction would be expected through the addition of types of unstructured data or detailed pre-processing of input data.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.107-119
• /
• 2023

The technology that recognizes a soldier's motion and movement status has recently attracted large attention as a combination of wearable technology and artificial intelligence, which is expected to upend the paradigm of troop management. The accuracy of state determination should be maintained at a high-end level to make sure of the expected vital functions both in a training situation; an evaluation and solution provision for each individual's motion, and in a combat situation; overall enhancement in managing troops. However, when input data is given as a timer series or sequence, existing feedforward networks would show overt limitations in maximizing classification performance. Since human behavior data (3-axis accelerations and 3-axis angular velocities) handled for military motion recognition requires the process of analyzing its time-dependent characteristics, this study proposes a high-performance data-driven classifier which utilizes the long-short term memory to identify the order dependence of acquired data, learning to classify eight representative military operations (Sitting, Standing, Walking, Running, Ascending, Descending, Low Crawl, and High Crawl). Since the accuracy is highly dependent on a network's learning conditions and variables, manual adjustment may neither be cost-effective nor guarantee optimal results during learning. Therefore, in this study, we optimized hyperparameters using Bayesian optimization for maximized generalization performance. As a result, the final architecture could reduce the error rate by 62.56% compared to the existing network with a similar number of learnable parameters, with the final accuracy of 98.39% for various military operations.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.4
• /
• pp.149-158
• /
• 2023

As data utilization is becoming more important recently, the importance of data centers is also increasing. However, the data center is a problem in terms of environment and economy because it is a massive power-consuming facility that runs 24 hours a day. Recently, studies using deep learning techniques to reduce power used in data centers or servers or predict traffic have been conducted from various perspectives. However, the amount of traffic data processed by the server is anomalous, which makes it difficult to manage the server. In addition, many studies on dynamic server management techniques are still required. Therefore, in this paper, we propose a dynamic server management technique based on Long-Term Short Memory (LSTM), which is robust to time series data prediction. The proposed model allows servers to be managed more reliably and efficiently in the field environment than before, and reduces power used by servers more effectively. For verification of the proposed model, we collect transmission and reception traffic data from six of Wikipedia's data centers, and then analyze and experiment with statistical-based analysis on the relationship of each traffic data. Experimental results show that the proposed model is helpful for reliably and efficiently running servers.

• Korean Journal of Remote Sensing
• /
• v.39 no.6_2
• /
• pp.1541-1551
• /
• 2023

Radiometric calibration is a fundamental step in ocean color remote sensing since the step to derive solar radiance spectrum in visible to near-infrared wavelengths from the sensor-observed electromagnetic signals. Generally, satellite sensor suffers from degradation over the mission period, which results in biases/uncertainties in radiometric calibration and the final ocean products such as water-leaving radiance, chlorophyll-a concentration, and colored dissolved organic matter. Therefore, the importance of radiometric calibration for the continuity of ocean color satellites has been emphasized internationally. This study introduces an approach to improve the radiometric calibration algorithm for the visible bands of the Geostationary Ocean Color Imager-II (GOCI-II) satellite with a focus on stability. Solar Diffuser (SD) measurements were employed as an on-orbit radiometric calibration reference, to obtain the continuous monitoring of absolute gain values. Time series analysis of GOCI-II absolute gains revealed seasonal variations depending on the azimuth angle, as well as long-term trends by possible sensor degradation effects. To resolve the complexities in gain variability, an azimuth angle correction model was developed to eliminate seasonal periodicity, and a sensor degradation correction model was applied to estimate nonlinear trends in the absolute gain parameters. The results demonstrate the effects of the azimuth angle correction and sensor degradation correction model on the spectrum of Top of Atmosphere (TOA) radiance, confirming the capability for improving the long-term stability of GOCI-II data.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_4
• /
• pp.1125-1134
• /
• 2023

Out of the total 17,000 reservoirs in Korea, 13,600 small agricultural reservoirs do not have hydrological measurement facilities, making it difficult to predict water storage volume and appropriate operation. This paper examined univariate and multivariate long short-term memory (LSTM) modeling to predict the storage rate of agricultural reservoirs using remote sensing and artificial intelligence. The univariate LSTM model used only water storage rate as an explanatory variable, and the multivariate LSTM model added n-day accumulative precipitation and date of year (DOY) as explanatory variables. They were trained using eight years data (2013 to 2020) for Idong Reservoir, and the predictions of the daily water storage in 2021 were validated for accuracy assessment. The univariate showed the root-mean square error (RMSE) of 1.04%, 2.52%, and 4.18% for the one, three, and five-day predictions. The multivariate model showed the RMSE 0.98%, 1.95%, and 2.76% for the one, three, and five-day predictions. In addition to the time-series storage rate, DOY and daily and 5-day cumulative precipitation variables were more significant than others for the daily model, which means that the temporal range of the impacts of precipitation on the everyday water storage rate was approximately five days.

• The Journal of Industrial Distribution & Business
• /
• v.9 no.5
• /
• pp.63-72
• /
• 2018

Purpose - This work analyzes, in detail, the specification of vector error correction model (VECM) and thus examines the relationships and impact among seven economic variables for USA - balance on current account (BCA), index of stock (STOCK), gross domestic product (GDP), housing price indices (HOUSING), a measure of the money supply that includes total currency as well as large time deposits, institutional money market funds, short-term repurchase agreements and other larger liquid assets (M3), real rate of interest (IR_REAL) and household credits (LOAN). In particular, we search for the main explanatory variables that have an effect on stock and real estate market, respectively and investigate the causal and dynamic associations between them. Research design, data, and methodology - We perform the time series vector error correction model to infer the dynamic relationships among seven variables above. This work employs the conventional augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) unit root techniques to test for stationarity among seven variables under consideration, and Johansen cointegration test to specify the order or the number of cointegration relationship. Granger causality test is exploited to inspect for causal relationship and, at the same time, impulse response function and variance decomposition analysis are checked for both short-run and long-run association among the seven variables by EViews 9.0. The underlying model was analyzed by using 108 realizations from Q1 1990 to Q4 2016 for USA. Results - The results show that all the seven variables for USA have one unit root and they are cointegrated with at most five and three cointegrating equation for USA. The vector error correction model expresses a long-run relationship among variables. Both IR_REAL and M3 may influence real estate market, and GDP does stock market in USA. On the other hand, GDP, IR_REAL, M3, STOCK and LOAN may be considered as causal factors to affect real estate market. Conclusions - The findings indicate that both stock market and real estate market can be modelled as vector error correction specification for USA. In addition, we can detect causal relationships among variables and compare dynamic differences between countries in terms of stock market and real estate market.

• The Journal of Engineering Geology
• /
• v.32 no.4
• /
• pp.697-723
• /
• 2022

Data-driven models to predict groundwater levels 30 days in advance were developed for 12 groundwater monitoring stations in the middle-Jeju watershed, Jeju Island. Stacked long short-term memory (stacked-LSTM), a deep learning technique suitable for time series forecasting, was used for model development. Daily time series data from 2001 to 2022 for precipitation, groundwater usage amount, and groundwater level were considered. Various models were proposed that used different combinations of the input data types and varying lengths of previous time series data for each input variable. A general procedure for deep-learning-based model development is suggested based on consideration of the comparative validation results of the tested models. A model using precipitation, groundwater usage amount, and previous groundwater level data as input variables outperformed any model neglecting one or more of these data categories. Using extended sequences of these past data improved the predictions, possibly owing to the long delay time between precipitation and groundwater recharge, which results from the deep groundwater level in Jeju Island. However, limiting the range of considered groundwater usage data that significantly affected the groundwater level fluctuation (rather than using all the groundwater usage data) improved the performance of the predictive model. The developed models can predict the future groundwater level based on the current amount of precipitation and groundwater use. Therefore, the models provide information on the soundness of the aquifer system, which will help to prepare management plans to maintain appropriate groundwater quantities.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.26 no.2
• /
• pp.186-194
• /
• 2020

In this study, spectral analysis was conducted to identify environmental factors af ecting short-term changes in water temperature in the East, West and South coasts of Korea. The data used in the spectrum analysis is the 2016 summer water temperature, air temperature, tide level and wind data provided by Korea Hydrographic & Oceanographic Agency. In power spectrum results, peaks of water temperature and tide level were observed at same periods in West Sea (Incheon, Pyeungteak, Gunsan and Mokpo) and South Sea (Wando, Goheung, Yeosu, Tongyeong and Masan) where mean tidal range was more than 100 cm. On the other hand, periodicity of water temperature did not appear in East Sea and Busan where the mean tidal range was small. Coherence analysis showed that water temperature was highly correlated with tide in West Sea and three stations(Wando, Goheung and Tongyeong) of South Sea. Especially, correlation between water temperature and tide level in Wando and Tongyeong presented 0.96 at semi-diurnal period. Water temperature in Yeosu seems to have influenced by tide and inflow of fresh water. In Masan, water temperature is influenced by south wind, tide and inflow of fresh water. In East Sea, influence of tide on water temperature is small due to current and small tidal range. As a result of comparing the time series graph, stations where the correlation between water temperature and tide is high show that relatively cold water was inputted at flood tide and flow out at ebb tide. short-term variation of water temperature was affected by tide, but long-term variation over a month was affected by air temperature.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.34 no.4
• /
• pp.1125-1138
• /
• 2014

In this study, we applied an advanced non-parametric low-flow frequency analysis using boundary kernel by Representative Concentration Pathways (RCPs) climate change scenarios through Arc-SWAT long-term runoff model simulation at the Gwangdong storage reservoir located in Taeback, Gangwondo. The results show that drought frequency under RCPs was expected to increase due to reduced runoff during the near future, and the variation of low-flow time series was appeared greatly under RCP8.5 scenario, respectively. The result from drought frequency of Median flow in the near future (2030s) compared historic period, the case of 30-year low-flow frequency was increased (the RCP4.5 shows +22.4% and the RCP8.5 shows +40.4%), but in the distant future (2080s) expected increase of drought frequency due to the reduction of low-flow (under RCP4.5: -4.7% and RCP8.5: -52.9%), respectively. In case of Quantile 25% flow time series data also expected that the severe drought frequency will be increased in the distant future by reducing low-flow (the RCP4.5 shows -20.8% to -60.0% and the RCP8.5 shows -30.4% to -96.0%). This non-parametric low-flow frequency analysis results according to the RCPs scenarios have expected to consider to take advantage of as a basis data for water resources management and countermeasures of climate change in the mid-watershed over the Korean Peninsula.