• Proceedings of the Korean Society of Crop Science Conference
• /
• 2017.06a
• /
• pp.315-315
• /
• 2017

In Japan, more than 80 % of soybean growing area is converted fields and excess water is one of the major problems in soybean production. For example, recent study (Yoshifuji et al., 2016) suggested that in the fields of shallow groundwater level (GWL) (< 1m depth), rising GWL even in a short period (e.g. 1 day) causes inhibition of soybean growth. Thus it becomes more and more important to predict GWL and soil moisture in detail. In addition to conventional surface drainage and underdrain, FOEAS (Farm Oriented Enhancing Aquatic System), which is expected to control GWL in fields adequately, has been developed recently. In this study we attempted to predict GWL and soil moisture condition at the converted field with FOEAS in Biwa lake reclamation area, Shiga prefecture, near the center of the main island of Japan. Two dimensional HYDRUS model (Simuinek et al., 1999) based on common Richards' equation, was used for the calculation of soil water movement. The calculation domain was considered to be 10 and 5 meter in horizontal and vertical direction, respectively, with two layers, i.e. 20cm-thick of plowed layer and underlying subsoil layer. The center of main underdrain (10 cm in diameter) was assumed to be 5 meter from the both ends of the domain and 10-60cm depth from the surface in accordance with the field experiment. The hydraulic parameters of the soil was estimated with the digital soil map in "Soil information web viewer" and Agricultural soil-profile physical properties database, Japan (SolphyJ) (Kato and Nishimura, 2016). Hourly rainfall depth and daily potential evapo-transpiration rate data were given as the upper boundary condition (B.C.). For the bottom B.C., constant upward flux, which meant the inflow flux to the field from outside, was given. Seepage face condition was employed for the surrounding of the underdrain. Initial condition was employed as GWL=60cm. Then we compared the simulated and observed results of volumetric water content at depth of 15cm and GWL. While the model described the variation of GWL well, it tended to overestimate the soil moisture through the growing period. Judging from the field condition, and observed data of soil moisture and GWL, consideration of soil structure (e.g. cracks and clods) in determination of soil hydraulic parameters at the plowed layer may improve the simulation results of soil moisture.

• Journal of Korean Society of Forest Science
• /
• v.104 no.3
• /
• pp.403-410
• /
• 2015

This study was conducted to analyze the forest environmental characteristics on a total of 20 forest environmental factors affecting the debris flow against 272 sites of risk areas. In the case of environmental factors, it showed the high risk of debris flow under the following conditions such as soil depth of less than 30cm, west slope, altitude of 200~300 m, mountain average slope of $25{\sim}30^{\circ}$, sandy loam, igneous rocks, and composite slope. Among the rainfall factors, 50~100 mm of maximum hourly rain fall and 300 mm of maximum rain fall per day have been shown the high risk of debris flow. Furthermore, the high risk of debris flow was related to the river-bed average slope of $10{\sim}20^{\circ}$, the river-bed average width of >10 m, the small amount of debris in river-bed (less than 20% of river-bed structure), the drainage density of >$1km/km^2$, the 40~60% of area with more than $20^{\circ}$ slope, and the 40~60% of areas with risk grade 2 of landslide. In addition, forest environmental factors including the driftwood, soil erosion control structures, age-class 3, crown density (density), and mixed forest were important factors causing the high risk of debris flow.

• Journal of the Korean Geographical Society
• /
• v.40 no.6 s.111
• /
• pp.730-747
• /
• 2005

This study examines relationships between climatic factors and spatial-temporal patterns and recent changes of tropical night phenomenon(TN) occurring through nighttime stages in South Korea. Frequencies of daily TN at different times of night are extracted from long term(1973-2004) 6 hourly nighttime(9PM and 3AM) temperature and daily minimum temperature data at 61 weather stations. Temporally, the occurrences of TN are more pronounced in the evening(9PM) and during the Changma Break period(late July - early August). Spatially, the TNs in the evening frequently occur in the urbanized inland cities at low latitudes due to urban heat islands, whereas the TNs in the middle of night(3AM) or at dawn frequently appeared along the coastal areas within 30km from ocean due to the thermal inertia of ocean. By contrast, the evening(dawn) TN is not seen in the highlands whose elevation is greater than 800m(300m) along the Taebaek and Sobaek mountain ridges due to temperature lapse rates with height Correlation and multivariate regression analyses reveal that the impacts of human or physical climatic factors, such as latitude, elevation, proximity to ocean, and population density, are diverse on the frequencies of TN according to nighttime stages. Recent temporal changes of the late Changma period and intensified urbanization during the 1990s have increased the occurances of TN in urban areas. Therefore, strategies to mitigate the increasing urban TN should be prepared in the near future.

• Asian-Australasian Journal of Animal Sciences
• /
• v.13 no.7
• /
• pp.941-947
• /
• 2000

Three fistulated Malaysian local bulls were used in a $3{\times}3$ Latin square design to determine the effects of different levels of concentrate with oil palm (Elaeis guineensis Jacq.) frond (OPF) on rumen pH and $NH_3$-N concentration, and DM degradability of different fractions of OPF. Three diets namely, 60% OPF pellet and 40% concentrate (Diet 1), 50% OPF pellet and 50% concentrate (Diet 2) and 40% OPF pellets and 60% concentrate (Diet 3) were used. The levels of concentrate in the diets affected rumen pH and $NH_3$-N concentration. The pH and $NH_3$-N concentration almost in all hourly samples did not show any difference (p>0.05) among the diets except the 6 h and 9 h samples. The highest (p<0.01) $NH_3$-N concentration was obtained on Diet 3 followed by Diet 2 and Diet 1, but there was a slightly higher (p>0.05) pH on Diet 1. The $NH_3$-N concentrations of rumen liquor at 9 h sampling on Diet 1 and Diet 2 were below the critical level (50 mg/liter) required for efficient fermentation of fibrous feeds. The in sacco DM degradation of different fractions of OPF was affected by diets. The DM degradation of fractions of OPF was higher on Diet 3, which showed differences (p<0.01) with the other diets. It was found that a higher level of concentrate (60%) with OPF gave a higher rumen $NH_3$-N concentration that increased the DM degradation of OPF fractions. The results showed that OPF could support an efficient rumen function in terms of $NH_3$-N concentration and pH when ${\leq}50%$ in the diet. A higher level of OPF (>50%) does not support an efficient rumen fermentation in terms of $NH_3$-N concentration, and resulted in lower DM degradation values of the fractions. The results suggested that there is a need to supplement additional nitrogen to OPF based diets.

• Journal of Korean Society for Atmospheric Environment
• /
• v.32 no.1
• /
• pp.114-126
• /
• 2016

Many cohort studies have reported associations of individual-level long-term exposures to $PM_{10}$ and health outcomes. Individual exposures were often estimated by using exposure prediction models relying on $PM_{10}$ data measured at national regulatory monitoring sites. This study explored spatial and temporal characteristics of regulatory $PM_{10}$ measurement data in South Korea and suggested $PM_{10}$ concentration metrics as long-term exposures for assessing health effects in cohort studies. We obtained hourly $PM_{10}$ data from the National Institute of Environmental Research for 2001~2012 in South Korea. We investigated spatial distribution of monitoring sites using the density and proximity in each of the 16 metropolitan cities and provinces. The temporal characteristics of $PM_{10}$ measurement data were examined by annual/seasonal/diurnal patterns across urban background monitoring sites after excluding Asian dust days. For spatial characteristics of $PM_{10}$ measurement data, we computed coefficient of variation (CV) and coefficient of divergence (COD). Based on temporal and spatial investigation, we suggested preferred long-term metrics for cohort studies. In 2010, 294 urban background monitoring sites were located in South Korea with a site over an area of $415.0km^2$ and distant from another site by 31.0 km on average. Annual average $PM_{10}$ concentrations decreased by 19.8% from 2001 to 2012, and seasonal $PM_{10}$ patterns were consistent over study years with higher concentrations in spring and winter. Spatial variability was relatively small with 6~19% of CV and 21~46% of COD across 16 metropolitan cities and provinces in 2010. To maximize spatial coverage and reflect temporal and spatial distributions, our suggestion for $PM_{10}$ metrics representing long-term exposures was the average for one or multiple years after 2009. This study provides the knowledge of all available $PM_{10}$ data measured at national regulatory monitoring sites in South Korea and the insight of the plausible longterm exposure metric for cohort studies.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.10
• /
• pp.8-15
• /
• 2020

A forecasting method using deep learning does not have consistent results due to the differences in the characteristics of the dataset, even though they have the same forecasting models and parameters. For example, the forecasting model X optimized with dataset A would not produce the optimized result with another dataset B. The forecasting model with the characteristics of the dataset needs to be optimized to increase the accuracy of the forecasting model. Therefore, this paper proposes novel optimization steps for outlier removal, dataset classification, and a CNN-LSTM-based hyperparameter tuning process to forecast the daily power usage of a university campus based on the hourly interval. The proposing model produces high forecasting accuracy with a 2% of MAPE with a single power input variable. The proposing model can be used in EMS to suggest improved strategies to users and consequently to improve the power efficiency.

• Journal of Korean Society of Forest Science
• /
• v.77 no.2
• /
• pp.178-185
• /
• 1988

This study is carried out for analyzing the factors causing several disasters occurring around beach area in order to set a plan for an afforestation which will fulfill its function as facilities for prevention of disasters and for relaxation around beach resort at Tong-ri, Pokil-my$\bar{o}$n, Wando-gun. The results are as follows : 1. The main wind direction was summer was SE. 2. The first class in the rate of frequency of the hourly average wind speed and the instantaneous wind speed in summer were 2.1-3.0m/sec (29.2%) and 1.1-2.0m/sec (30.6%) respectively. 3. The particle sizes of the dune sands was a little small (82.5% in 0.125 to 0.25mm, $D_{50}=0.178mm$). 4. The mineral composition and the chemical components were as follows : The main mineral : Quartz The accessary minerals : Calcite, Feldspars(Orthoclase), Sericite. The chemical components : $SiO_2$ ; 75.6%. $Al_2O_3$ ; 8.1%, CaO ; 7.76%, Ign. loss ; 6.8%, MgO ; 0.23%, $K_2O$ ; 0.72%, $Na_2O$ ; 0.41%, $Fe_2O_3$ ; 0.32%. 5. The threshold friction velocity equation for the diameter class of the dune sands was $y=4.191x^{0.221}$. 6. The amount of floating salt was maximum at the point of 65m away from beach line, dropped abruptly at 135m and remained almost same on its way to the inland. The equation of the amount of floating salt was $y=28.181{\times}(-0.369^x)$. 7. The amount of extracted salt(33%) in seawater in summer was the same as the one of the nearest seawater of Korea.

• Journal of Korean Society of Forest Science
• /
• v.83 no.3
• /
• pp.322-330
• /
• 1994

The Universal Soil Loss Equation (USLE) has been widely used to predict long-term soil loss by incorporating several erosion factors, such as rainfall, soil, topography, and vegetation. This study is aimed to introduce the LISLE within geographic information system(GIS) environment. The Kwangneung Experimental Forest located in Kyongki Province was selected for the study area. Initially, twelve years of hourly rainfall records that were collected from 1982 to 1993 were processed to obtain the rainfall factor(R) value for the LISLE calculation. Soil survey map and topographic map of the study area were digitized and subsequent input values(K, L, S factors) were derived. The cover type and management factor (C) values were obtained from the classification of Landsat Thematic Mapper(CM) satellite imagery. All these input values were geographically registered over a common map coordinate with $25{\times}25m^2$ ground resolution. The USLE was calculated for every grid location by selecting necessary input values from the digital base maps. Once the LISLE was calculated, the resultant soil loss values(A) were represented by both numerical values and map format. Using GIS to run the LISLE, it is possible to pent out the exact locations where soil loss potential is high. In addition, this approach can be a very effective tool to monitor possible soil loss hazard under the situations of forest changes, such as conversion of forest lands to other uses, forest road construction, timber harvesting, and forest damages caused by fire, insect, and diseases.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.36 no.8
• /
• pp.803-809
• /
• 2012

It is necessary to predict the heating load in order to determine the optimal scheduling control of district heating systems. Heating loads are affected by many complex parameters, and therefore, it is necessary to develop an efficient, flexible, and easy to use prediction method for the heating load. In this study, simple specifications included in a building design document and the estimated temperature and humidity are used to predict the heating load on the next day. To validate the performance of the proposed method, heating load data measured from a benchmark district heating system are compared with the predicted results. The predicted outdoor temperature and humidity show a variation trend that agrees with the measured data. The predicted heating loads show good agreement with the measured hourly, daily, and monthly loads. During the heating period, the monthly load error was estimated to be 4.68%.

• Journal of Korea Water Resources Association
• /
• v.41 no.9
• /
• pp.959-967
• /
• 2008

Disaggregation techniques are widely used to transform observed daily rainfall values into hourly ones, which serve as important inputs for flood forecasting purposes. However, an important limitation with most of the existing disaggregation techniques is that they treat the rainfall process as a realization of a stochastic process, thus raising questions on the lack of connection between the structure of the models on one hand and the underlying physics of the rainfall process on the other. The present study introduces a nonlinear deterministic (and specifically chaotic) framework to study the dynamic characteristics of rainfall distributions across different temporal scales (i.e. weights between scales), and thus the possibility of rainfall disaggregation. Rainfall data from the Seoul station (recorded by the Korea Meteorological Administration) are considered for the present investigation, and weights between only successively doubled resolutions (i.e., 24-hr to 12-hr, 12-hr to 6-hr, 6-hr to 3-hr) are analyzed. The correlation dimension method is employed to investigate the presence of chaotic behavior in the time series of weights, and a local approximation technique is employed for rainfall disaggregation. The results indicate the presence of chaotic behavior in the dynamics of weights between the successively doubled scales studied. The modeled (disaggregated) rainfall values are found to be in good agreement with the observed ones in their overall matching (e.g. correlation coefficient and low mean square error). While the general trend (rainfall amount and time of occurrence) is clearly captured, an underestimation of the maximum values are found.