• Journal of The Korean Society of Agricultural Engineers
• /
• v.65 no.1
• /
• pp.15-26
• /
• 2023

This study aimed to evaluate the performance of water level classification from CCTV images in agricultural facilities such as reservoirs. Recently, the CCTV system, widely used for facility monitor or disaster detection, can automatically detect and identify people and objects from the images by developing new technologies such as a deep learning system. Accordingly, we applied the ResNet-50 deep learning system based on Convolutional Neural Network and analyzed the water level of the agricultural reservoir from CCTV images obtained from TOMS (Total Operation Management System) of the Korea Rural Community Corporation. As a result, the accuracy of water level detection was improved by excluding night and rainfall CCTV images and applying measures. For example, the error rate significantly decreased from 24.39 % to 1.43 % in the Bakseok reservoir. We believe that the utilization of CCTVs should be further improved when calculating the amount of water supply and establishing a supply plan according to the integrated water management policy.

• Journal of Korea Water Resources Association
• /
• v.39 no.8 s.169
• /
• pp.669-676
• /
• 2006

Hydrological factors, especially the spatial distribution of interpretation on precipitation is often topic of interest in studying of water resource. The popular methods such as Thiessen method, inverse distance method, and isohyetal method are limited in calculating the spatial continuity and geographical characteristics. This study was intended to overcome those limitations with improved method that will yield higher accuracy. The monthly and yearly precipitation data were produced and compared with the observed daily precipitation to find correlation between them. They were then used as secondary variables in Co-kriging method, and the result was compared with the outcome of existing methods like inverse distance method and kriging method. The comparison of the data showed that the daily precipitation had high correlation with corresponding year's average monthly amounts of precipitation and the observed average monthly amounts of precipitation. Then the result from the application of these data for a Co-kriging method confirmed increased accuracy in the modeling of spatial distribution of precipitation, thus indirectly reducing inconsistency of the spatial distribution of hydrological factors other than precipitation.

• Journal of Korean Society of Forest Science
• /
• v.105 no.4
• /
• pp.429-434
• /
• 2016

Slope failures in forest roads often result in social and economic loss as well as environmental damage. This study was carried out to assess susceptibility of slope failures of forest roads in Hongcheon-gun, Gangwon-do where many slope failures occurred after heavy rainfall in 2013 using GIS and logistic regression analysis. The results showed that sandy soil (6.616) in soil texture type had the highest susceptibility to slope failures while medium class (-3.282) in tree diameter showed the lowest susceptibility. A error matrix for both slope failure and non-slope failure area was made and a model was developed showing a classification accuracy of 74.6%. Non-slope failures area in the forest roads were classified mostly in the range of >0.7 which was higher values than the classification criteria (0.5) used by the logistic regression model. It is suggested that considering forest environment and site factors related to forest road failures would improve the accuracy in predicting susceptibility of slope failures.

• Journal of Korea Water Resources Association
• /
• v.49 no.1
• /
• pp.19-28
• /
• 2016

The objective of this study is the generation of the gridded flash flood index using the gridded hydrologic components of TOPLATS land surface model and statistic flash flood index model. The accuracy of this method is also examined in this study. The study area is the national capital region of Korea, and 38 flash flood damages had occurred from 2009 to 2012. The spatio-temporal resolutions of land surface model are 1 h and 1 km, respectively. The gridded meteorological data are generated using the inverse distance weight method with automatic weather stations (AWSs) of Korea Meteorological Administration (KMA). The hydrological components (e.g., surface runoff, soil water contents, and water table depth) of cells corresponding to the positions of 38 flood damages reasonably respond to the cell based hourly rainfalls. Under the total rainfall condition, the gridded flash flood index shows 71% to 87% from 4 h to 6 h in the lead time based on the rescue request time and 42% to 52% of accuracy at 0 h which means that the time period of the lead time is in a limited rescue request time. From these results, it is known that the gridded flash flood index using the cell based hydrological components from land surface model and the statistic flash flood index model have a capability to predict flash flood in the mountainous area.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.11
• /
• pp.5417-5422
• /
• 2013

This research was performed in order to improve the goodness of fit of paddy rice production forecasting using MODIS images and to find out appropriate explanatory variables in the forecasting model. The aim of this paper is to review the use of satellite images for the survey of paddy rice production in Korea. Many developed countries, including the United States, Australia, and Japan, have been using satellite images to produce agricultural statistics such as crop production, cultivated acreage, etc. The survey accuracy of crop production by using satellite images, however, is not satisfied in practical use. In this paper, we reviewed several methods to increase the survey accuracy of rice production statistics, gained from satellite images. Rice was selected for this study because its cultivated area and production amount could be more easily identified than other crops by using satellite images. The MODIS images were used because they involved more appropriate images to estimate and analyze rice production. This study estimated yield functions by using the NDVIs, gained from paddy rice yields and annual average isothermal lines, and the meteorological variables such as sunshine hours, rainfall, and temperature during ripening stage. As a result of yield function estimation, the goodness of fit(R-squared) for the models was shown from 0.768 to 0.891. In this study, it is noteworthy academically and practically that vegetation index(NDVIs) identified by annual average isothermal lines and meteorological variables are very useful for estimating yield functions.

• Journal of Korea Water Resources Association
• /
• v.52 no.8
• /
• pp.531-543
• /
• 2019

Constructing a reliable runoff model and reducing model runtime are important in research of real-time urban flood forecasting to reduce the repetitive flood damage. Sewer networks in the major urban basin such as Seoul are vast and complex so that it is not suitable for real-time urban flood forecasting. Therefore, the rainfall-runoff model should be simplified. However, the runoff results due to the simplification of sewer networks can vary depending on the subjectivity and simplification method of the researcher and there is a significant difference especially in 2-D inundation analysis. In this study, the sewer networks in various urban basins with different numbers and distributions of sewer networks were simplified to certain criteria. The accuracy of the simplification model according to the sewer network scale is evaluated by 2-D inundation analysis. The runoff models of Gwanak, Sillim, and Dorimcheon, frequently inundated basins were simplified based on four simplification ranges due to the cumulative drainage area set as a criterion for calculating the simplification range. This study will be expected that the inundation result of simplification models estimated through the analysis can contribute to the construction of a reasonable and accurate runoff model suitable for real-time flood forecasting.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.33 no.3
• /
• pp.101-109
• /
• 2021

Wave observations using a marine X-band radar are conducted by analyzing the backscattered radar signal from sea surfaces. Wave parameters are extracted using Modulation Transfer Function obtained from 3D wave number and frequency spectra which are calculated by 3D FFT of time series of sea surface images (42 images per minute). The accuracy of estimation of the significant wave height is, therefore, critically dependent on the quality of radar images. Wave observations during Typhoon Maysak and Haishen in the summer of 2020 show large errors in the estimation of the significant wave heights. It is because of the deteriorated radar images due to raindrops falling on the sea surface. This paper presents the algorithm developed to increase the accuracy of wave heights estimation from radar images by adopting convolution neural network(CNN) which automatically classify radar images into rain and non-rain cases. Then, an algorithm for deriving the Hs is proposed by creating different ANN models and selectively applying them according to the rain or non-rain cases. The developed algorithm applied to heavy rain cases during typhoons and showed critically improved results.

• Journal of Korea Water Resources Association
• /
• v.54 no.10
• /
• pp.795-805
• /
• 2021

In this study, after developing an LSTM-based deep learning model for estimating daily runoff in the Soyang River Dam basin, the accuracy of the model for various combinations of model structure and input data was investigated. A model was built based on the database consisting of average daily precipitation, average daily temperature, average daily wind speed (input up to here), and daily average flow rate (output) during the first 12 years (1997.1.1-2008.12.31). The Nash-Sutcliffe Model Efficiency Coefficient (NSE) and RMSE were examined for validation using the flow discharge data of the later 12 years (2009.1.1-2020.12.31). The combination that showed the highest accuracy was the case in which all possible input data (12 years of daily precipitation, weather temperature, wind speed) were used on the LSTM model structure with 64 hidden units. The NSE and RMSE of the verification period were 0.862 and 76.8 m³/s, respectively. When the number of hidden units of LSTM exceeds 500, the performance degradation of the model due to overfitting begins to appear, and when the number of hidden units exceeds 1000, the overfitting problem becomes prominent. A model with very high performance (NSE=0.8~0.84) could be obtained when only 12 years of daily precipitation was used for model training. A model with reasonably high performance (NSE=0.63-0.85) when only one year of input data was used for model training. In particular, an accurate model (NSE=0.85) could be obtained if the one year of training data contains a wide magnitude of flow events such as extreme flow and droughts as well as normal events. If the training data includes both the normal and extreme flow rates, input data that is longer than 5 years did not significantly improve the model performance.

• Journal of Korea Water Resources Association
• /
• v.55 no.11
• /
• pp.931-939
• /
• 2022

In Korea, about two-thirds of the precipitation is concentrated in the summer season, so the problem of turbidity in the summer flood season varies from year to year. Concentrated rainfall due to abnormal rainfall and extreme weather is on the rise. The inflow of turbidity caused a sudden increase in turbidity in the water, causing a problem of turbidity in the dam reservoir. In particular, in Korea, where rivers and dam reservoirs are used for most of the annual average water consumption, if turbidity problems are prolonged, social and environmental problems such as agriculture, industry, and aquatic ecosystems in downstream areas will occur. In order to cope with such turbidity prediction, research on turbidity modeling is being actively conducted. Flow rate, water temperature, and SS data are required to model turbid water. To this end, the national measurement network measures turbidity by measuring SS in rivers and dam reservoirs, but there is a limitation in that the data resolution is low due to insufficient facilities. However, there is an unmeasured period depending on each dam and weather conditions. As a sensor for measuring turbidity, there are Optical Backscatter Sensor (OBS) and YSI, and a sensor for measuring SS uses equipment such as Laser In-Situ Scattering and Transmissometry (LISST). However, in the case of such a high-tech sensor, there is a limit due to the stability of the equipment. Therefore, there is an unmeasured period through analysis based on the acquired flow rate, water temperature, SS, and turbidity data, so it is necessary to develop a relational expression to calculate the SS used for the input data. In this study, the AEM3D model used in the Water Resources Corporation SURIAN system was used to improve the accuracy of prediction of turbidity through the turbidity-SS relationship developed based on the measurement data near the dam outlet.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.20 no.3
• /
• pp.262-276
• /
• 2018

Models to predict Leaf Wetness Duration (LWD) were evaluated using the observed meteorological and dew data at the 11 citrus orchards in Jeju, South Korea from 2016 to 2017. The sensitivity and the prediction accuracy were evaluated with four models (i.e., Number of Hours of Relative Humidity (NHRH), Classification And Regression Tree/Stepwise Linear Discriminant (CART/SLD), Penman-Monteith (PM), Deep-learning Neural Network (DNN)). The sensitivity of models was evaluated with rainfall and seasonal changes. When the data in rainy days were excluded from the whole data set, the LWD models had smaller average error (Root Mean Square Error (RMSE) about 1.5hours). The seasonal error of the DNN model had the similar magnitude (RMSE about 3 hours) among all seasons excluding winter. The other models had the greatest error in summer (RMSE about 9.6 hours) and the lowest error in winter (RMSE about 3.3 hours). These models were also evaluated by the statistical error analysis method and the regression analysis method of mean squared deviation. The DNN model had the best performance by statistical error whereas the CART/SLD model had the worst prediction accuracy. The Mean Square Deviation (MSD) is a method of analyzing the linearity of a model with three components: squared bias (SB), nonunity slope (NU), and lack of correlation (LC). Better model performance was determined by lower SB and LC and higher NU. The results of MSD analysis indicated that the DNN model would provide the best performance and followed by the PM, the NHRH and the CART/SLD in order. This result suggested that the machine learning model would be useful to improve the accuracy of agricultural information using meteorological data.