• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.4B
• /
• pp.389-398
• /
• 2006

The present study compositely applied the self-organizing map (SOM), which is a kind of artificial neural networks (ANNs), and the back propagation algorithm (BPA) for the rainfall-runoff prediction model taking account of the irregular variation of the spatiotemporal distribution of rainfall. To solve the problems from the previous studies on ANNs, such as the overestimation of low flow during the dry season, the underestimation of runoff during the flood season and the persistence phenomenon, in which the predicted values continuously represent the preceding runoffs, we introduced SOM theory for the preprocessing in the prediction model. The theory is known that it has the pattern classification ability. The method proposed in the present research initially includes the classification of the rainfall-runoff relationship using SOM and the construction of the respective models according to the classification by SOM. The individually constructed models used the data corresponding to the respectively classified patterns for the runoff prediction. Consequently, the method proposed in the present study resulted in the better prediction ability of runoff than that of the past research using the usual application of ANNs and, in addition, there were no such problems of the under/over-estimation of runoff and the persistence.

• Journal of Wetlands Research
• /
• v.15 no.1
• /
• pp.115-125
• /
• 2013

This study established a methodology for the application of downscaling technique in a mountainous area having large spatial variations of rainfall and tried to estimate the change of rainfall characteristics in the future under climate change using the established method. The Namhan river basin, which is in the mountainous area of the Korean peninsula, has been chosen as the study area. Artificial Neural Network - Simple Kriging with varying local means (ANN-SKlm) has been built by combining artificial neural network, which is one of the general downscaling techniques, and SKlm technique, which can reflect the geomorphologic characteristics like elevation of the study area. The evaluation of SKlm technique was done by using the monthly rainfalls at six weather stations which KMA(Korea Meteorological Administration) is managing in the basin. The ANN-SKlm technique was compared with the Thiessen technique and ordinary kriging(OK) technique. According to the evaluation result of each technique the SKlm technique showed the best result.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.63 no.1
• /
• pp.103-116
• /
• 2021

Analysis of runoff is substantial for effective water management in the watershed. Runoff occurs by reaction of a watershed to the rainfall and has non-linearity and uncertainty due to the complex relation of weather and watershed factors. ANN (Artificial Neural Network), which learns from the data, is one of the machine learning technique known as a proper model to interpret non-linear data. The performance of ANN is affected by the ANN's structure, the number of hidden layer nodes, learning rate, and activation function. Especially, the activation function has a role to deliver the information entered and decides the way of making output. Therefore, It is important to apply appropriate activation functions according to the problem to solve. In this paper, ANN models were constructed to estimate runoff with different activation functions and each model was compared and evaluated. Sigmoid, Hyperbolic tangent, ReLU (Rectified Linear Unit), ELU (Exponential Linear Unit) functions were applied to the hidden layer, and Identity, ReLU, Softplus functions applied to the output layer. The statistical parameters including coefficient of determination, NSE (Nash and Sutcliffe Efficiency), NSEln (modified NSE), and PBIAS (Percent BIAS) were utilized to evaluate the ANN models. From the result, applications of Hyperbolic tangent function and ELU function to the hidden layer and Identity function to the output layer show competent performance rather than other functions which demonstrated the function selection in the ANN structure can affect the performance of ANN.

• The Journal of Engineering Geology
• /
• v.20 no.1
• /
• pp.61-70
• /
• 2010

In case of groundwater recharge estimation using water table fluctuation method, specific yield affects the accuracy and confidence level of recharge rate. Nevertheless, there have been few studies on the method for the accurate estimation of specific yield in Korea. Specific yield estimated from the relationship between rainfall and groundwater levels is reasonable compared to the other methods. However, lots of factors such as artificial pumping, evapotranspiration by the plants, and a sudden increase in water levels by a heavy rainfall can affect the pattern of groundwater levels' fluctuation and make an over-estimated or under-estimated specific yield. This study obtained a reasonable specific yield by using a daily or 12 hourly average of rainfall and groundwater levels measured in a dry season.

• Korean Journal of Ecology and Environment
• /
• v.42 no.4
• /
• pp.487-494
• /
• 2009

We compared spatial and temporal variations of water chemistry between high-flow year ($HF_y$) and low-flow year ($LF_y$) in an artificial lentic ecosystem of Daechung Reservoir. The differences in the rainfall distributions explained the variation of the annual inflow and determined flow characteristics and water residence time and modified chemical and biological conditions, based on TP, suspended solids, and chlorophylla, resulting in changes of ecological functions. The intense rainfall and inflow from the watershed resulted in partial disruption of thermal structure in the metalimnion depth, ionic dilution, high TP, and high suspended solids. This condition produced a reduced chlorophyll-a in the headwaters due to low light availability and rapid flushing. In contrast, reduced inflow and low rainfall by drought resulted in strong thermal difference between the epilimnion and hypolimnion, low inorganic solids, high total dissolved solids, and low phosphorus in the ambient water. The riverine conditions dominated the hydrology in the monsoon of $HF_y$ and lacustrine conditions dominated in the $HF_y$. Overall data suggest that effective managements of the flow from the watershed may have an important role in the eutrophication processes.

• 한국방재학회:학술대회논문집
• /
• 2008.02a
• /
• pp.369-372
• /
• 2008

Recently, landslides have frequently occurred on natural slopes during periods of intense rainfall. With a rapidly increasing population on or near steep terrain in Korea, landslides have become one of the most significant natural hazards. Thus, it is necessary to protect people from landslides and to minimize the damage of houses, roads and other facilities. To accomplish this goal, many landslide prediction methods have been developed in the world. In this study, a simple landslide prediction system that enables people to escape the endangered area is introduced. The system is focused to debris flows which happen frequently during periods of intense rainfall. The system is based on the wireless sensor network (WSN) that is composed of sensor nodes, gateway, and server system. Sensor nodes comprising a sensing part and a communication part are developed to detect ground movement. Sensing part is designed to measure inclination angle and acceleration accurately, and communication part is deployed with Bluetooth (IEEE 802.15.1) module to transmit the data to the gateway. To verify the feasibility of this landslide prediction system, a series of experimental studies was performed at a small-scale earth slope equipped with an artificial rainfall dropping device. It is found that sensing nodes installed at slope can detect the ground motion when the slope starts to move. It is expected that the landslide prediction system by wireless senor network can provide early warnings when landslides such as debris flow occurs.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.143-143
• /
• 2022

Precipitation plays an essential role in water resources management and disaster prevention. Therefore, the understanding related to spatiotemporal characteristics of rainfall is necessary. Nowadays, highly accurate precipitation is mainly obtained from gauge observation systems. However, the density of gauge stations is a sparse and uneven distribution in mountainous areas. With the proliferation of technology, satellite-based precipitation sources are becoming increasingly common and can provide rainfall information in regions with complex topography. Nevertheless, satellite-based data is that it still remains uncertain. To overcome the above limitation, this study aims to take the strengthens of machine learning to generate a new reanalysis of precipitation data by fusion of multiple satellite precipitation products (SPPs) with gauge observation data. Several machine learning algorithms (i.e., Random Forest, Support Vector Regression, and Artificial Neural Network) have been adopted. To investigate the robustness of the new reanalysis product, observed data were collected to evaluate the accuracy of the products through Kling-Gupta efficiency (KGE), probability of detection (POD), false alarm rate (FAR), and critical success index (CSI). As a result, the new precipitation generated through the machine learning model showed higher accuracy than original satellite rainfall products, and its spatiotemporal variability was better reflected than others. Thus, reanalysis of satellite precipitation product based on machine learning can be useful source input data for hydrological simulations in ungauged river basins.

• Korean Journal of Hydrosciences
• /
• v.3
• /
• pp.11-29
• /
• 1992

Recently, rapid industrialization, urbanization and higher living standards accelerate to increase groundwater consumption resulting in continuously dropping groundwater elevations. To maintain enough groundwater volume without dropping groundwater elevations, the proper groundwater rechatge is necessary. The groundwater rechatge can be classified into two categories which are natural rechatge and artiticial rechatge. Even though the natural rechatge through by dired infiltration from the rainfall is desirable, the artificial groundwater rechatge is necessaty when the increment of groundwater consumption exceeds natural recharge rate. Well method and scattering method are utilized as artificial rechatging method, a severe disadvantage, which is the reduction of the void of soil surface, is indicated in the well method. Recently, the underground piping method, which is a scattering method, is receiving increasing attention as a proper recharging method. The method is indirectly to supply water to the underground using an underground piping system. Therefore, the void of soil surface is not severely reduced and better infiltration rate can be achieved. In this paper, the artificial groundwater rechatge using underground piping method is investigated through experiments and numerical analysis. The influence of the groundwater by underground piping method is evaluated through comparing recharging heights. Good agreements between experiments and numerical analysis are obtained and the artificial groundwater recharge by underground piping method is well tested and verified.

• Korean Journal of Hydrosciences
• /
• v.4
• /
• pp.1-9
• /
• 1993

It is a major objective for the management and operation of water resources system to forecast streamflows. The applicability of artificial neural network model to hydrologic system is analyzed and the performance is compared by statistical method with observed. Multi-layered perception was used to model rainfall-runoff process at Pyung Chang River Basin in Korea. The neural network model has the function of learning the process which can be trained with the error backpropagation (EBP) algorithm in two phases; (1) learning phase permits to find the best parameters(weight matrix) between input and output. (2) adaptive phase use the EBP algorithm in order to learn from the provided data. The generalization results have been obtained on forecasting the daily and hourly streamflows by assuming them with the structure of ARMA model. The results show validities in applying to hydrologic forecasting system.

• Journal of the Korean Geotechnical Society
• /
• v.29 no.1
• /
• pp.49-59
• /
• 2013

Heavy rainfall intensity may cause shallow slope failures and debris flow by rill erosion and scour on land surface. The paper represents the difference between native soil (weathered soil) and reinforced soil, which is mixed by hardening agent with flyash as main material, for investigating experimental findings of rill erosion and erosion. Results obtained from artificial rainfall simulator show that erosion rate of reinforced soil mixed with hardening agent is reduced by 20% because an amount of eroded soil on slope surface is inversely proportional to the increase of soil strength. For example, rainfall of 45mm (at the elapsed time of 25mins in rainfall intensity of 110mm/hr) triggers rill erosion on native soil surface, but the rill erosion on reinforced soil surface does not even occur at 330mm rainfall (at the elapsed time of 3hrs in rainfall intensity of 110mm/hr). As a result of slope stability analysis, it was found that the construction method for reinforced soil surface would be more economical, easy and fast construction technology than conventional reinforcement method.