• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.135-135
• /
• 2021

In water resources management, rainfall prediction with high accuracy is still one of controversial issues particularly in countries facing heavy rainfall during wet seasons in the monsoon climate. The aim of this study is to develop an artificial neural network (ANN) for predicting future six months of rainfall data (from April to September 2020) from daily meteorological data (from 1971 to 2019) such as rainfall, temperature, wind speed, and humidity at Seoul, Korea. After normalizing these data, they were trained by using a multilayer perceptron (MLP) as a class of the feedforward ANN with 15,000 neurons. The results show that the proposed method can analyze the relation between meteorological datasets properly and predict rainfall data for future six months in 2020, with an overall accuracy over almost 70% and a root mean square error of 0.0098. This study demonstrates the possibility and potential of MLP's applications to predict future daily rainfall patterns, essential for managing flood risks and protecting water resources.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.2
• /
• pp.63-74
• /
• 2019

The objective of this study was to evaluate the influence of rainfall observation network on daily dam inflow using artificial neural networks(ANNs). Chungju Dam and Soyangriver Dam were selected for the study watershed. Rainfall and dam inflow data were collected as input data for construction of ANNs models. Five ANNs models, represented by Model 1 (In watershed, point rainfall), Model 2 (All in the Thiessen network, point rainfall), Model 3 (Out of watershed in the Thiessen network, point rainfall), Model 1-T (In watershed, area mean rainfall), Model 2-T (All in the Thiessen network, area mean rainfall), were adopted to evaluate the influence of rainfall observation network. As a result of the study, the models that used all station in the Thiessen network performed better than the models that used station only in the watershed or out of the watershed. The models that used point rainfall data performed better than the models that used area mean rainfall. Model 2 achieved the highest level of performance. The model performance for the ANNs model 2 in Chungju dam resulted in the $R^2$ value of 0.94, NSE of 0.94 $NSE_{ln}$ of 0.88 and PBIAS of -0.04 respectively. The model-2 predictions of Soyangriver Dam with the $R^2$ and NSE values greater than 0.94 were reasonably well agreed with the observations. The results of this study are expected to be used as a reference for rainfall data utilization in forecasting dam inflow using artificial neural networks.

• 한국방재학회:학술대회논문집
• /
• 2008.02a
• /
• pp.101-104
• /
• 2008

In this study, rainfall adjust and forecasting using artificial neural network(ANN) which includes a correlation coefficient is application in Seoul region. It analyzed one-hour rainfall data which has been reported in 25 region in seoul during from 2000 to 2006 at rainfall observatory by AWS. The ANN learning algorithm apply for input data that each region using cross-correlation will use the highest correlation coefficient region. In addition, rainfall adjust analyzed the minimum error based on correlation coefficient and determination coefficient related to the input region. ANN model used back-propagation algorithm for learning algorithm. In case of the back-propagation algorithm, many attempts and efforts are required to find the optimum neural network structure as applied model. This is calculated similar to the observed rainfall that the correlation coefficient was 0.98 in missing rainfall adjust at 10 region. As a result, ANN model has been for suitable for rainfall adjust. It is considered that the result will be more accurate when it includes climate data affecting rainfall.

• Journal of the Korean earth science society
• /
• v.31 no.5
• /
• pp.465-474
• /
• 2010

Real-time rainfall monitoring is of great practical importance over the highly populated Indochina area, which is prone to natural disasters, in particular in association with rainfall. With the goal of d etermining near real-time half-hourlyrain estimates from satellite, the three-layer, artificial neural networks (ANN) approach was used to train the brightness temperatures at 6.7, 11, and $12-{\mu}m$ channels of the Japanese geostationary satellite MTSAT against passive microwavebased rain rates from Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) and TRMM Precipitation Radar (PR) data for the June-September 2005 period. The developed model was applied to the MTSAT data for the June-September 2006 period. The results demonstrate that the developed algorithm is comparable to the PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks) results and can be used for flood monitoring across the Indochina area on a half-hourly time scale.

• Journal of Korean Society on Water Environment
• /
• v.32 no.3
• /
• pp.310-317
• /
• 2016

It has been a big problem to estimate rainfall for the studies of mud-debris flows because the estimated rainfall from the nearest AWS (Automatic Weather Station) can tend to be quite inaccurate at individual sites. This study attempts to improve this problem through accurate rainfall depth estimation by applying an artificial neural network with radar rainfall data. For this, three models were made according to utilizing methodologies of rainfall data. The first model uses the nearest rainfall, observing the site from an ungauged site. The second uses only radar rainfall data and the third model integrates the above two models using both radar and observed rainfall at the sites around the ungauged site. This methodology was applied to the metropolitan area in Korea. It appeared as though the third model improved rainfall estimations by the largest margin. Therefore, the proposed methodology can be applied to forecast mud-debris flows in ungageed sites.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.12
• /
• pp.828-835
• /
• 2019

An artificial rainfall system is used widely as a research tool for generating model experiment data. Artificial rainfall devices have been used in many studies, but studies of the rainfall distribution are not considered as important issues. To simulate various rainfall characteristics, it should be possible to simulate from low to high intensity, and the homogeneity of the rainfall distribution should be ensured. In this study, the maximum rainfall intensity was set to 130mm/hr and controlled by 10mm/hr. In addition, the aim was to secure a uniform coefficient value of 80% or more. To this end, rainfall tests were performed according to the nozzle type, diameter, position, and pump pressure. The rainfall test showed that the circular nozzle was suitable, and the nozzle size was 1.9mm and 1.4mm. The optimal pump pressure was found to be 3~6kg/㎠. The rainfall intensity tended to increase linearly with increasing pump pressure. Based on the rainfall test results, a rainfall control manual was produced with variables, such as pump pressure, nozzle type, and number of nozzles. As a result of rainfall verification, rainfall intensity showed a 3.1% error with a uniformity coefficient of 86%.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.45 no.2
• /
• pp.45-57
• /
• 2003

An artificial neural network model was developed to analyze and forecast Short-term river runoff from the Naju watershed, in Korea. Error back propagation neural networks (EBPN) of hourly rainfall and runoff data were found to have a high performance In forecasting runoff. The number of hidden nodes were optimized using total error and Bayesian information criterion. Model forecasts are very accurate (i.e., relative error is less than 3% and $R^2$is greater than 0.99) for calibration and verification data sets. Increasing the time horizon for application data sets, thus mating the model suitable for flood forecasting. decreases the accuracy of the model. The resulting optimal EBPN models for forecasting hourly runoff consists of ten rainfall and four runoff data(ANN0410 model) and ten rainfall and ten runoff data(ANN1010 model). Performances of the ANN0410 and ANN1010 models remain satisfactory up to 6 hours (i.e., $R^2$is greater than 0.92).

• Proceedings of the Korean Geotechical Society Conference
• /
• 2001.03a
• /
• pp.199-206
• /
• 2001

Landslides generally occur due to influences of the internal and external factors. Internal factors include ground characteristics, terrain and so on. External factors can also be divided into natural factors such as rainfall, ground water, earthquake and so on, and artificial factors resulting from cutting and embankments. Among these factors, rainfall becomes the most important external factors by means of which landslides occur in Korea. To appropriately deal with tile effects of pore water pressures due to rainfall, the method using the pore water pressure ratio(r$\_$u/) is generally applied in slope stability analysis or the design of slope reinforcements. Since tire value of r,, is in general not constant over the whole cross section, in most slope stability analyses the average values are used with little loss in accuracy. However, determination of the average values of r$\_$u/ to applied in the design is difficult problem. Therefore, in this study, tile average values of r$\_$u/ according to the intensity of rainfall and slope inclination is suggested based on results of the small scaled model tests using the artificial rainfall apparatus. It is found from the model tests that the average values of r$\_$u/ is about 0.07∼0.18(in case of the intensity of rainfall is 50mm/hr.), about 0.10∼0.28(in case of the intensity of rainfall is 100mm/hr.), and about 0.10∼0.33(in case of the intensity of rainfall is 150mm/hr.).

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.2
• /
• pp.227-235
• /
• 2018

In this study, we developed a relational formula for observing high - resolution rainfall using vehicle rain sensor. The vehicle rain sensor consists of eight channels. Each channel generates a sensor signal by detecting the amount of rainfall on the windshield of the vehicle when rainfall occurs. The higher the rainfall, the lower the sensor signal is. Using these characteristics of the sensor signal generated by the rain sensor, we developed a relational expression. In order to generate specific rainfall, an artificial rainfall generator was constructed and the change of the sensor signal according to the variation of the rainfall amount in the artificial rainfall generator was analyzed. Among them, the optimal sensor channel which reflects various rainfall amounts through the sensitivity analysis was selected. The sensor signal was generated in 5 minutes using the selected channel and the representative values of the generated 5 - minute sensor signals were set as the average, 25th, 50th, and 75th quartiles. The calculated rainfall values were applied to the actual rainfall data using the constructed relational equation and the calculated rainfall amount was compared with the rainfall values observed at the rainfall station. Although the reliability of the relational expression was somewhat lower than that of the data of the verification result data, it was judged that the experimental data of the residual range was insufficient. The rainfall value was calculated by applying the developed relation to the actual rainfall, and compared with the rainfall value generated by the ground rainfall observation instrument observed at the same time to verify the reliability. As a result, the rain sensor showed a fine rainfall of less than 0.5 mm And the average observation error was 0.36mm.

• Geomechanics and Engineering
• /
• v.22 no.2
• /
• pp.153-163
• /
• 2020

Laboratory experiments were conducted with two different soil conditions to investigate rainfall infiltration characteristics. The soil layer materials that were tested were weathered granite soil and weathered gneiss soil. Artificial rainfall of 80 mm/hr was reproduced through the use of a rainfall device, and the volumetric water content and matric suction were measured. In the case of the granite soil, the saturation velocity and the moving direction of the wetting front were fast and upward, respectively, whereas in the case of the weathered gneiss soil, the velocity and direction were slow and downward, respectively. Rainfall penetrated and saturated from the bottom to the top as the hydraulic conductivity of the granite soil was higher than the infiltration capacity of the artificial rainfall. In contrast, as the hydraulic conductivity of the gneiss soil was lower than the infiltration capacity of the rainfall, ponding occurred on the surface: part of the rainfall first infiltrated, with the remaining rainfall subsequently flowing out. The unsaturated hydraulic conductivity function of weathered soils was determined and analyzed with matric suction and the effective degree of saturation.