• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.40 no.3
• /
• pp.303-314
• /
• 2020

The hydrological characteristics of watersheds and hydraulic systems of urban and river floods are highly nonlinear and contain uncertain variables. Therefore, the predicted time series of rainfall-runoff data in flood analysis is not suitable for existing neural networks. To overcome the challenge of prediction, a NARX (Nonlinear Autoregressive Exogenous Model), which is a kind of recurrent dynamic neural network that maximizes the learning ability of a neural network, was applied to forecast a flood in real-time. At the same time, NARX has the characteristics of a time-delay neural network. In this study, a hydrological model was constructed for the Taehwa river basin, and the NARX time-delay parameter was adjusted 10 to 120 minutes. As a result, we found that precise prediction is possible as the time-delay parameter was increased by confirming that the NSE increased from 0.530 to 0.988 and the RMSE decreased from 379.9 ㎥/s to 16.1 ㎥/s. The machine learning technique with NARX will contribute to the accurate prediction of flow rate with an unexpected extreme flood condition.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.22 no.4
• /
• pp.39-58
• /
• 2019

This study performed the dam watershed storm runoff modeling using GPM(Global Precipitation Measurement) satellite rain and KIMSTORM2(KIneMatic wave STOrm Runoff Model 2) distributed model. For YongdamDam watershed(930㎢), three heavy rain events of 25th August 2014, 11th September 2017, and 26th June 2018 were selected and tested for 4 cases of spatial rainfalls such as (a) Kriging interpolated data using ground observed data at 7 stations, (b) original GPM data, (c) GPM corrected by CM(Conditional Merging), and GPM corrected by GDA(Geographical Differential Analysis). For the 4 kinds of data(Kriging, GPM, CM-GPM, and GDA-GPM), the KIMSTORM2 was calibrated respectively using the observed flood discharges at 3 water level gauge stations(Cheoncheon, Donghyang, and Yongdam) with parameters of initial soil moisture contents, stream Manning's roughness coefficient, and effective hydraulic conductivity. The total average Nash-Sutcliffe efficiency(NSE) for the 3 events and 3 stations was 0.94, 0.90, 0.94, and 0.94, determination coefficient(R²) was 0.96, 0.92, 0.97 and 0.96, the volume conservation index(VCI) was 1.03, 1.01, 1.03 and 1.02 for Kriging, GPM, CM-GPM, and GDA-GPM applications respectively. The CM-GPM and GDA-GPM showed better results than the original GPM application for peak runoff and runoff volume simulations, and they improved NSE, R², and VCI results.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.294-294
• /
• 2019

The system for predicting flood of river at Flood Control Office is made up of a rainfall-runoff model and FLDWAV model. This system is mainly operating to predict the excess of the flood watch or warning level at flood forecast points. As the demand for information of the management and operation of riverside, which is being used as a waterfront area such as parks, camping sites, and bike paths, high-level forecasts of watch and warning at certain points are required as well as production of lowland flood forecast information that is used as a waterfront within the river. In this study, a technology to produce flood forecast information in lowland areas of the river used as a waterfront was developed. Based on the results of the 1D hydraulic analysis, a model for performing spatial operations based on high resolution grid was constructed. A model was constructed for Andong district, and the inundation conditions and level were analyzed through a virtual outflow scenarios of Andong and Imha Dam.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.4
• /
• pp.17-24
• /
• 2018

The HEC-HMS model was applied to identify the rainfall-runoff processes for the Anseongchun basin, where the lower part of the stream has been damaged severely by tropical storms in the past. Modeling processes include incorporating with the SCS-CN model for loss, Clark's UH model for transformation, exponential recession model for baseflow, and Muskingum model for channel routing. The parameters were calibrated through an optimization technique using a trial and error method. Sensitivity analysis after calibration was performed to understand the effects of parameters, such as the time of concentration, storage coefficient, and base flow related constants. Two storm water events were simulated by the model and compared with the corresponding observations. Good accuracy in predicting the runoff volume, peak flow, and the time to peak flow was achieved using the selected methods. The results of this study can be used as a useful tool for decision makers to determine a master plan for regional flood control management.

• Journal of Environmental Science International
• /
• v.25 no.6
• /
• pp.799-815
• /
• 2016

In this study, according to the reference setting based on the runoff video of 9:00 where the highest water level of 3.94 m has been recorded during the runoff of Cheon-mi Stream in Jeju Island by the attack of Typhoon no. 16 Sanba on September $17^{th}$, 2012, the error rate of long-distance and short-distance velocimetry and real-distance change rate by input error have been calculated and the input range value of reference point by stream has been suggested. In the reference setting process, if a long-distance reference point input error occurs, the real-distance change rate of 0.35 m in the x-axis direction and 1.35 m in y-axis direction is incurred by the subtle input error of 2~11 pixels, and if a short-distance reference point input error occurs, the real-distance change rate of 0.02 m in the x-axis direction and 0.81 m in y-axis direction is incurred by the subtle input error of 1~11 pixels. According to the long-distance reference point setting variable, the velocity error rate showed the range of fluctuation of at least 14.36% to at most 76.06%, and when calculating flux, it showed a great range of fluctuation of at least 20.48% to at most 78.81%.

• Journal of Korea Water Resources Association
• /
• v.38 no.10 s.159
• /
• pp.895-906
• /
• 2005

The diversity of observed hydrologic data and the development of geographic information system leads significant progress for developing distributed runoff models in the world. One of the typical examples is TOPMODEL, but the spatial coverage of its application Is limited on small headwater basins. The purpose of this study attempts to overcome its limitation and consequently develops a semi-distributed TOPMODEL. The developed model is composed of two components: a watershed runoff component for a lumped representation of hydrologic runoff process on the catchment scale and a kinematic wave type hydraulic channel routing component lot routing the catchment outflows. The application basin is the $2,703km^2$ upper Soyang dam site and several daily and hourly events are selected for model calibrations and verifications. The model parameters are estimated on 1990 daily event. The model performance on correlation coefficient between observed and computed flows are above 0.90 for the verification events. It is concluded that the developed model in this study can be used for flood analysis in large drainage basins.

• Journal of Korean Society of Disaster and Security
• /
• v.14 no.4
• /
• pp.9-15
• /
• 2021

The frequency of flash floods in mountainous areas is increasing due to the abnormal weather that occurs increasingly in the recent, and it causes human and material damages is increasing. Various plans for disaster mitigation have been established, but artificial plans such as raising embankment and dredging operation are inappropriate for valleys and rivers in national parks that prioritize nature protection. In this study, flood risk assessment was conducted for Gyeryongsan National Park in Korea using the WMS (Watershed Modeling System)which is rainfall runoff model for valleys and rivers in the catchment. As the result, it was simulated that it is flooding in three sub-catchments (Jusukgol, Sutonggol, Dinghaksa) of a total in Gyeryongsan National Park when rainfall over the 50 years return period occurs, and it was confirmed that the risk of trails and facilities what visitors are using was high. The risk of trails in national parks was quantitatively presented through the results of this study, and we intend to present the safe management guidelines of national parks in the future.

• Ecology and Resilient Infrastructure
• /
• v.3 no.1
• /
• pp.1-7
• /
• 2016

River restoration has recently progressed in consideration of ecological functions along with flood controls and conservation. For river restorations that consider ecological health and diversity, it is important to contemplate the recovery of hydraulic and hydrologic connectivity in isolated spaces by longitudinal structures. In this study, as a first step for the provision of hydraulic and hydrologic data, which is necessary for the ecological connection analysis in isolated spaces, we developed a one-dimensional numerical model for rainfall runoff and channel routing and applied it to the Cheongmi watershed. The developed numerical model can simulate hydraulic and hydrologic analysis at the same time using the rainfall data. Numerical results were compared with observed data and other numerical results. As a result, a very reasonable agreement is observed. The results of this study will be improved so that the long-term hydrologic and hydraulic analysis is possible to predict ecological change.

• Journal of Korea Water Resources Association
• /
• v.49 no.4
• /
• pp.335-346
• /
• 2016

This study produced the parameter maps of the Modified Bartlett-Lewis Rectangular Pulse (MBLRP) stochastic rainfall generation model across South Korea and developed and validated the web application that automates the process of rainfall generation based on the produced parameter maps. To achieve this purpose, three deferent sets of parameters of the MBLRP model were estimated at 62 ground gage locations in South Korea depending on the distinct purpose of the synthetic rainfall time series to be used in hydrologic modeling (i.e. flood modeling, runoff modeling, and general purpose). The estimated parameters were spatially interpolated using the Ordinary Kriging method to produce the parameter maps across South Korea. Then, a web application has been developed to automate the process of synthetic rainfall generation based on the parameter maps. For validation, the synthetic rainfall time series has been created using the web application and then various rainfall statistics including mean, variance, autocorrelation, probability of zero rainfall, extreme rainfall, extreme flood, and runoff depth were calculated, then these values were compared to the ones based on the observed rainfall time series. The mean, variance, autocorrelation, and probability of zero rainfall of the synthetic rainfall were similar to the ones of the observed rainfall while the extreme rainfall and extreme flood value were smaller than the ones derived from the observed rainfall by the degree of 16%-40%. Lastly, the web application developed in this study automates the entire process of synthetic rainfall generation, so we expect the application to be used in a variety of hydrologic analysis needing rainfall data.

• Journal of Korea Water Resources Association
• /
• v.45 no.1
• /
• pp.65-73
• /
• 2012

Recently climate change and increase of surface runoff caused the urban flooding. Traditional way of dealing with urban flooding has been to increase the sewer capacity or construction of pumping stations, however, it is practically almost impossible because of time, money and traffic problems. Multipurpose DRMS (Decentralized Rainwater Management System) is a new paradigm proposed and recommended by NEMA (National Emergency Management Agency) for both flood control and water conservation. Suwon City has already enacted the ordinance on sound water cycle management by DRMS. In this study, a flood prone area in Suwon is selected and analysis of DRMS has been made using XP-SWMM for different scenarios of RT installation with same total rainwater tank volume and location. Installing one rainwater tank of 3,000$m^3$ can reduce the peak flow rate by 15.5%. Installing six rainwater tanks of 500$m^3$ volume in the area can reduce the peak flow rate by 28%. Three tanks which is concentrated in the middle region can reduce peak rate more than evenly distributed tanks. The method and results found from this study can be used for the design and performance prediction of DRMS at a flood prone area by supplementing the existing sewer system without increase of the sewer capacity.