• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.147-147
• /
• 2018

In this article, we use an open source software library: TensorFlow, developed for the purposes of conducting very complex machine learning and deep neural network applications. However, the system is general enough to be applicable in a wide variety of other domains as well. The proposed model based on a deep neural network model, LSTM (Long Short-Term Memory) to predict the river water level at Okcheon Station of the Guem River without utilization of rainfall - forecast information. For LSTM modeling, the input data is hourly water level data for 15 years from 2002 to 2016 at 4 stations includes 3 upstream stations (Sutong, Hotan, and Songcheon) and the forecasting-target station (Okcheon). The data are subdivided into three purposes: a training data set, a testing data set and a validation data set. The model was formulated to predict Okcheon Station water level for many cases from 3 hours to 12 hours of lead time. Although the model does not require many input data such as climate, geography, land-use for rainfall-runoff simulation, the prediction is very stable and reliable up to 9 hours of lead time with the Nash - Sutcliffe efficiency (NSE) is higher than 0.90 and the root mean square error (RMSE) is lower than 12cm. The result indicated that the method is able to produce the river water level time series and be applicable to the practical flood forecasting instead of hydrologic modeling approaches.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.101-101
• /
• 2020

Traditional stochastic simulation of hydroclimatological variables often underestimates the variability and correlation structure of larger timescale due to the difficulty in preserving long-term memory. However, the Long Short-Term Memory (LSTM) model illustrates a remarkable long-term memory from the recursive hidden and cell states. The current study, therefore, employed the LSTM model in stochastic generation of hydrologic and climate variables to examine how much the LSTM model can preserve the long-term memory and overcome the drawbacks of conventional time series models such as autoregressive (AR). A trigonometric function and the Rössler system as well as real case studies for hydrological and climatological variables were tested. Results presented that the LSTM model reproduced the variability and correlation structure of the larger timescale as well as the key statistics of the original time domain better than the AR and other traditional models. The hidden and cell states of the LSTM containing the long-memory and oscillation structure following the observations allows better performance compared to the other tested conventional models. This good representation of the long-term variability can be important in water manager since future water resources planning and management is highly related with this long-term variability.

• Journal of Korea Water Resources Association
• /
• v.51 no.6
• /
• pp.503-514
• /
• 2018

In recent, the hydrological regime of the Mekong river is changing drastically due to climate change and haphazard watershed development including dam construction. Information of hydrologic feature like streamflow of the Mekong river are required for water disaster prevention and sustainable water resources development in the river sharing countries. In this study, runoff simulations at the Kratie station of the lower Mekong river are performed using SWAT (Soil and Water Assessment Tool), a physics-based hydrologic model, and LSTM (Long Short-Term Memory), a data-driven deep learning algorithm. The SWAT model was set up based on globally-available database (topography: HydroSHED, landuse: GLCF-MODIS, soil: FAO-Soil map, rainfall: APHRODITE, etc) and then simulated daily discharge from 2003 to 2007. The LSTM was built using deep learning open-source library TensorFlow and the deep-layer neural networks of the LSTM were trained based merely on daily water level data of 10 upper stations of the Kratie during two periods: 2000~2002 and 2008~2014. Then, LSTM simulated daily discharge for 2003~2007 as in SWAT model. The simulation results show that Nash-Sutcliffe Efficiency (NSE) of each model were calculated at 0.9(SWAT) and 0.99(LSTM), respectively. In order to simply simulate hydrological time series of ungauged large watersheds, data-driven model like the LSTM method is more applicable than the physics-based hydrological model having complexity due to various database pressure because it is able to memorize the preceding time series sequences and reflect them to prediction.

• Journal of Wetlands Research
• /
• v.12 no.2
• /
• pp.13-23
• /
• 2010

According to recent researches on climate change, the global warming is obvious to increase rainfall intensity. Damage caused by extreme hydrologic events due to global change is steadily getting bigger and bigger. Recently, frequently occurring heavy rainfalls surely affect the trend of rainfall observations. Probability precipitation estimation method used in designing and planning hydrological resources assumes that rainfall data is stationary. The stationary probability precipitation estimation method could be very weak to abnormal rainfalls occurred by climate change, because stationary probability precipitation estimation method cannot reflect increasing trend of rainfall intensity. This study analyzed temporal variation of trend in rainfall time series at 51 stations which are not significant for statistical trend tests. After modeling rainfall time series with maintaining observed statistical characteristics, this study also estimated whether rainfall data is significant for the statistical trend test in near future. It was found that 13 stations among sample stations will have trend within 10 years. The results indicate that non-stationary probability precipitation estimation method must be applied to sufficiently consider increase trend of rainfall.

• Environmental Engineering Research
• /
• v.15 no.2
• /
• pp.123-126
• /
• 2010

Modeling non-point pollution across multiple scales has become an important environmental issue. As a more representative and practical approach in quantifying and qualifying surface water, a modular neural network (MNN) was implemented in this study. Two different site-scales ($1.5\;{\times}\;10^5$ and $1.62\;{\times}\;10^6\;m^2$) with the same plants, soils, and paddy field management practices, were selected. Hydrologic data (rainfall, irrigation and surface discharge) and water quality data (time-series nutrient loadings) were continuously monitored and then used for the verification of MNN performance. Correlation coefficients (R) for the results predicted from the networks versus measured values were within the range of 0.41 to 0.95. The small block could be extrapolated to the large field for the rainfall-surface drainage process. Nutrient prediction produced less favorable results due to the complex phenomena of nutrients in the drainage water. However, the feasibility of using MNN to generate improved prediction accuracy was demonstrated if more hydrologic and environmental data are provided. The study findings confirmed the estimation accuracy of the upscaling from a small-segment block to large-scale paddy field, thereby contributing to the establishment of water quality management for sustainable agriculture.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.9 no.6
• /
• pp.153-159
• /
• 2009

In order to estimate release water from reservoirs located on ungaged watersheds, an algorithm was suggested based on hydrologic reservoir routing and real time calibrating watershed parameters. A prototype - simple computer program was developed to implement the algorithm with Genetic Algorithm technic. The program was applied to a mid-size reservoir and its ungauged watershed area using observed rainfall data, spillway gates operation data and reservoir water stage time series data under a existing storm event. The result shows that the algorithm and the prototype would be useful to simulate released water from reservoirs.

• Water for future
• /
• v.14 no.4
• /
• pp.27-34
• /
• 1981

The design flow of the urban strom drainage systems has been assessed largely on a basis of empirical relations between rainfall and runoff, and the rational formula has been widely used for the cities in our country. In order to estimate it more accurately, the urban runoff simulation model based on the RRl method has been developed and applied to the sample basin in this study. The rainfall hyetograph of the design stromfor the design flow has been obtained by the determination of the total rainfall and the temporal distributions of that rainfall. The total rainfall has been assessed from the empirical formula of rainfall intensity and the temporal distribution of that rainfall determined on the basis of Huff's method from the historical rainfall data of the basin. The virtual inflow hydrograph to each inlet of the basin has been constructed by computing the series of discharges in each time increment, using design strom hyetograph and time-area diagram. The actual runoff hydrograph at the basin outlet has been computed from the virtual inflow hydrographs by developing a relations between discharge and storage for the watershed. The discharge data for verification of the simulated runoff hydrograph are not available in the sample basin and so the sensitivity analysis of the simulation model has not been possible. The peak discharge for the design of drainage systems has been estimated from the computed runoff hydrograph at the basin outlet and compared to thatl obtained form the rational formula.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.28 no.1B
• /
• pp.1-9
• /
• 2008

The impact of snow accumulation and snowmelt in rainfall-runoff modelling was analyzed for the Soyanggang dam basin by comparing the measured and simulated discharges simulated by the NWS-PC model. Sugawara's conceptual model was used to simulate the snow accumulation and snowmelt phenomena and NWS-PC model was employed to simulate rainfall-runoff. Parameters in model calibration were estimated by the Multi-step Automated Calibration Scheme and optimized using SCE-UA algorithm in each step. The results of the model calibration and verification show that the model considering snowmelt process is better than the one without consideration of snowmelt under the performance criteria such as RMSE, PBIAS, NSE, and PME. The measured discharge time series has over 60 days of persistence. Correlograms for each simulation showed that the simulated discharge with snowmelt model reproduce the persistence closely to the measured discharge's while the one without snow accumulation and snowmelt model reproduce only 20 days of persistence. The study result indicates that the inclusion of snow accumulation and snowmelt model is important for the accurate simulation of rainfall-runoff phenomena in the Soyanggang dam basin.

• Journal of Wetlands Research
• /
• v.10 no.2
• /
• pp.39-51
• /
• 2008

In recent days, the cases of using wetlands in treating waste water, storm events, mining leachate, and agriculture effluents are increasing. But there is the lack of the data for wetlands because of the difficulty in long term monitoring. Such an aspect makes the proper use of wetland impractical. In this study for the purpose of generating a long term hydrologic data, the time series of storage amount for Upo, Mokpo, Sajipo, and Jjokjibeol in Topyeong watershed is simulated using SWAT model. Based on the SWAT-Topyeong model involved in several scenarios for constructing new washlands in Topyeong watershed, the temporal behavior of new washlands is analyzed. It is also revealed that the constructed washland can affect the Upo in some degrees.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.308-308
• /
• 2019

지구상 해양, 대기 및 대륙 상호간의 연속적인 물의 거동을 나타내는 물의 순환의 주요 과정 중 하나인 유량 자료는 경년부터 수십년간의 다양한 기상학적 변동성을 내포하며 해당 지역의 수문기상학적 특성을 반영한다. 이러한 기상학적 변동성 중에서 비교적 긴 시간 주기를 나타내는 저주파 진동은 전지구적 기후변화의 장기적 영향을 나타내며 해수면 상승, 홍수 또는 가뭄과 같은 극한 수문사상을 나타내는 매우 주요한 지표로 활용되고 있지만 관측된 수문 시계열의 짧은 자료길이로 인하여 통계적 분석의 신뢰성에 한계를 보여왔다. 따라서 과거 수문 시계열의 확장으로 인하여 부재의 영역으로 남아있던 자료 기간의 한계가 보완되면 보다 정확하고 신뢰도 있는 분석이 가능할 것이다. 나무나이테를 활용한 고기후 복원 등의 연구가 증가하고 있지만 공학 분야에서 이를 실제로 활용한 연구는 아직 미비하다. 따라서 본 연구에서는 과거 기후의 정보를 바탕으로 복원된 수문 시계열을 활용하여 수문 시계열에 내재된 장기 기후 변동성을 통계적으로 분석하기 위한 문헌들을 조사하고, 장기적인 시간 흐름에 내재된 잠재적인 경향 및 변동성을 통계적 분석을 파악하고자 한다. 이를 위해 주어진 수문 시계열에 내재된 저주파 신호을 추출하기 위한 경험적 모드분해법을 활용하여 수문 자료에 내재된 장기 변동성을 추출하였으며, 산업화 이전부터 연장된 수문 시계열의 공학적 활용성을 분석하고자 한다.