• Journal of Korea Water Resources Association
• /
• v.33 no.1
• /
• pp.87-98
• /
• 2000

The objective of this study is to test the flood forecasting capability of TOPMODEL on a single watershed in Korea. The selected study area is the Soyang River basin with outlet at Soyang Dam site. The three daily hydrographs and the three hourly flood events during 1990~1996 are selected for model calibrations and performance tests. The model parameters are estimated on 1990 daily event by manual fitting technique and the effects of topographic index distribution to river flow simulations are investigated on the study area. The model performance on correlation coefficient between the observed and the simulated flows for the verification periods are above 0.77 on the 95-, 96-daily events, while above 0.87 for 90-, 95-, 96-hourly events. By the consideration of flood flow characteristics in Korea, the physical interpretation of the model concept, and the model performance, it can be concluded that the TOPMODEL is feasible as a flood forecasting model in Korea. Korea.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.12 no.3
• /
• pp.139-151
• /
• 1992

A reservoir operation model was established under the varying restricted water level(r.w.l.) subject to the inflow distributions in flood period. The optimization model consists of 2 sub-models. One model minimizes deviations of releases from the expected release and the other minimizes capacity requirement for flood control. In order to make deterministic equivalents, the inflow distribution of reservoir is assumed to be 2-parameter Lognormal, and its parameters are estimated by the maximum likelihood method. The model is applied to joint operation of Soyang and Chungju dam. The results show that Soyang was designed for larger flood event than that for Chungju. The operation under the varying r.w.l. turns out to be more effective than one under the uniform r.w.l. Such effect is more obvious at Chungju compared with Soyang. Release pattern shows diminishing and delaying effect in a period of high inflows and larger discharges than actual in a period of low inflows.

• Korean Journal of Ecology and Environment
• /
• v.33 no.4 s.92
• /
• pp.328-335
• /
• 2000

Seasonal and vertical variations of organic carbon and chemical oxygen demand (COD$_{Mn}$ were investigated to estimate a suitability of COD$_{Mn}$ methods as an index of the quantity of organic matter in Lake Soyang. Organic matter in the lake may be composed matters resistant to oxidation by COD$_{Mn}$ method, especially in de데 layers. Rations of COD$_{Mn}$ to TOC varied with season and depth layer in Lake Soyang. In the inflowing river, the ratios after summer monsoon were very low and significantly different compared with other seasons (p<0.0001 between after summer monsoon and other seasons by unpaired sample t-test). At the dam site the ratio was low in the deep layer. The ratios in deep layer were significantly different compared with those of upper and middle layers (p<0.0001 by paired sample t-test). In Lake Soyang COD$_{Mn}$ yielded only 40$ of TOC and the oxidation rate of COD$_{Mn}$ varied largely with season and depth. These results implied that the COD$_{Mn}$ does not seem to be an index or organic matter in lakes.

• Korean Journal of Ecology and Environment
• /
• v.37 no.2 s.107
• /
• pp.205-212
• /
• 2004

Phosphorus cycle was studied in a deep stratified reservoir in summer monsoon area (Lake Soyang, Korea) by surveying phosphorus input from the watershed and the movement of phosphorus within the reservoir. And the spatial and temporal distribution of phosphorus was modeled with a 2-dimensional water quality model (CE-QUAL-W2), Phosphorus loading was calculated by measuring TP in the main inflowing river (the Soyang River) accounting for 90% of watershed discharge. TP of the Soyang River showed a large daily variation with the flow rate. High phosphorus loading occurred during a few episodic storm runoff laden with suspended sediments and phosphorus. Because storm runoff water on rainy days have lower temperature, it plunges into a depth of same temperature (usually below 20m depth), forming an intermediate turbidity layer with a thickness of 20 ${\sim}$ 30 m. Because of stable thermal stratification in summer the intermediate layer water of high phosphorus content was discharged from the dam through a mid-depth outlet without diffusing into epilimnion. The movement of runoff water within the reservoir, and the subsequent distribution of phosphorus were well simulated by the water quality model showing a good accuracy. The major parameter for the calibration of phosphorus cycle was a settling velocity of detritus, which was calibrated to be 0.75 m ${\cdot}$ $day^{-1}$. It is concluded that the model can be a good simulator of limnological phenomena in reservoirs of summer monsoon area.

• Journal of Environmental Science International
• /
• v.23 no.5
• /
• pp.839-852
• /
• 2014

The Ministry of Environment (MOE) has made more effort in managing point source pollution rather than in nonpoint source pollution in order to improve water quality of the four major rivers. However, it would be difficult to meet water quality targets solely by managing the point source pollution. As a result of the comprehensive measures established in 2004 under the leadership of the Prime Minister's Office, a variety of policies such as the designation of control areas to manage nonpoint source pollution are now in place. Various action plans to manage nonpoint source pollution have been implemented in the Soyang-dam watershed as one of the control areas designed in 2007. However, there are no tools to comprehensively assess the effectiveness of the action plans. Therefore, this study would assess the action plans (especially, BMPs) designed to manage Soyang-dam watershed with the WinHSPF and the CE-QUAL-W2. To this end, we simulated the rainfall-runoff and the water quality (SS) of the watershed and the reservoir after conducting model calibration and the model validation. As the results of the calibration for the WinHSPF, the determination coefficient ($R^2$) for the flow (Q, $m^3/s$) was 0.87 and the $R^2$ for the SS was 0.78. As the results of the validation, the former was 0.78 and the latter was 0.67. The results seem to be acceptable. Similarly, the calibration results of the CE-QUAL-W2 showed that the RMSE for the water level was 1.08 and the RMSE for the SS was 1.11. The validation results(RMSE) of the water level was 1.86 and the SS was 1.86. Based on the daily simulation results, the water quality target (turbidity 50 NTU) was not exceeded for 2009~2011, as results of maximum turbidity in '09, '10, and '11 were 3.1, 2.5, 5.6 NTU, respectively. The maximum turbidity in the years with the maximum, the minimum, and the average of yearly precipitation (1982~2011) were 15.5, 7.8, and 9.0, respectively, and therefore the water quality target was satisfied. It was discharged high turbidity at Inbuk, Gaa, Naerin, Gwidun, Woogak, Jeongja watershed resulting of the maximum turbidity by sub-basins in 3years(2009~2011). The results indicated that the water quality target for the nonpoint source pollution management should be changed and management area should be adjusted and reduced.

• Korean Journal of Ecology and Environment
• /
• v.35 no.1 s.97
• /
• pp.36-44
• /
• 2002

The spatial and temporal trends of water qualities in Lake Soyang was statistically analyzed in this study. The water qualities include nutrients, ionic contents and chlorophyll-a (Chl-a) measured during 1993${\sim}$2000. The rainfall intensity and runoff from the catchment appeared to play an important role in water quality trends in the lake. According to seasonal Mann-Kendall test, conductivity, TP, and Ctl-a did not show any trends of increase or decrease over the 8 year period, while TN declined slightly. It was found that the variation of TP was a function of interannual inflow and rainfall. In the analyses of spatial trend, conductivity, based on the mean by site, showed a downlake decline over the eight year period. Minimum conductivity was found in the headwaters during summer monsoon of July to August and near the dam during October. This result indicates a time-lag phenomenon that the headwater is diluted by rainwater immediately after summer monsoon rain and then the lake water near the dam is completely diluted in October. During summer period, TP and TN had an inverse relation with conductivity values. Concentrations of TP peaked during July to September in the headwaters and during September in the downlake. Also, TN increase during the summer and was more than 1.5 mg/L regardless of season and location, indicating a consistent eutrophic state. Values of Chl-a varied depending on location and season, but peaked in the midlake rather than in the headwaters during the monsoon. Regression analyses of log-transformed seasonal Chl-a against TP showed that value of $R^2$ was below 0.003 in the premonsoon and monsoon seasons but was 0.82 during the postmonsoon, indicating a greater algal response to the phosphorus during the postmonsoon. In contrast, TN had no any relations with Chl-a during all seasons.

• Korean Journal of Ecology and Environment
• /
• v.38 no.3 s.113
• /
• pp.297-303
• /
• 2005

The analysis of fish community structure in the Peace Dam, Korea, was investigated from April 2003 to September 2004. Eight families and 31 species were collected during the period surveyed. Korean endemic species were 13 and the relative abundance was 41.9% of the total, along with Hemibarbus mylodon, Acheilognathus yamatsutae and Coreoleuciscus splendidus. Dominant was Hemibarbus labeo (14.1%) and subdominant species was Zacco temmincki (13.7%). Also, Z. platypus (11.9%), H. longirostris (9.3%), and Siniperca scherzeri (8.1%), were numerous. The biomass of collected fishes were H. labeo (24,278 g), S. scherzeri (19,487 g), Anguilla japonica (10,400 g), Cyprinus carpio (8,418 g), and Carassius auratus (4,020 g), According to the community analyses, the artificial reservoir with in the Bukhan River system were divided into 2 groups by unweighted pair-group method analysis (UPGMA), and it revealed that the community structure in Peace Dam was similar to those in Soyang and Paro Reservoirs.

• Journal of Korea Water Resources Association
• /
• v.37 no.2
• /
• pp.121-131
• /
• 2004

SIMHYD and TANK model are used to predict time series of daily rainfall-runoff of Soyang Dam and Youngcheon Dam watershed. The performances of SIMHYD model with 7 parameters and TANK model with17 parameters are compared. Three optimization methods (Genetic algorithm, Pattern search multi-start and Shuffled Complex Evolution algorithm) were applied to study-areas with 3 different types of objective functions. Efficiency of TANK model is higher than that of SIMHYD. Among different types of objective function, Nash-sutcliffe coefficient is found to be the most appropriateobjective function to evaluate applicability of model.

• 한국방재학회:학술대회논문집
• /
• 2008.02a
• /
• pp.723-726
• /
• 2008

It is not always easy to estimate the parameters in hydrologic models due to insufficient hydrologic data when hydraulic structures are designed or water resources plan are established, uncertainty analysis, therefore, are inevitably needed to examine reliability for the estimated results. With regard to this point, this study applies a Bayesian Markov Chain Monte Carlo scheme to the NWS-PC rainfall-runoff model that has been widely used, and a case study is performed in Soyang Dam watershed in Korea. The NWS-PC model is calibrated against observed daily runoff, and thirteen parameters in the model are optimized as well as posterior distributions associated with each parameter are derived. The Bayesian Markov Chain Monte Carlo shows a improved result in terms of statistical performance measures and graphical examination. The patterns of runoff can be influenced by various factors and the Bayesian approaches are capable of translating the uncertainties into parameter uncertainties. One could provide against an expected runoff event by utilizing information driven by Bayesian methods. Therefore, the rainfall-runoff analysis coupled with the uncertainty analysis can give us an insight in evaluating flood risk and dam size in a reasonable way.

• Journal of Korea Water Resources Association
• /
• v.55 no.7
• /
• pp.495-504
• /
• 2022

With the recent development of artificial intelligence, a Long Short-Term Memory (LSTM) model that is efficient with time-series analysis is being used to increase the accuracy of predicting the inflow of dams. In this study, we predict the inflow of the Soyang River dam, using the LSTM model with the Sequence-to-Sequence (LSTM-s2s) and attention mechanism (LSTM-s2s with attention) that can further improve the LSTM performance. Hourly inflow, temperature, and precipitation data from 2013 to 2020 were used to train the model, and validate and test for evaluating the performance of the models. As a result, the LSTM-s2s with attention showed better performance than the LSTM-s2s in general as well as in predicting a peak value. Both models captured the inflow pattern during the peaks but detailed hourly variability is limitedly simulated. We conclude that the proposed LSTM-s2s with attention can improve inflow forecasting despite its limits in hourly prediction.