• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.165-165
• /
• 2023

Streamflow prediction is a critical task in water resources management and essential for planning and decision-making purposes. However, the streamflow prediction is challenging due to the complexity and non-linear nature of hydrological processes. The transfer learning is a powerful technique that enables a model to transfer knowledge from a source domain to a target domain, improving model performance with limited data in the target domain. In this study, we apply the transfer learning using the Informer model, which is a state-of-the-art deep learning model for streamflow prediction. The model was trained on a large-scale hydrological dataset in the source basin and then fine-tuned using a smaller dataset available in the target basin to predict the streamflow in the target basin. The results demonstrate that transfer learning using the Informer model significantly outperforms the traditional machine learning models and even other deep learning models for streamflow prediction, especially when the target domain has limited data. Moreover, the results indicate the effectiveness of streamflow prediction when knowledge transfer is used to improve the generalizability of hydrologic models in data-sparse regions.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2009.05a
• /
• pp.2176-2178
• /
• 2009

• Journal of Environmental Science International
• /
• v.18 no.6
• /
• pp.609-619
• /
• 2009

SWAT-K model is a modified version of the original SWAT, and is known to more accurately estimate the streamflows and pollutant loadings in Korean watersheds. In this study, its hydrological components were compared with those of HSPF in order to analyse the differences in total runoff including evapotranspiration(ET), surface flow, lateral flow and groundwater flow from the Chungju Dam watershed during $2000{\sim}2006$. Averaged annual runoff with SWAT-K overestimated by 1%, and HSPF underestimated it by 3% than observed runoff. Determination coefficients($R^2$) for observed and simulated daily streamflows by both the models were relatively good(0.80 by SWAT-K and 0.82 by HSPF). Potential ET and actual ET by HSPF were lower in winter, but similar or higher than those by SWAT-K. And though there were some differences in lateral and groundwater flows by two models because of the differences in hydrological algorithms, the results were to be reasonable. From the results, it was suggested that we should utilize a proper model considering the characteristic of study area and purposes of the model application because the simulated results from same input data could be different with models used. Also we should develop a novel model appropriate to Korean watersheds by enhancing limitations of the existing models in the future.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.42 no.4
• /
• pp.67-75
• /
• 2000

A simulation model for reservoir sizing was developed to be applied in a region with insufficient hydrological data. Reservoir storage balance equation was formulated on a monthly basis. Gajiyama equation was generalized to estimate monthly reservoir inflow more accurately. Monthly evaporation equation on a reservoir water surface was introduced , which was functioned with monthly mean temperature. Generalized Gajiyama equation was applied to estmate reservoir inflow of the Sayeon dam. Nash-Sutcliffe's model efficiency was 0.793. Using developed model for reservoir sizing, water supply capacity was analyzed with 118.000㎥/day on the Sayeon dam. This showed a reasonable result as compared with 110000㎥/day in other technical report. For general application of developed model, a virtual reservoir was considered and its dta of surface area and volume by elevation was prepared using DEM. Using the model, size of reservoir was determined and water supply capacity was anlayzed on a virtual reservoir.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.44 no.7
• /
• pp.25-35
• /
• 2002

In this study, the effect of DEM (Digital Elevation Model) resolution (15m, 30m, 50m, 70m, 100m, 200m, 300m) on the hydrological simulation was examined using the BASINS (Better Assessment Science Integrating point and Nonpoint Source) for the Heukcheon watershed (303.3 ㎢) data from 1998 to 1999. Generally, as the cell size of DEM increased, topographical changes were observed as the original range of elevation decreased. The processing time of watershed delineation and river network needed more time and effort on smaller cell size of DEM. The larger DEM demonstrated had some errors in the junction of river network which might affect on the simulation of water quantity and quality. The area weighted average watershed slope became milder but the length weighted average channel slope became steeper as the DEM size increased. DEM resolution affected substantially on the topographical parameter but less on the hydrological simulation. Considering processing time and accuracy on hydrological simulation, DEM grid size of 100m is recommended for this range of watershed size.

• Advances in environmental research
• /
• v.7 no.3
• /
• pp.161-175
• /
• 2018

Krishna River is significantly affected due to Srisailam dam from past 30 years. The impact of this hydraulic structure drastically reduced the minimum flow regime on the downstream, which made the river nearing to decaying stage. In the present paper, Environmental Flow called minimum flow values released for the dam are estimated with the help of three hydrological methods viz., Range of variability Approach (RVA), Desktop Reserve Model (DRM), and Global Environmental Flow Calculator (GEFC). DRM method suggested considering the intermediate values obtained from among the three methods to preserve the ecosystem on the downstream of the river, which amounts to an average annual allocation of 9378 Million Cubic Meter (MCM) which is equal to 23.11% of mean annual flow (MAF). In this regard GEFC and RVA methods accounted for 22% and 31.04% of MAF respectively. The results indicate that current reservoir operation policy is causing a severe hydrological alteration in the high flow season especially in the month of July. The study concluded that in the case of non-availability of environmental information, hydrological indicators can be used to provide the basic assessment of environmental flow requirements. It is inferred from the results obtained from the study, that the new reservoir operations can fulfil human water needs without disturbing Environmental Flow Requirements.

• Journal of the Korean Society of Safety
• /
• v.27 no.6
• /
• pp.166-171
• /
• 2012

This study aims to analyze the hydrological characteristics of downstream areas by the dam discharge of Soyanggang dam by using HEC-RAS Model. As a result of analyzing the data of dam discharge divided into hydropeaking discharge and total discharge, it as found that the maximum hydro-peaking discharge and the maximum total discharge have been 254.4 CMS and 1567.7 CMS respectively for the past 11 years. When the hydro-peaking discharge was applied to HEC-RAS Model, there occurred some sections where the water level rapidly changed, but the velocity of moving water was quite stable in the range between 0.23 m/sec and 1.16m/sec. Besides, when the total discharge was applied to this model, the submersible bridge along the dam downstream was flooded, and in some sections, the water level increased over the flood plain. Accordingly, this study judged that it is required to necessarily consider all the influence made by an increase of Soyanggang Dam's discharge when waterfronts are installed or used at dam downstream areas.

• Environmental Engineering Research
• /
• v.17 no.4
• /
• pp.191-196
• /
• 2012

Analysis of hydrological safety as well as the determination of many substance concentrations are necessary when bioreactor systems are introduced to landfill operations. Therefore, hydrological and substance balance model was developed since it can be applied to various bioreactor landfill operation systems. For the final evaluation of the model's effectiveness, four different methods of injections (leachate alone, leachate and organic waste water, leachate and reverse osmosis concentrate, and all the above three combination) was applied to 1st landfill site of Sudokwon landfill. As a result, the water content of the hypothetical cases for four different systematic bioreactors is projected to be increased up to 35.5% in next 10 years, and this indicated that there will be no problems in meeting the hydrological safety. Also, the final $Cl^-$ concentration after 10-yr time period was projected to be between from minimum 126 to maximum 3,238 mg/L, which could be still a decrease from the original value of 3,278 mg/L. According to the proposed model, whether the substance concentration becomes increased or decreased largely depends on the ratio of initial quantity of inner landfill leachate and the rate of injection.

• Journal of Korea Water Resources Association
• /
• v.40 no.5
• /
• pp.419-430
• /
• 2007

Integrated modelling of surface water and groundwater has become important to satisfy the growing demands for sustainable water resources and improved water quality. In this study, the integrated model of the semi-distributed watershed model, SWA T and the fully-distributed groundwater flow model, MODFLOW is applied to Musirn river basin for the purpose of investigating its applicability to reproduce watershed-scale hydrological processes. This objective is accomplished by first demonstrating good agreement between the simulated discharge hydrographs with the measured hydrographs for the period of 2001 -2004 while simultaneously calibrating the calculated groundwater level distribution to observation wells. Next, the integrated model is used to evaluate the effect of different temporal precipitation averages on hydrodynamic processes of streamflow, percolation, recharge and groundwater discharge. Moreover, comprehensive simulations are performed to present the relationships between monthly precipitation and each hydrological component, and to analyze the temporal-spatial variability of recharge. The results show that the components are highly interrelated, and that the heterogeneity of watershed characteristics such as subbasin slope, land use, soil type causes a significant spatial variation of recharge. Overall it is concluded that the model is capable of reproducing the temporally and spatially varied surface and subsurface hydrological processes at the watershed scale.

• Journal of Korea Water Resources Association
• /
• v.50 no.11
• /
• pp.769-779
• /
• 2017

As the occurrence of drought is recently on the rise, the reliable drought forecasting is required for developing the drought mitigation and proactive management of water resources. This study developed a probabilistic hydrological drought forecasting method using the Bayesian Networks and drought propagation relationship to estimate future drought with the forecast uncertainty, named as the Propagated Bayesian Networks Drought Forecasting (PBNDF) model. The proposed PBNDF model was composed with 4 nodes of past, current, multi-model ensemble (MME) forecasted information and the drought propagation relationship. Using Palmer Hydrological Drought Index (PHDI), the PBNDF model was applied to forecast the hydrological drought condition at 10 gauging stations in Nakdong River basin. The receiver operating characteristics (ROC) curve analysis was applied to measure the forecast skill of the forecast mean values. The root mean squared error (RMSE) and skill score (SS) were employed to compare the forecast performance with previously developed forecast models (persistence forecast, Bayesian network drought forecast). We found that the forecast skill of PBNDF model showed better performance with low RMSE and high SS of 0.1~0.15. The overall results mean the PBNDF model had good potential in probabilistic drought forecasting.