• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.35-35
• /
• 2022

최근에 딥 러닝(Deep learning) 기반의 많은 방법들이 수문학적 모형 및 예측에서 의미있는 결과를 보여주고 있지만 더 많은 연구가 요구되고 있다. 본 연구에서는 수자원의 가장 대표적인 모델링 구조인 강우유출의 관계의 규명에 대한 모형을 Long Short-Term Memory (LSTM) 기반의 변형 된 방법으로 제시하고자 한다. 구체적으로 본 연구에서는 반응변수인 유출량에 대한 직접적인 고려가 아니라 그의 1차 도함수 (First derivative)로 정의되는 Delta기반으로 모형을 구축하였다. 또한, Attention 메카니즘 기반의 모형을 사용함으로써 강우유출의 관계의 규명에 있어 정확성을 향상시키고자 하였다. 마지막으로 확률 기반의 예측를 생성하고 이에 대한 불확실성의 고려를 위하여 Denisty 기반의 모형을 포함시켰고 이를 통하여 Epistemic uncertainty와 Aleatory uncertainty에 대한 상대적 정량화를 수행하였다. 본 연구에서 제시되는 모형의 효용성 및 적용성을 평가하기 위하여 미국 전역에 위치하는 총 507개의 유역의 일별 데이터를 기반으로 모형을 평가하였다. 결과적으로 본 연구에서 제시한 모형이 기존의 대표적인 딥 러닝 기반의 모형인 LSTM 모형과 비교하였을 때 높은 정확성뿐만 아니라 불확실성의 표현과 정량화에 대한 유용한 것으로 확인되었다.

• Journal of Korean Society on Water Environment
• /
• v.23 no.6
• /
• pp.897-905
• /
• 2007

Approximately 70% of Korea is composed of forest areas. Especially 48% of agricultural field is practiced at highland areas over 400 m in elevation in Kangwon province. Over 90% of highland agricultural farming is located at Kangwon province. Runoff characteristics at the mountainous area such as Kangwon province are largely affected by steep slopes, thus runoff estimation considering field slopes needs to be utilized for accurate estimation of direct runoff. Although many methods for runoff estimation are available, the Soil Conservation Service (SCS), now Natural Resource Conservation Service (NRCS), Curve Number (CN)-based method is used in this study. The CN values were obtained from many plot-years dataset obtained from mid-west areas of the United States, where most of the areas have less than 5% in slopes. Thus, the CN method is not suitable for accurate runoff estimation where significant areas are over 5% in slopes. Therefore, the CN values were adjusted based on the average slopes (25.8% at Doam-dam watershed) depending on the 5-day Antecedent Moisture Condition (AMC). In this study, the CN-based Long-Term Hydrologic Impact Assessment (L-THIA) direct runoff estimation model used and the Web-based Hydrograph Analysis Tool (WHAT) was used for direct runoff separation from the stream flow data. The $R^2$ value was 0.65 and the Nash-Sutcliffe coefficient value was 0.60 when no slope adjustment was made in CN method. However, the $R^2$ value was 0.69 and the Nash-Sutcliffe value was 0.69 with slope adjustment. As shown in this study, it is strongly recommended the slope adjustment in the CN direct runoff estimation should be made for accurate direct runoff prediction using the CN-based L-THIA model when applied to steep mountainous areas.

• Journal of Korean Society on Water Environment
• /
• v.24 no.2
• /
• pp.221-229
• /
• 2008

This study revised a model for hydrologically analyzing rainwater harvesting facilities considering their rainfall-runoff properties and the data available. This model has only a few parameters, which can be estimated with rather poor measurements available. The model has a non-linear module for rainfall loss, and the remaining rainfall excess (effective rainfall) is assumed to be inflow to the storage tank. This model has been applied for the rainwater harvesting facilities in Seoul National University, Korea Institute of Construction Technology, and the Daejon World Cup Stadium. As a result, the runoff coefficients estimated were about 0.9 for the building roof as a rainwater collecting surface and about 0.18 for the playground. This result is coincident with that for designing the rainwater harvesting facilities to show the accuracy of model and the simulation results.

• Environmental Engineering Research
• /
• v.17 no.1
• /
• pp.47-56
• /
• 2012

A long-term nationwide nonpoint-source pollution monitoring program was initiated by the Ministry of Environment Republic of Korea (ME) in 2007. Monitoring devices including rain gauges, flow meters, and automatic samplers were installed in monitoring sites to collect dynamic runoff data in 2008-2009. More than 10 rainfall events with three or more antecedent dry days were monitored per year. More than 10 samples were collected and analyzed per event. So far, five land use types (single family, apartments, education facilities, power plants, and other public facilities) have been monitored 23 to 24 times each. Characterization of the runoff from different land use types will aid unit load estimation in Korea and hopefully in other countries with similar land use. The monitoring results will be reported regularly at national and international levels.

• Journal of Korean Society on Water Environment
• /
• v.23 no.6
• /
• pp.829-836
• /
• 2007

For the purpose of managing non-point sources, water quality control basins (WQCBs) are often designed to capture rainfall events smaller than extreme events. The design rainfall statistics and runoff capture rates for sizing a WQCB should be derived from the local long-term continuous rainfall record. In this study, the 31-year continuous rainfall data recorded in Busan is analyzed to derive the synthesized runoff capture curve incorporated with SCS curve number.

• Water for future
• /
• v.28 no.1
• /
• pp.81-90
• /
• 1995

This study attempts to develop a stochastic system model for extension and prediction of monthly runoff series in river basins where the observed runoff data are insufficient although there are long-term hydrometeorological records. For this purpose, univariate models of a seasonal ARIMA type are derived from the time series analysis of monthly runoff, monthly precipitation and monthly evaporation data with trend and periodicity. Also, a causual model of multiple input-single output relationship that take monthly precipitation and monthly evaporation as input variables-monthly runoff as output variable is built by the cross-correlation analysis of each series. The performance of the univariate model and the multiple input-output model were examined through comparisons between the historical and the generated monthly runoff series. The results reveals that the multiple input-output model leads to the improved accuracy and wide range of applicability when extension and prediction of monthly runoff series is required.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.51 no.5
• /
• pp.107-113
• /
• 2009

The purpose of this study is to estimate the design flood runoff for ungaged forest watershed to reduce the flood damage in national park. Daewonsa watershed in Jirisan National Park was selected as study watershed, of which characteristic factors were obtained from GIS data. Flood runoff was simulated using SCS unit hydrograph module in HEC-HMS model. SCS Curve Number (CN) was calculated from forest type area weighted average method. Huff's time distribution of second-quartile storm of the Sancheong weather station, which is nearest from study watershed, was used for design flood runoff estimation. Critical storm duration for the study watershed was 3 hrs. Based on the critical duration, the peak runoff for each sub-watershed were simulated. It is recommended to monitor the long-term flow data for major stream stations in National Park for a better reliable peak runoff simulation results.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.53 no.6
• /
• pp.171-183
• /
• 2011

Temporal and spatical surface runoff by heavy rainfall during 25~28 July, 2011 causing urban flooding at Seoul were analyzed using Long-Term Hydrologic Impact Assessment (L-THIA). L-THIA was calibrated for 1988~1997 and validated for 1998~2007 using monthly observed data at Hangangseoul watershed which covers 90 % of Seoul city. As a results of calibration and validation of L-THIA at Hangangseoul watershed, Nash-Sutcliffe coefficients were 0.99 for calibration and 0.99 for validation. The simulated values were good agreement with observed data and both calibrated and validated levels were "very good" based on calibration criteria. The calibrated curve number (CN) values of residential and other urban area represented 87 % and 93 % of impervious area, respectively, which were maximum percentage of impervious area. As a result of L-THIA application at Seoul city during 25~28 July, 2011, most of rainfall (54 %, 287.49 mm) and surface runoff (65 %, 247.32) were generated at 27 July, 2011 and a significant amount of rainfall and surface runoff were occurred at southeastern Seoul city. As a result of bi-hourly spatial and temporal analysis during 27 July, 2011, surface runoff during 2:00~4:00 and 8:00~10:00 were much higher than those during other times and surface runoff located at Seocho-gu during 6:00~8:00 represented maximum value with maximum rainfall intensity which caused landslide from Umyun mountain.

• Water for future
• /
• v.24 no.3
• /
• pp.95-104
• /
• 1991

The study was focused on developing a new model to estimate annual runoff. This model can be used to estimate the available water resources for ungaged catchments for long-term water resources development planning. Data used in the model development were daily rainfall and daily runoff of the sample basin with record length from 1945 to 1988 years in Korea. The sample basin selected by consideration whether the flow is virgin and quality of discharge data is good. As a result, 46 stage gaging station were selected. Annual runoff was determined by sum of daily runoff calculated by daily stage data of the sample basin. Also, the annual mean precipitation by using daily rainfall data was estimated and the annual runoff ratio for each sample basin was calculated, and the annual mean runoff ratio was estimated. The linear regression model was proposed and calibrated using auunal mean precipitation values and geomorphological characteristics of the basins. To verify reasonableness of this model, the regression model was applied to the gaging stations which have historical data.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2005.05b
• /
• pp.1199-1203
• /
• 2005

Combined sewer overflows(CSOs) are themselves a significant source of water pollution. Therefore, the control of urban drainage for CSOs reduction and receiving water quality protection is needed. Examples in combined sewer systems include downstream storage facilities that detain runoff during periods of high flow and allow the detained water to be conveyed by an interceptor sewer to a centralized treatment plant during periods of low flow. The design of such facilities as stormwater detention storage is highly dependant on the temporal variability of storage capacity available(which is influenced by the duration of interevent dry periods) as well as the infiltration capacity of soil and recovery of depression storage. As a result, a contiunous approach is required to adequately size such facilities. This study for the continuous long-term analysis of urban dranage system used analytical Probabilistic model based on derived probability distribution theory. As an alternative to the modeling of urban drainage system for planning or screening level analysis of runoff control alternatives, this model have evolved that offer much ease and flexibility in terms of computation while considering long-term meteorology. This study presented rainfall and runoff characteristics or the subject area using analytical Probabilistic model. Runoff characteristics manifasted the unique characteristics of the subject area with the infiltration capacity of soil and recovery of depression storage and was examined appropriately by sensitivity analysis. This study presented the average annual COSs and number of COSs when the interceptor capacity is in the range 3xDWF(dry weather flow). Also, calculated the average annual mass of pollutant lost in CSOs using Event Mean Concentration. Finally, this study presented a dicision of storage volume for CSOs reduction and water quality protection.