• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.170-170
• /
• 2020

Due to the development of technologies, complex computation of huge data set is possible with a prevalent personal computer. Therefore, machine learning methods have been widely applied in the hydrologic field such as regression-based regional frequency analysis (RFA). The main purpose of this study is to compare two frameworks of RFA based on the artificial neural network (ANN) models: quantile regression technique (QRT-ANN) and parameter regression technique (PRT-ANN). As an output layer of the ANN model, the QRT-ANN predicts quantiles for various return periods whereas the PRT-ANN provides prediction of three parameters for the generalized extreme value distribution. Rainfall gauging sites where record length is more than 20 years were selected and their annual maximum rainfalls and various hydro-meteorological variables were used as an input layer of the ANN model. While employing the ANN model, 70% and 30% of gauging sites were used as training set and testing set, respectively. For each technique, ANN model structure such as number of hidden layers and nodes was determined by a leave-one-out validation with calculating root mean square error (RMSE). To assess the performances of two frameworks, RMSEs of quantile predicted by the QRT-ANN are compared to those of the PRT-ANN.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.57 no.5
• /
• pp.139-152
• /
• 2015

Along with climate change, it is reported that the scale and frequency of extreme climate events show unstable tendency of increase. Thus, to comprehend the change characteristics of precipitation data, it is needed to consider non-stationary. The main objectives of this study were to estimate future design floods for Wonpyeongcheon watershed based on RCP (Representative Concentration Pathways) scenario. Wonpyeongcheon located in the Keum River watershed was selected as the study area. Historical precipitation data of the past 35 years (1976~2010) were collected from the Jeonju meteorological station. Future precipitation data based on RCP4.5 were also obtained for the period of 2011~2100. Systematic bias between observed and simulated data were corrected using the quantile mapping (QM) method. The parameters for the bias-correction were estimated by non-parametric method. A non-stationary frequency analysis was conducted with moving average method which derives change characteristics of generalized extreme value (GEV) distribution parameters. Design floods for different durations and frequencies were estimated using rational formula. As the result, the GEV parameters (location and scale) showed an upward tendency indicating the increase of quantity and fluctuation of an extreme precipitation in the future. The probable rainfall and design flood based on non-stationarity showed higher values than those of stationarity assumption by 1.2%~54.9% and 3.6%~54.9%, respectively, thus empathizing the necessity of non-stationary frequency analysis. The study findings are expected to be used as a basis to analyze the impacts of climate change and to reconsider the future design criteria of Wonpyeongcheon watershed.

• Journal of Korea Water Resources Association
• /
• v.50 no.10
• /
• pp.661-671
• /
• 2017

The objective of this study was to analyze the flood stage considering the uncertainty caused by the river roughness coefficients and discharge. The methodology of this study involved the GLUE (Generalized Likelihood Uncertainty Estimation) to quantify the uncertainty bounds applying three different storm events. The uncertainty range of the roughness was 0.025~0.040. In case of discharge, the uncertainty stemmed from parameters in stage-discharge rating curve, if h represents stage for discharge Q, which can be written as $Q=A(h-B)^C$. Parameters in rating curve (A, B and C) were estimated by non-linear regression model and assumed by t distribution. The range of parameters in rating curve was 5.138~18.442 for A, -0.524~0.104 for B and 2.427~2.924 for C. By sampling 10,000 parameter sets, Monte Carlo simulations were performed. The simulated stage value was represented by 95% confidence interval. In storm event 1~3, the average bound was 0.39 m, 0.83 m and 0.96 m, respectively. The peak bound was 0.52 m, 1.36 m and 1.75 m, respectively. The recurrence year of each storm event applying the frequency analysis was 1-year, 10-year and 25-year, respectively.

• Journal of Korea Water Resources Association
• /
• v.49 no.8
• /
• pp.645-656
• /
• 2016

The study performed a regional flood frequency analysis and proposed a regression equation to estimate design floods corresponding to return periods for ungauged basins in Geum-river basin. Five preliminary tests were employed to investigate hydrological independence and homogeneity of streamflow data, i.e. the lag-one autocorrelation test, time homogeneity test, Grubbs-Beck outlier test, discordancy measure test ($D_i$), and regional homogeneity measure (H). The test results showed that streamflow data were time-independent, discordant and homogeneous within the basin. Using five probability distributions (generalized extreme value (GEV), three-parameter log-normal (LN-III), Pearson type 3 (P-III), generalized logistic (GLO), generalized Pareto (GPA)), comparative regional flood frequency analyses were carried out for the region. Based on the L-moment ratio diagram, average weighted distance (AWD) and goodness-of-fit statistics ($Z^{DIST}$), the GLO distribution was selected as the best fit model for Geum-river basin. Using the GLO, a regression equation was developed for estimating regional design floods, and validated by comparing the estimated and observed streamflows at the Ganggyeong station.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.40 no.4
• /
• pp.45-57
• /
• 1998

This study was conducted to derive optimal design floods by Generalized Extreme Value (GEV) distribution for the annual maximum series at ten watersheds along Han, Nagdong, Geum, Yeongsan and Seomjin river systems. Adequacy for the analysis of flood data used in this study was established by the tests of Independence, Homogeneity, detection of Outliers. L-coefficient of variation, L-skewness and L-kurtosis were calculated by L-moment ratio respectively. Parameters were estimated by the Methods of Moments and L-Moments. Design floods obtained by Methods of Moments and L-Moments using different methods for plotting positions in GEV distribution were compared by the Relative Mean Errors(RME) and Relative Absolute Errors(RAE). The results were analyzed and summarized as follows. 1. Adequacy for the analysis of flood data was acknowledged by the tests of Independence, Homogeneity and detection of Outliers. 2. GEV distribution used in this study was found to be more suitable one than Pearson type 3 distribution by the goodness of fit test using Kolmogorov-Smirnov test and L-Moment ratios diagram in the applied watersheds. 3. Parameters for GEV distribution were estimated using Methods of Moments and L-Moments. 4. Design floods were calculated by Methods of Moments and L-Moments in GEV distribution. 5. It was found that design floods derived by the method of L-Moments using Weibull plotting position formula in GEV distribution are much closer to those of the observed data in comparison with those obtained by method of moments using different formulas for plotting positions from the viewpoint of Relative Mean Errors and Relative Absolute Errors.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.44 no.5
• /
• pp.41-53
• /
• 2002

This study was conducted to derive the regional design rainfall by the regional frequency analysis based on the regionalization of the precipitation suggested by the first report of this project. According to the regions and consecutive durations, optimal design rainfalls were derived by the regional frequency analysis for L-moment in the second report of this project. Using the LH-moment ratios and Kolmogorov-Smirnov test, the optimal regional probability distribution was identified to be the Generalized extreme value (GEV) distribution among applied distributions. regional and at-site parameters of the GEV distribution were estimated by the linear combination of the higher probability weighted moments, LH-moment. Design rainfall using LH-moments following the consecutive duration were derived by the regional and at-site analysis using the observed and simulated data resulted from Monte Carlo techniques. Relative root-mean-square error (RRMSE), relative bias (RBIAS) and relative reduction (RR) in RRMSE for the design rainfall were computed and compared in the regional and at-site frequency analysis. Consequently, it was shown that the regional analysis can substantially more reduce the RRMSE, RBIAS and RR in RRMSE than at-site analysis in the prediction of design rainfall. Relative efficiency (RE) for an optimal order of L-moments was also computed by the methods of L, L1, L2, L3 and L4-moments for GEV distribution. It was found that the method of L-moments is more effective than the others for getting optimal design rainfall according to the regions and consecutive durations in the regional frequency analysis. Diagrams for the design rainfall derived by the regional frequency analysis using L-moments were drawn according to the regions and consecutive durations by GIS techniques.

• Journal of Astronomy and Space Sciences
• /
• v.16 no.2
• /
• pp.241-254
• /
• 1999

Differential time-series observations of BL Camelopardalis classified as a double mode SX Phoenicis type variable were secured with a charge coupled device. The observed photometric data was reduced using the IRAF Package and the differential magnitudes were obtained through aperture photometry. The periods of BL Cam were analyzed with the Generalized Least-Square Method by Vanicek (1971) and the Fourier Decomposition Method. It was found that the first and second period of BL Cam were 0.0391 day respectively which lead the period ratio of P1/P0=0.81. This period ratio is much different from 0.78 determined by other investigators and also much more larger than that of other double-mode SX Phe type variables. In addition, this period ratio is much different from the value expected from the relation between the metallicity and period ratio. From these results, it can be confirmed that BL Cam is the most extreme case among all double-mode SX Phe type variables.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.45 no.1
• /
• pp.33-44
• /
• 2003

This study was conducted to estimate the design flood by the determination of best fitting order for LH-moments of the annual maximum series at fifteen watersheds. Using the LH-moment ratios and Kolmogorov-Smirnov test, the optimal regional probability distribution was identified to be the Generalized Extreme Value (GEV) in the first report of this project. Parameters of GEV distribution and flood flows of return period n years were derived by the methods of L, L1, L2, L3 and L4-moments. Frequency analysis of flood flow data generated by Monte Carlo simulation was performed by the methods of L, L1, L2, L3 and L4-moments using GEV distribution. Relative Root Mean Square Error. (RRMSE), Relative Bias (RBIAS) and Relative Efficiency (RE.) using methods of L, Ll , L2, L3 and L4-moments for GEV distribution were computed and compared with those resulting from Monte Carlo simulation. At almost all of the watersheds, the more the order of LH-moments and the return periods increased, the more RE became, while the less RRMSE and RBIAS became. The Absolute Relative Reduction (ARR) for the design flood was computed. The more the order of LH-moments increased, the less ARR of all applied watershed became It was confirmed that confidence efficiency of estimated design flood was increased as the order of LH-moments increased. Consequently, design floods for the appled watersheds were derived by the methods of L3 and L4-moments among LH-moments in view of high confidence efficiency.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.102-102
• /
• 2015

최근 중소규모 하천은 하천수 및 하천변 지하수 이용의 증가와 토지이용변화 등 유역 내 수문인자의 특성 변화로 인해 하천의 건천화가 점증하고 있어 하천 환경이 악화되고 있는 실정이다. 본 연구에서는 격자기반의 분포형 장기수문 모델(PGA-CC)을 이용하여 과거 수십년동안 토지이용변화에 따른 하천 건천화 영향을 평가하였다. 실제 건천화가 진행되고 있는 삽교천 상류유역($358.8km^2$)을 선정하였고 토지이용변화분석을 위해 과거 1975년 토지이용도(Past), 현재 2008년 토지이용도(Present)를 구축하였다. 각각의 토지이용 항목 중 변화율이 가장 높은 도시비율은 과거 토지이용도에서는 2.6 %였으며 현재 토지이용도에서는 11.3 %로 8.7 %가 증가하였다. 모델 검보정은 최근 7년(2005-2011)동안 최종유역출구지점에서 유출 검보정을 실시하였다. 그 결과 NSE (Nash-Sutcliffe model efficiency)은 평균 0.71로 유출량의 모의값과 실측값이 유효한 것으로 나타났다. 건천화를 평가하기 위해 시험유역에서의 5 WPs (Watching Points)를 선정하여 과거 및 현재 토지이용조건을 모의하고 유황분석을 통한 갈수 변화량 분석을 실시하였다. 건천화 빈도분석을 위해 GEV (Generalized Extreme Value) 갈수빈도분석을 실시하여 과거 토지이용 모의결과 산정된 평균 갈수량($m^3/s$) 이하로 낮아지는 유출량 일수를 계산하였다. 최종유역출구에서 과거 및 현재 토지이용도에서 모의된 평균갈수량은 각각 $3.27m^3/s$ 및 $3.11m^3/s$로 나타났다. GEV 갈수빈도분석결과 과거 토지이용조건에서의 평균갈수량은 $3.20m^3/s$(재현기간 2.33년)으로 나타났다. 도시증가에 따른 인구증가는 지하수 사용량에 증가를 가져온다. 이는, 건천화에 영향을 미치며 본 연구에서는 지하수이용량 자료(1998-2011)를 이용하여 도시면적과 지하수이용량의 선형회귀분석을 실시하여 과거 22년 지하수 사용량을 예측하였다. 그 결과 지하수사용량 증가는 토지이용변화와 복합적으로 상류유역에 하천의 변화를 가속시키는 것으로 나타냈다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.596-596
• /
• 2015

수공구조물 설계의 기본이 되는 설계홍수량은 정상성 (Stationary) 가정 하에 산정되고 있다. 정상성은 분포형의 매개변수들이 시간에 따라 변화하지 않는 것을 의미한다. 그러나 최근 기후변화로 인한 극치사상의 크기와 빈도가 비정상적인 증가 추세를 나타내고 있어 강우자료의 변화 특성을 정확하게 파악하기 위해서는 비정상성 (Non-Stationary)에 대한 고려가 필요한 실정이다. 따라서 본 연구에서는 비정상성을 고려한 미래 IDF (Intensity-Duration-Frequency) 곡선을 산정하고, 기후변화에 의한 IDF 곡선의 특성 변화를 분석하고자 한다. 연구대상지로 충청남도 공주시에 위치한 계룡저수지를 선정하였고, 계룡저수지에 인접한 대전 관측소의 과거 강우자료 (1981-2010년)를 수집하였다. 또한 기상청에서 제공하는 RCP4.5 시나리오 기반의 미래 강우자료를 편의보정하여 3개 기간 (2011-2040년, 2041-2070년, 2071-2100년)에 대한 미래 강우자료를 구축하였고, 지속시간별 연최대치 강우자료를 추출하여 경향성 분석 및 비정상성 빈도해석을 실시하였다. 강우자료의 확률분포형과 매개변수 추정방법으로는 GEV (Generalized Extreme Value)분포와 L-모멘트법을 선정하였다. 이를 바탕으로 과거 및 미래 기간별 IDF 곡선을 산정하였으며, 그 특성을 비교분석하였다. 본 연구에서 도출한 IDF 곡선은 계룡저수지의 설계한도를 재검토하는데 활용될 수 있으며, 본 연구에서 제시한 방법은 기후변화에 따른 농업용저수지의 안정성 검토에 사용할 수 있을 것으로 사료된다.