Jichul Ryu

Water Quality Prediction at Mandae Watershed using SWAT and Water Quality Improvement with Vegetated Filter Strip

Mandae watershed in Gangwon province has been known as one of soil erosion hot spot watersheds within Hanggang basin. Thus numerous efforts have been made to reduce soil erosion and pollutant loads into receiving watershed. However, proper best management practices have not been suggested because no monitoring flow and water quality data were available. Thus, modeling technique could not be utilized to evaluate water quality issue properly at Mandae watershed to develop and implement the best management practices. In this study, the SWAT model was applied to the Mandae watershed, Gangwon province to evaluate the SWAT prediction ability and water quality improvement with vegetated filter strip (VFS) in this study. The Nash-Sutcliffe model efficiency (NSE) and Coefficient of determination () values for flow simulation were 0.715 and 0.802, respectively, and the NSE and values were 0.903 and 0.920 for T-P simulation indicating the SWAT can be used to simulate flow and T-P with acceptable accuracies. The SWAT model, calibrated for flow and T-P, was used to evaluate water quality improvement with the VFS in agricultural fields. It was found that approximately 56.19 % of T-P could be reduced with vegetated filter strip of 5 m at the edge of agricultural fields within the watershed (34.86 % reduction with VFS of 1m, 48.29 % with VFS of 3 m). As shown in this study, the T-P, which plays key roles in eutrophication in the waterbodies, can be reduced with proper installation of the VFS.

Analysis of Soil Erosion Reduction Effect of Rice Straw Mat by the SWAT Model

The purpose of this study is to evaluate sediment yield reduction under various field slope conditions with rice straw mat. The Vegetative Filter Strip Model-W (VFSMOD-W) and Soil and Water Assessment Tool (SWAT) were used for simulation of sediment yield reduction effect of rice straw mat. The Universe Soil Loss Equation Practice factor (USLE P factor), being able to reflect simulation of rice straw mat in the agricultural field, were estimated for each slope with VFSMOD-W and measured soil erosion values under 5, 10, and 20 % slopes. Then with the regression equation for slopes, USLE P factor was derived and used as input data for each Hydrological Response Unit (HRU) in the SWAT model. The SWAT Spatially Distributed-HRU (SD-HRU) pre-processor module was utilized, moreover, in order to consider spatial location and topographic features (measured topographic features by field survey) of all HRU within each subwatershed in the study watershed. Result of monthly sediment yield without rice straw mat (Jan. 2000 - Aug. 2007) was 814.72 ton/month, and with rice straw mat (Jan. 2000 - Aug. 2007) was 526.75 ton/month, which was reduced as 35.35 % compared without it. Also, during the rainy season (from Jun. to Sep. 2000 - 2007), when without vs. with rice straw mat, monthly sediment indicated 2,109.54 ton and 1,358.61 ton respectively. It showed about 35.60 % was reduced depending on rice straw mat. As shown in this study, if rice straw mat is used as a Best Management Practice (BMP) in the sloping fields, rainfall-driven sediment yield will be reduced effectively.

Projecting Future Climate Change Scenarios Using Three Bias-Correction Methods

We performed bias correction in future climate change scenarios to provide better accuracy of models through adaptation to future climate change. The proposed combination of the change factor (CF) and quantile mapping (QM) methods combines the individual advantages of both methods for adjusting the bias in global circulation models (GCMs) and regional circulation models (RCMs). We selected a study site in Songwol-dong, Seoul, Republic of Korea, to test and assess our proposed method. Our results show that the combined CF

Application of the Load Duration Curve (LDC) to Evaluate the Achievement Rate of Target Water Quality in the Han-River Watersheds

AbstractWater quality in four major river basin in Korea was managed with Total Maximum Daily Load (TMDL) System. The unit watershed in TMDL system has been evaluated with Target Water Quality (TWQ) assessment using average water quality, without considering its volume of water quantity. As results, although unit watershed are obtained its TWQ, its allocated loads were not satisfied and vice versa. To solve these problems, a number of TWQ assessments with using Load Duration Curve (LDC) have been studied at other watersheds. The purpose of this study was to evaluate achievement of TWQ with Flow Duration Curve (FDC) and Load Duration Curve(LDC) at 26 unit watersheds in Han river basin. The results showed that achievement rates in TWQ assessment with current method and with LDC were 50~56 % and 69~73%, respectively. Because of increasing about 20% of achievement rates with using LDC, the number of exceeded unit watershed at Han river Basin was decreased about 4~6 unit watersheds.Key words:Han-River, LDC (Load Duration Curve), Target Water Quality, TMDLs (Total Maximum Daily Loads)

Enhancement of Estimation Method on the Land T-P Pollutant Load in TMDLs Using L-THIA

In this study, the uncertainty analysis of present land pollutant load estimation with simplified land category in TMDLs was performed and the enhanced method for land pollutant load estimation with level II land cover consisting of 23 categories was suggested, which was verified by L-THIA model. For land TP load estimation in Jinwi stream basin, the result of comparison between existing method with simplified land category (Scenario 1) and enhanced method with level II land cover (Scenario 2) showed high uncertainty in existing method. TP loads estimated by Scenario 2 for land covers included in the site land category were in the range of 3.45 to 56.69 kg/day, in which TP loads differed by sixteen times as much among them. For application of scenario 2 to TMDLs, Land TP loads were estimated by matching level II land cover to 28 land categories in serial cadastral map (Scenario 3). In order to verify accuracy of TP load estimation by scenario 3, the simulation result of L-THIA was compared with that and the difference between the two was as little as 10%. The result of this study is expected to be used as primary data for accurate estimation of land pollutant load in TMDLs.

Evaluation of SWAT Flow and Sediment Estimation and Effects of Soil Erosion Best Management Practices

Soil erosion and sediment from agricultural farmland has caused various negative impacts on environment in recent years. The effect of rice straw mat on soil erosion has been investigated by many researchers these days. In this study, the SWAT model was applied to Hongcheon watershed to evaluate SWAT flow and sediment, and the effect of rice straw mat on sediment yield at watershed outlet was evaluated. The Nash-Sutcliffe model efficiency (NSE) and coefficient of determination () values for flow simulation (calibration period) were 0.66 and 0.67, and the NSE values for sediment was 0.90. The calibrated parameters were used to analyze the reduction of sediment yield in the farmland with rice straw mat. Average daily sediment yield without rice straw mat was 49.8 ton/day and sediment yield with rice straw mat was 25.5 ton/day, and the reduction rate was 38.7 %. Also, average daily sediment yield with/without rice straw mat were 97.5 ton/day and 190.7 ton/day during the rainy season (Jun. 2008 - Aug. 2009), with the reduction rate 46.3 %.

Analysis of Baseflow Contribution to Streamflow at Several Flow Stations

Abstract Streamflow is typically divided into two components that are direct runoff and baseflow, it is required to analyze and estimate behaviors of those two flow components to understand watershed characteristics so that watershed management plan can be effective in pollutant reductions. Since pollutant load behaviors in a stream or river are variable by flow component behaviors , best management practices need to be applied in a watershed based on the pollutant load behaviors varying with flow components. Thus, baseflow behaviors were analyzed separating baseflow from streamflow data collected from fifteen streamflow gaging stations in the 4 major river watersheds which are the Han river, Nakdong river, Guem river, and Yeongsan·Somjin river watersheds. Moreover, precipitation trends throughout the 4 River Systems were investigated, thus daily precipitation data were collected from sixty-five locations. The Hank river watershed displayed the largest precipitation (925.2 mm) in summer but the lowest precipitation (71.8 mm) in winter, indicating the watershed has the most fluctuating precipitation characteristic. While the precipitation trends in the Four River Systems varied, a distinct feature in baseflow trends was not found, moreover baseflow percentages to streamflow were typically greater than 50% in the Four River Systems. As shown in this study, it would be expected significant amount of pollutants could be contributed to the stream in the form of baseflow at the watershed.

Evaluation of Runoff‧Peak Rate Runoff and Sediment Yield under Various Rainfall Intensities and Patterns Using WEPP Watershed Model

Recently, changes in rainfall intensity and patterns have been causing increasing soil loss worldwide. As a result, the water ecosystem becomes worse and crops yield are reduced with soil loss and nutrient loss with it. Many studies have been proposed to estimate runoff and soil loss to predict or decrease non-point source pollution. Although the USLE has been used for many years in estimating soil losses, the USLE cannot reflect effects on soil loss of changes in rainfall intensity and patterns. The WEPP, physically based model, is capable of predicting soil loss and runoff using various rainfall intensity. In this study, the WEPP model was simulated for sediment yield, runoff and peak runoff using data of 5, 10, 30, 60 minute term rainfall, Huff`s method and design rainfall. In case of rainfall interval of 5 minutes and 60 minutes, the sediment and runoff values decreased by 24% and 19%, respectively. The peak rate runoff values decreased by 16% when rainfall interval changed from 5 minutes to 60 minutes, indicating the peak rate runoff values are affected by rainfall intensity to some degrees. As a result of simulating using Huff`s method, all values (sediment yield, runoff, peak runoff) were found to be the greatest at third quartile. According to the analysis under various design rainfall conditions (2, 3, 5, 10, 20, 30, 50, 100, 200, 300 years frequency), sediment yield, runoff, and peak runoff of 906.2%, 249.4% and 183.9% were estimated using 2 year to 300 year frequency rainfall data.

Development of Automatic Extraction Model of Soil Erosion Management Area using ArcGIS Model Builder

Due to increased human activities and intensive rainfall events in a watershed, soil erosion and sediment transport have been hot issues in many areas of the world. To evaluate soil erosion problems spatially and temporarily, many computer models have been developed and evaluated over the years. However, it would not be reasonable to apply the model to a watershed if topography and environment are different to some degrees. Also, source codes of these models are not always public for modification. The ArcGIS model builder provides ease-of-use interface to develop model by linking several processes and input/output data together. In addition, it would be much easier to modify/enhance the model developed by others. Thus, simple model was developed to decide soil erosion hot spot areas using ArcGIS model builder tool in this study. This tool was applied to a watershed to evaluate model performance. It was found that sediment yield was estimated to be 13.7 ton/ha/yr at the most severe soil erosion hot spot area in the study watershed. As shown in this study, the ArcGIS model builder is an efficient tool to develop simple models without professional programming abilities. The model, developed in this study, is available at http://www.EnvSys.co.kr/~sateec/toolbox for free download. This tool can be easily modified for further enhancement with simple operations within ArcGIS model builder interface. Although very simple soil erosion and sediment yield were developed using model builder and applied to study watershed for soil erosion hot spot area in this study. The approaches shown in this study provides insights for model development and code sharing for the researchers in the related areas.

Development of Desktop-Based LDC Evaluation System for Effectiveness TMDLs

Load Duration Curve (LDC) can be used as a method for load management of point and non-point pollution source because the LDC easily assesses the water quality corresponding to hydrological changes in a watershed. Recently, the application of LDC to total pollution load management is a growing interest in Korea. In this regard, A desktop-based LDC assessment system was developed in this study to provide convenience to users in water quality evaluation. The developed system can simply produce the LDC by using streamflow and water quality data involved in its database. Also, The system can quantitatively inform the success or failure of the achievement for a target water quality at monthly scale. Furthermore, seasonal water quality and point/non-point pollution load in a watershed can be estimated by this system. We expect that the developed system will contribute to establish local and national policies regarding water management and total pollution load management because of its advantages such as the pollution tracking investigation and the analysis of water quality and pollution loading amount in an ungauged watershed.