The grid-based water balance of watershed scale was assessed in the mountainous area of Pyosun catchment in Jeju Island after analyzing precipitation, evapotranspiration, and runoff from January 2008 to December 2013. The existing results of direct runoff, evapotranspirtion, and groundwater recharge comparing to precipitation were presented 22.0%, 25.6%, and 52.4%, respectively, in Pyosun catchment. However, this study indicated each component shows 14.5%, 24.2%, and 61.0%, respectively, in the mountainous area of Pyosun catchment. Therefore, groundwater recharge rate in the mountainous area appears higher than 10% comparing to the overall catchment. It would be analyzed that the amount of direct runoff is relatively small. Moreover, this difference could be generated because of the spatial discontinuities in the process of estimating the total amount of precipitation in the mountainous area. Therefore, the grid-based spatial analysis to maximize the spatial continuity would be useful for providing a more reasonable result when the total amount of water resources are evaluated in mountainous areas in the future.

최근 국내에는 불규칙한 강우로 인해 수자원관리가 어려워 집중호우와 돌발홍수로 인한 피해가 급증하고 있는 상황이다. 이러한 피해를 최소화하기 위해 홍수 예경보 시스템구축을 위한 실시간 모니터링이나 강우-유출 모의 실험장치를 이용하여 모의 강우 적용 시 발생하는 유출량을 분석하여 그 대안을 찾고자 한다. 기존의 강우-유출 모의실험장치는 규모와 유지관리비용 측면에서 부담이 되는 실정이라 본 연구에서는 강우-유출 모의실험장치의 장단점을 보완한 소형강우모의장치의 설계와 개발 및 활용을 통해 인공강우 분사에 의한 시료의 포화와 지표 및 지하수 유출을 관측하였으며, TDR-300로 측정된 시료의 함수비는 강우강도를 112mm/hr로 설정하여 측정했고 실험 전, 후 차이를 비교 했을 때 모래의 경우 20.87% 증가하였고, 흙은 2.32% 증가하는 것을 확인 할 수 있었다. 또한 물수지방정식으로 분석한 결과 총 유입량이 28891.5 cm3였고 모래의 경우 지표유출량이 발생하지 않았으며 침투량은 26941.5 cm3, 함수량은 1950 cm3 로 산정되었고 흙의 경우 지표유출량이 22080 cm3였고, 침투량과 함수량은 각각 3480 cm3, 3331.5 cm3 였다. 이를 통해 소형강우모의장치의 효용성을 평가하였다.

과거 SWMM 모형은 수십 개소 주요하수관거의 매개변수를 추정하고, 유량측정자료를 활용한 검보정을 통하여 구축된다. 현재 하수관로 관련 수치지도의 양적, 질적 자료 향상으로 분포형 모형과 같은 수준의 SWMM 모형의 초기값 추정이 쉽지만, 매개변수 검보정 시간은 단축되지 않는다. 최적화된 매개변수와 유사한 매개변수의 초기값을 산정하면, 매개변수의 검보정 시간단축으로 분포형 모형과 같은 수준의 SWMM 모형 구축이 쉽다. 본 연구에서는 GIS 툴을 이용하여 SWMM 모형의 소유역의 매개변수를 산정하며, 동일한 산정방법을 적용하여 하수관거 정보를 단순화시킨 관로정보 개소수의 변화와 유역분할 방법에 따른 유출특성의 변화를 비교분석한다. 모의시간과 예측정밀도를 요구하는 홍수예경보시스템의 도시유출모형 구축을 위하여 기존 연구 결과에서 SWMM 모형의 매개변수 민감도 분석결과에서 검토하지 못했던 소유역 분할방법과 모의시간단축과 관계가 깊은 하수관거 정보 개소수 변화에 따른 유출특성의 변화를 도출하는 민감도 분석을 실시한다. 본 연구의 결과는 홍수예경보시스템 구축을 위하여 정밀하고 모의시간이 짧은 최적의 SWMM 모형 구축을 위한 자료로 이용될 것이다.

The rainfall-runoff characteristics in Jeju Island significantly differ from those in inland, due to highly permeable geologic features driven by volcanic island. Streams are usually sustained in the dry conditions and thereby the rainfall-runoff characteristics changes in terms of initiating stream discharge and its types, depending highly on the antecedent precipitation. Among various the rainfall-runoff characteristics, lag time mainly used for flood warning system in river and direct runoff ratio for determining water budget to estimate groundwater recharge quantity are practically crucial. They are expected to vary accordingly with the given antecedent precipitation. This study assessed the lag time in the measured hydrograph and direct runoff ratio, which are especially in the upstream watershed having the outlet as 2nd Dongsan bridge of Han stream, Jeju, based upon several typhoon events such as Khanun, Bolaven, Tembin, Sanba as well as a specific heavy rainfall event in August 23, 2012. As results, considering that the lag time changed a bit over the rainfall events, the averaged lag time without antecedent precipitation was around 1.5 hour, but it became increased with antecedent precipitation. Though the direct run-off ratio showed similar percentages (i.e., 23%)without antecedent precipitation, it was substantially increased up to around 45% when antecedent precipitation existed. In addition, the direct run-off ration without antecedent precipitation was also very high (43.8%), especially when there was extremely heavy rainfall event in the more than five hundreds return period such as typhoon Sanba.

The MFFn(Mass first flush), EMCs(Event mean concentrations) and runoff loads were analyzed for various rainy events(monitoring data from 2011 to 2012) in transportation area(rail road in station). The pollutant EMCs by volume of stormwater runoff showed the BOD5 9.6 ㎎/L, COD 29.9 ㎎/L, SS 16.7 ㎎/L, T-N 3.271 ㎎/L, T-P 0.269 ㎎/L in the transportation areas(Railroad in station). The average pollutant loading by unit area of stormwater runoff showed the BOD5 27.26 kg/㎢, COD 92.55 kg/㎢, SS 50.35 kg/㎢, T-N 10.13 kg/㎢ and T-P 10.13 kg/㎢ in the transportation areas. Estimated NCL-curve(Normalized cumulated-curve) was evaluated by comparison with observed MFFn. MFFn was estimated by varying n-value from 10% to 90% on the rainy events. The n-value increases, MFFn is closed to '1'. As time passed, the rainfall runoff was getting similar to ratio of pollutants accumulation. The result of a measure of the strength of the linear relationship between observed data and expected data under model was good.

본 연구에서는 기상청에서 제공하는 RCP 기후변화시나리오를 이용하여 기후변화와 유역유출특성에 따른 환경영향을 평가하였다. SWAT모형을 이용한 미래 댐 유입량 평가, HEC-ResSim 모형을 이용한 댐 안전성 평가 및 하류 유황분석을 수행하였다. 또한, 기존 수질관측자료를 이용하여 Seasonal-Kendall Test를 통해 수질변화 추세에 대해 분석하였고, SWAT과 HEC-ResSim 모형으로 도출된 방류량 및 지류유출량을 Qual2E모형에 적용하여 미래 수질변화 추세에 대해 분석하였다. 다음과 같은 유역 통합환경검토 기법은 하천환경에 대한 과학적 물환경 관리 체계를 확보하고, 기후변화 등 새로운 환경문제에 선제적으로 대응하기 위한 지침을 마련할 수 있을 것이다.

도심지 지표에 내린 우수유출은 자연유역과는 다르게 다양한 지형지물의 영향을 받아 유하하다가 통상 도로의 빗물받이를 통해 관망으로 유입되며 이후 관망시스템을 통해 하천이나 저류지로 방류되는 형태로 이루어진다. 하지만 현재 실무에서 활용되고 있는 1차원 상용모형은 이와 같은 현상을 명확하게 고려하지 못하고 있다. 우리나라에서 현재 도심지 유출해석시 주로 이용하고 있는 SWMM모형의 경우 지표부분에 대한 해석은 소유역을 구분하고 소유역별로 비선형저류방정식을 적용하여 소유역에 대한 유출량을 산정하는 방식을 활용하고 있기 때문에 도심지의 다양한 지형지물에 대한 고려가 어려울 수밖에 없으며, 특히 빗물받이를 통해 관망으로 유입되는 부분은 생략되어 계산되고 있는 실정이다. 따라서 본 연구에서는 도심지 배수특성을 고려한 격자기반 유출해석기법을 개발하고 Test-Bed에 적용해 보았다.

안전한 하천구조물을 설계하기 위해서는 신뢰도 있는 홍수량 산정이 필요하다. 신뢰도 있는 홍수량 산정은 유역의 정확한 지역적 특성의 반영이 중요하지만, 실제적으로는 유역에 대한 지역적 특성의 정밀한 측정이 어렵다. 본 연구에서는 홍수량 산정 시 발생 할 수 있는 오차들을 최대한으로 줄이기 위하여 여러 개념적 강우 유출 모형을 적용 및 평가 하고자 한다. 이를 위하여 3개의 토양저류함수모형(Soil Moisture Accounting, SMA) 및 3개의 유역유출개념모형(Routing Modules)을 조합하여 총 9개의 모형을 금강의 22개소 유역에 적용한다. 적용결과 3개의 토양 저류 함수 모형 중 확률분포모형(Probability Distribution Model, PDM)이 Nash Surcliffe Efficiency (NSE*) 결과에서 타 토양 저류 함수 모형보다 우수하게 나타난다. 또한 유역유출모형에 대해서 우수한 NSE*값을 나타낸 모형은 2PMP(Micro-pre Approach parallel structure)모형으로 확인되었다. 따라서 금강 22개소 유역에서 두 모형의 조합인 PDM-2PMP모형이 적용 가능하다고 평가 할 수 있다. 이는 향후 금강유역의 지역화 연구 및 우리나라 전 유역에 적용 가능한 강우 유출모형 개발을 위한 기초연구로서 높은 활용가치를 가질 것으로 판단된다.

This study conducted a laboratory simulation using artificial and natural rainfall in order to investigate the runoff characteristics of livestock resources through the analysis of the surface runoff water and infiltration water by rainfall intensity and fertilization level. Cattle manure and pig liquid fertilizer was used as livestock resource. As a result of this study, it was observed that the surface runoff occurred over 32 mm/hr rainfall intensity, and flow rate of the surface runoff water and the runoff ratio of contaminant parameters from livestock resource was increased, as rainfall intensity was stronger. With doubled fertilization level, T-N increased in compost and the amount of CODMn runoff also considerably increased in liquefied fertilizer. In the case of natural rainfall, the runoff ratio of T-P clearly increased in compost and the T-N of final surface runoff ratio in compost and liquefied fertilizer was ranged from 0.13047 to 0.13623 with stronger rainfall intensity.

The objective of this paper is compare to landuse type for calculating peak flood and soil loss in rapidly expansion urban area. This study compares two landuse maps, including numerical landuse map and aerial photograph landuse map, for calculating the ratio of urban and agriculural area, curve number, time of concentration, peak flood discharge, and soil loss. It is found that flood discharge calculated using aerial photograph landuse map are larger than that calculated using numerical landuse map, and soil loss calculated using aerial photograph landuse map are smaller than that calculated using numerical landuse map. Results also indicate that landuse chage in rapidly expansion urban area significantly influences flood discharge and soil loss.

For Kangjeong stream and Akgeun stream in the central part of the southern Jeju Island, on-site discharge estimation was carried out for approximately 10 months (July 2011-April 2012) twice a month on a regular basis by using ADCP (acoustic doppler current profiler) and long term rate of discharge was calculated by using SWAT (soil and water assessment tool) model. The discharge was 0.28-1.30 m 3 /sec for Kangjeong stream and 0.10-1.54 m 3 /sec for Akgeun stream. It showed the maximum in the summer and the minimum in the winter. As a result of parameter sensitivity analysis of SWAT model, CN (NRCS runoff curve number for moisture condition Ⅱ), SOL_AWC (available water capacity of the soil layer), and ESCO (soil evaporation compensation factor) showed sensitive responses. By using the result, the model was corrected and the rate of discharge was calculated. As a result, the annual discharge rate was 27.12-31.86(%) at the Akgeun basin and 23.55-28.43(%) at the Kangjeong basin.

Jeju Island, the heaviest raining area in Korea, is a volcanic Island located at the southernmost of Korea, but most streams are of the dry due to its hydrological/geological characteristics different from those of inland areas. Therefore, there are limitations in applying the results from the mainland to the studies on stream run-off characteristics analysis and water resource analysis of Jeju Island. In this study, the SWAT(soil & water assessment tool) model is used for the Hwabuk stream watershed located east of the downtown to calculate the long-term stream run-off rate, and WMS(watershed modeling system) and HEC-HMS(hydrologic modeling system) models are used to figure out the stream run-off characteristics due to short-term heavy rainfall. As the result of SWAT modelling for the long-term rainfall-runoff model for Hwabuk stream watershed in 2008, 5.66% of the average precipitation of the entire basin was run off, with 3.47% in 2009, 8.12% in 2010, and root mean square error(RMSE) and determination coefficient(R2) was 496.9 and 0.87, respectively, with model efficient(ME) of 0.72. From the results of WMS and HEC-HMS models which are short-term rainfall-runoff models, unless there was a preceding rainfall, the runoff occurred only for rainfall of 40mm or greater, and the run-off duration averaged 10～14 hours.

In this study, a method of simulating ephemeral stream runoff characteristics in Jeju watershed is newly suggested. The process based conceptual-physical scheme is established based on the SWAT-K and applied to Cheonmi-cheon watershed which shows the typical pattern of ephemeral stream runoff characteristics. For the proper simulation of this runoff, the intermediate flow and baseflow are controlled to make downward percolation should be dominant. The result showed that surface runoff simulated by using the modified scheme showed good agreement with observed runoff data. In addition, it was found that the estimated runoff directly affected the groundwater recharge rate. This conceptual model should be continuously progressed including rainfall interception, spatially estimated evapotranspiration and so forth for the reasonable simulation of the hydrologic characteristics in Jeju island.

This study was to investigate the runoff characteristics of non-point pollutants source at the urban area in Suwon city. The highest T-N and T-P concentration of rainfall runoff observed in agricultural area. In residential area, the highest BOD5 and SS concentration of rainfall runoff was investigated. During rainfall events, the peak concentrations of SS and BOD5 were observed after 1~2 hours of rainfall in urban area. Whereas, the peak concentrations occurred within 1~2 hours after rainfall and then the highest concentrations of SS and BOD5 sharply decreased, showing strong first flush effect in urban area. The EMC results indicated that the highest value of T-N and T-P in agricultural area was observed. While residential area was shown the lowest EMC value as T-N and T-P. Non-point pollutant loads on the land use types in urban area were investigated in the order of residential>industrial>agricultural>highway. BOD5 and SS loads on urban watershed were investigated in the order of Suwon>Hwangguji>Seoho>Wonchunri. Whereas, T-N and T-P loads on urban watershed were investigated in the order of Hwangguji>Suwon>Wonchunri>Seho.

In this study, non-point source(NPS) contribution was investigated based on flow rates and water qualities of streams into the lake during rainfall events. Event mean concentration(EMC) and the pollution loads were calculated to establish a database for NPS control measurement in the survey area, and so on. The runoff characteristics of NPS were investigated and estimated on the basis of the ratio of an agricultural to forest area in the stream of sub-catch basin during rainfall events. Non-point source pollution loads were also calculated to establish a database for NPS control measure in the upstream lake Chinyang. At a rainfall event, BOD concentrations rise sharply at the early peak time of runoff, however, peaks of TSS concentration were observed at the similar time of peak flow. This was a phenomenon shown at the watersheds caused by forest and geological types. The discharged EMC range was 2.9-4.8 mg/L in terms of BOD. The discharged EMC range was 6.2-8.2 mg/L in terms of SS. The discharged EMCs of T-N and T-P were 1.4-2.5 mg/L and 0.059-0.233 mg/L, respectively. Total BOD loading rate through the 3 tributaries to the lake Chinyang was 1,136 kg/d during dry weather. The upper watershed area of the Nam-river dam in this study was divided into 14 catchment basins based on the Korean guideline for total maximum daily load(TMDL) of water quality pollutants. The higher the agricultural land-use ratio, the more NPS loading rate discharged, but the more occupied a forest area, the lower more NPS loading rate discharged. In an agricultural land-use area more than 20%, the increase of NPS loadings might be dramatically diffused by increasing the integrated complex-use like vinyl-house facilities and fertilizer use etc. according to the effective land-use utilization. The NPS loading rates were BOD 0.3 kg/ha․day, SS 0.21 kg/ha․day, TN 0.02 kg/ha․day, TP 0.005 kg/ha․day under less than 10% agricultural land-use. In agricultural land-use of 20%-50%, these values were investigated in the range of 0.32 kg/ha․day-0.73 kg/ha․day for BOD, 0.92 kg/ha․day-3.32 kg/ha․day for SS, 0.70 kg/ha․day-0.90 kg/ha․day TN, 0.03 kg/ha․day-0.044 kg/ha․day for TP.

This study analyzed the characteristics of stormwater runoff by rainfall type in orchard areas and transportation areas for 2 years(2010~2011year). Effluents were monitored to calculate the Event Mean Concentrations(EMCs) and runoff loads of each pollutant. The pollutant EMCs by volume of stormwater runoff showed the ranges of BOD 0.9~13.6 ㎎/L, COD 13.7~45.2 ㎎/L, SS 4.1~236.4 ㎎/L, T-N 2.123~21.111 ㎎/L, T-P 0.495~2.214 ㎎/L in the orchard areas, and was calculated as BOD 2.3~22.5㎎ /L, COD 4.4~91.1 ㎎/L, SS 4.3~138.3 ㎎/L, T-N 0.700~13.500 ㎎/L, T-P 0.082~1.345 ㎎/L in the transportation areas. The correlation coefficient of determination in the orchard area was investigated in the order of Total Rainfall(0.81) > Total Runoff(0.76) > Rainfall Intensity(0.56) > Rainfall Duration(0.46) > Antecedent Dry Days(0.27). Also, in the case of the transportation area was investigated in the order of Total Rainfall (0.55) > Total Runoff(0.54) > Rainfall Intensity(0.53) > Rainfall Duration(0.24) > Antecedent Dry Days(0.14). As the result, comparing valuables relating to runoff of non-pollutant source between orchard areas and transportation areas, orchard area(R2 ≥ 0.5 : X3, X4, X5) was investigated to have more influence of diverse independent valuables compared to the transportation area(R2 ≥ 0.5 : X3, X4) and the difference of discharge influence factor by the land characteristics appeared apparently.

홍수기 자연하천의 효율적인 홍수제어를 위해서는 정확한 유출량을 예측하는 것이 가장 중요하다. 본 연구에서는 신흥천 하류에 위치한 구재교 지점의 홍수기 유출특성을 HEC-HMS 모형을 이용하여 분석하였다. 분석을 위한 수문기초 자료인 유역의 물리적 인자들은 지형정보시스템을 이용하여 추출하였으며 유출모형의 초기 매개변수는 추출한 유역의 물리적 인자들을 이용하여 추정하여 사용하였다. HEC-HMS 모형을 이용하여 정확한 유출량을 예측하기 위해서 2년간의 관측유량자료와 모의유출량을 비교하여 최적화된 매개변수를 산정하였다. 최적화과정은 HEC-HMS 모형에서 제공하는 매개변수 최적화 과정을 이용하였으며 N&M(Nelder-Mead Method) 방법을 사용하였다. 최적화된 매개변수는 향후 구재교 지점에 대한 홍수유출을 모의할 때 기초자료로 제공하고자 한다.

최근 지구온난화에 따른 기후변화로 인하여 집중호우 및 가뭄의 발생빈도가 증가하고 있으며 더불어 홍수와 하천의 건천화가 지역의 물 관리에 영향을 미치고 있다. 하천 본류의 경우 지속적인 모니터링과 관리를 통하여 홍수기와 이수기의 유출특성을 파악하고 있으나 지류하천까지는 어려움이 따르고 있다. 본 연구에서는 지류하천의 홍수기 유출특성을 파악하기 위하여 만경강의 지류하천인 소양천 삼중일교 지점에서 유량측정을 실시하였으며, HEC-HMS 모형을 이용하여 분석된 모의유출량과 관측유량을 비교 분석하였다. 분석을 위한 수문기초 자료는 지형정보시스템을 이용하여 추출하였으며 추출된 지형인자를 기반으로 모형의 초기 매개변수를 적용하였다. 관측유량과 초기 매개변수를 적용하여 추정된 모의유량을 비교하여 모형의 보정을 실시하고 호우사상에 대한 평균상대오차를 나타내었다.

국내 최초의 분산형 생태환경저류지 시스템으로 방재기능뿐만 아니라 초기 비점오염원에 대한 생태적 수질정화 및 생태복원, 근린공원의 친수적 이용성, 토지이용도 제고 등의 다양한 기능을 융합하여 대상지 특성에 적합한 다목적용 저류지가 조성되었다. 복원 후 수리·수문 모니터링 결과는 서울지역 재현기간 10년의 강우량과 홍수량에서 첨두홍수량은 12% 감소하고 20분의 지체효과를 보였다. 생태환경수로의 하상 평균소류력의 범위는 86.29-376.18N/m2, 유속의 범위는 2.76-10.07m/s이었다. 하상보호공은 평균 소류력이 133.94N/m2 이하인 곳에는 자연석을, 133.94N/m2 이상인 곳에는 생태목틀이 적용되었다. 현장조사결과 생태목틀과 자연석이 적용된 생태환경수로는 자연석의 이탈이나 자연석 주변 하상 세굴과 토사의 퇴적이 발생하지 않았음을 확인하였다. 생태환경 저류지 시스템의 홍수저감효과와 생태환경 수로의 안정성을 입증하였다. 국내 융합적인 생태환경 저류지나 생태하천의 표준화와 적용에 필요한 중요한 기준을 확립하는데 기여할 수 있을 것이다. 다만 생태환경저류지 시스템 조성이후 생태적 안정화 기간인 2년 이상의 연구개발자(전문가)의 유지관리 참여가 필요하다. 이를 통해 일반기술자도 생태환경저류지의 조성 및 유지관리가 용이하도록 지침을 마련해야 할 것이다.

In this study, occurrence status of nonpoint pollutants and characteristics of discharge by each nonpoint pollutants were examined through monitoring on nonpoint pollutants caused when raining in vineyard belonging to the agricultural area of various land use patterns. Also, the first flush analysis limited to studies on the existing non-percolation area was applied to percolation area to ascertain availability and criteria of study. Various water quality and sluice of nonpoint pollutants were analyzed, based on which discharge of nonpoint pollutants in agricultural area was ascertained to be influenced greatly by artificial factors such as period, cultivation, management, etc. Meanwhile, the first flush phenomenon at agricultural area was ascertained to occur, and the first flush was quantified through calculation of the first flush ratio. If MFF30 is based, discharge load by each nonpoint pollutants caused when raining was investigated to include 40.8% on the basis of total discharge. In case of SS in pollutants showed the highest first flush phenomenon of 64.8%. Through such a result, calculation possibility of the initial rain criteria was ascertained, and it was determined that reliability-assured criteria were calculated through further monitoring.