• Korean Journal of Remote Sensing
• /
• v.37 no.6_1
• /
• pp.1589-1597
• /
• 2021

Submarine groundwater discharge (SGD) in coastal areas is gaining importance as a major transport route that bring nutrients and trace metals into the ocean. This paper describes the analysis of the seasonal changes and spatiotemporal characteristicsthrough the modeling monthly SGD for 35 years from 1986 to 2020 for the Nakdong river basin. In this study, we extracted 210 watersheds and SGD estimation points using the SRTM (Shuttle Radar Topography Mission) DEM (Digital Elevation Model). The average annual SGD of the Nakdong River basin was estimated to be 466.7 m²/yr from the FLDAS (Famine Early Warning Systems Network Land Data Assimilation System) recharge data of 10 km which is the highest resolution global model applicable to Korea. There was no significant time-series variation of SGD in the Nakdong river basin, but the concentrated period of SGD was expanded from summer to autumn. In addition, it was confirmed that there is a large amount of SGD regardless of the season in coastal area nearby large rivers, and the trend has slightly increased since the 1980s. The characteristics are considered to be related to the change in the major precipitation period in the study area, and spatially it is due to the high baseflow-groundwater in the vicinity of large rivers. This study is a precedentstudy that presents a modeling technique to explore the characteristics of SGD in Korea, and is expected to be useful as foundational information for coastal management and evaluating the impact of SGD to the ocean.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.88-88
• /
• 2021

최근 수자원과 수질관리 분야에 자료기반 머신러닝 모델과 딥러닝 모델의 활용이 급증하고 있다. 그러나 딥러닝 모델은 Blackbox 모델의 특성상 고전적인 질량, 운동량, 에너지 보존법칙을 고려하지 않고, 데이터에 내재된 패턴과 관계를 해석하기 때문에 물리적 법칙을 만족하지 않는 예측결과를 가져올 수 있다. 또한, 딥러닝 모델의 예측 성능은 학습데이터의 양과 변수 선정에 크게 영향을 받는 모델이기 때문에 양질의 데이터가 제공되지 않으면 모델의 bias와 variation이 클 수 있으며 정확도 높은 예측이 어렵다. 최근 이러한 자료기반 모델링 방법의 단점을 보완하기 위해 프로세스 기반 수치모델과 딥러닝 모델을 결합하여 두 모델링 방법의 장점을 활용하는 연구가 활발히 진행되고 있다(Read et al., 2019). Process-Guided Deep Learning (PGDL) 방법은 물리적 법칙을 반영하여 딥러닝 모델을 훈련시킴으로써 순수한 딥러닝 모델의 물리적 법칙 결여성 문제를 해결할 수 있는 대안으로 활용되고 있다. PGDL 모델은 딥러닝 모델에 물리적인 법칙을 해석할 수 있는 추가변수를 도입하며, 딥러닝 모델의 매개변수 최적화 과정에서 Cost 함수에 물리적 법칙을 위반하는 경우 Penalty를 추가하는 알고리즘을 도입하여 물리적 보존법칙을 만족하도록 모델을 훈련시킨다. 본 연구의 목적은 대청호의 수심별 수온을 예측하기 위해 역학적 모델과 딥러닝 모델을 융합한 PGDL 모델을 개발하고 적용성을 평가하는데 있다. 역학적 모델은 2차원 횡방향 평균 수리·수질 모델인 CE-QUAL-W2을 사용하였으며, 대청호를 대상으로 2017년부터 2018년까지 총 2년간 수온과 에너지 수지를 모의하였다. 기상(기온, 이슬점온도, 풍향, 풍속, 운량), 수문(저수위, 유입·유출 유량), 수온자료를 수집하여 CE-QUAL-W2 모델을 구축하고 보정하였으며, 모델은 저수위 변화, 수온의 수심별 시계열 변동 특성을 적절하게 재현하였다. 또한, 동일기간 대청호 수심별 수온 예측을 위한 순환 신경망 모델인 LSTM(Long Short-Term Memory)을 개발하였으며, 종속변수는 수온계 체인을 통해 수집한 수심별 고빈도 수온 자료를 사용하고 독립 변수는 기온, 풍속, 상대습도, 강수량, 단파복사에너지, 장파복사에너지를 사용하였다. LSTM 모델의 매개변수 최적화는 지도학습을 통해 예측값과 실측값의 RMSE가 최소화 되로록 훈련하였다. PGDL 모델은 동일 기간 LSTM 모델과 동일 입력 자료를 사용하여 구축하였으며, 역학적 모델에서 얻은 에너지 수지를 만족하지 않는 경우 Cost Function에 Penalty를 추가하여 물리적 보존법칙을 만족하도록 훈련하고 수심별 수온 예측결과를 비교·분석하였다.

• Korean Journal of Ecology and Environment
• /
• v.50 no.1
• /
• pp.1-15
• /
• 2017

Daecheong Reservoir was made by the construction of a large dam (>15 m in height) on the middle to downstream of the Geum River and the discharge systems have the watergate-spillway (WS), a hydropower penstock (HPP), and two intake towers. The purpose of this study was to investigate the limnological anomalies of turbid water reduction, green algae phenomenon, and oligotrophic state in the lower part of reservoir dam site, and compared with hydro-meteorological factors. Field surveys were conducted in two stations of near dam and the outlet of HPP with one week intervals from January to December 2000. Rainfall was closely related to the fluctuations of inflow, outflow and water level. The rainfall pattern was depended on the storm of monsoon and typhoon, and the increase of discharge and turbidity responded more strongly to the intensity than the frequency. Water temperature and DO fluctuations within the reservoir water layer were influenced by meteorological and hydrological events, and these were mainly caused by water level fluctuation based on temperature stratification, density current and discharge types. The discharges of WS and HPP induced to the flow of water bodies and the outflows of turbid water and nutrients such as nitrogen and phosphorus, respectively. Especially, when hypoxic or low-oxygen condition was present in the bottom water, the discharge through HPP has contributed significantly to the outflow of phosphorus released from the sediment into the downstream of dam. In addition, HPP effluent which be continuously operated throughout the year, was the main factor that could change to a low trophic level in the downreservoir (lacustrine zone). And water-bloom (green-tide) occurring in the lower part of reservoir was the result that the water body of upreservoir being transported and diffused toward the downreseroir, when discharging through the WS. Finally, the hydropower effluent was included the importance and dynamics that could have a temporal and spatial impacts on the physical, chemical and biological factors of the reservoir ecosystem.

• Economic and Environmental Geology
• /
• v.40 no.3 s.184
• /
• pp.277-293
• /
• 2007

Seasonal and spatial variations in the concentrations of trace elements, pH and Eh were found in a creek watershed affected by mine drainage and leachate from several waste rock dumps within the As-Pb-rich Indae mine site. Because of mining activity dating back to about 40 years ago and rupture of the waste rock dumps, this creek was heavily contaminated. Due to the influx of leachate and mine drainage, the water quality of upstream reach in this creek was characterized by largest seasonal and spatial variations in concentrations of Zn(up to $5.830 mg/{\ell}$), Cu(up to $1.333 mg/{\ell}$), Cd(up to $0.031 mg/{\ell}$) and $SO_4^{2-}$(up to $173 mg/{\ell}$), relatively acidic pH values (3.8-5.1) and highly oxidized condition. The most abundant metals in the leachate samples were in order of Zn($0.045-13.909 mg/{\ell}$), Fe($0.017-8.730mg/{\ell}$), Cu($0.010-4.154mg/{\ell}$) and Cd($n.d.-0.077mg/{\ell}$), with low pH(3.1-6.1), and high $SO_4^{2-}$(up to $310 mg/{\ell}$). The mine drainage also contained high concentrations of Zn, Cu, Cd and $SO_4^{2-}$ and remained constantly near-neutral pH values(6.5-7.0) in all the year. While the leachate and mine drainage might not affect short-term fluctuations in flow, it may significantly influence the concentrations of chemicals in the stream. The abundance and chemistry of Fe-(oxy)hydroxide within this creek indicated that the Fe-(oxy)hydroxide formation could be responsible for some removal of trace elements from the creek waters. Spatial and seasonal variations along down-stream reach of this creek were caused largely by the influx of water from uncontaminated tributaries. In addition, the trace metal concentrations in this creek have been decreased nearly down to the background level at a short distance from the discharge points without any artificial treatments after hydrologic mixing in a tributary. The nonconservative(i.e. precipitation, adsorption, oxidation, dissolution etc.) and conservative(hydrologic mixing) reactions constituted an efficient mechanism of natural attenuation which reduces considerably the transference of trace elements to rivers.

• Korean Journal of Ecology and Environment
• /
• v.49 no.3
• /
• pp.153-175
• /
• 2016

Daecheong Reservoir has suffered eutrophication and water-blooms by blue-green algae from initial impoundment, and algae alert system (AAS) was introduced in 1997. The purpose of this study was to investigate the effect of rainfall and hydrological factors in increase or decrease variability of green-tide and prolonged AAS, studied and analyzed the current situation of AAS has been operating for 19 years (1997~2015) in Daecheong Reservoir. The total issued number of AAS was 46 times, the most frequent period in August and September were 22 times (752 days) and 16 times (431 days), respectively, it accounted for 82.6%. Many number and frequency during this period were significantly associated with rainfall, various discharge and water level. Rainfall and hydrological events are associated with the rainy season of monsoon-Changma and the typhoon, it was concentrated in June~September, total rainfall in this period accounted for 69.9% of the annual rainfall. An increase in inflows was dependent on the intensity, frequency and the amount of rainfall. Accounted for 68.4% of the total annual inflow, it was a time when the most rapidly changing hydrological variability in the reservoir. The total outflow was closely related to rainfall, and compared the distinctive characteristics of hydropower generation and watergate-spillway discharge. In addition, the upreservoir zone of Daecheong Reservoir could be vulnerable to green-tide by regulating discharge of the upstream dam. The issue of AAS was strongly related to the with and without of watergate-spillway discharge. The watergate-spillway discharge had a total of 25 times, it was maximum 17 days from July to September in the year. And the opening times and each duration of the watergate were 1~4 times and the range of 3~37 days, respectively. When the watergate opened, the issue of AAS was maintained to 13 years and the movement of water bodies and green-tide was great about five times than that of non-open, had a profound effect on prolonged AAS within reservoir. In Daecheong Reservoir, Chusori (CHU) area of the So-ok Stream was still showing serious symptoms green-tide levels in the summer, but Janggye (JAN) waters of the main reservoir was pointed out that more important. AAS will be operated by an absolutely consider the rainfall and hydrological effects around the watergate-spillway discharge. The measures of green-tide will be included in the limnological studies more suited to the characteristics of the watershed and reservoir of the our country. Finally, from now on, we will prepare the systematic management and guidelines for vulnerable zone water-blooms that are the source within the reservoir before the monsoon rather than waiting for the arrival of green-tide on the operating stations of AAS.