• 한국환경과학회지
• /
• 제26권10호
• /
• pp.1135-1146
• /
• 2017

At riverbank filtration sites, groundwater levels of alluvial aquifers near rivers are sensitive to variation in river discharge and pumping quantities. In this study, the groundwater level fluctuation, pumping quantity, and streamflow rate at the site of a riverbank filtration plant, which produces drinking water, in the lower Nakdong River basin, South Korea were interrelated. The relationship between drawdown ratio and river discharge was very strong with a correlation coefficient of 0.96, showing a greater drawdown ratio in the wet season than in the dry season. Autocorrelation and cross-correlation were carried out to characterize groundwater level fluctuation. Autoregressive model analysis of groundwater water level fluctuation led to efficient estimation and prediction of pumping for riverbank filtration in relation to river discharge rates, using simple inputs of river discharge and pumping data, without the need for numerical models that require data regarding several aquifer properties and hydrologic parameters.

• 한국농공학회지
• /
• 제39권6호
• /
• pp.88-98
• /
• 1997

In this study, pumping capacity in the area of natural river estuary is estimated by the quantitative analysis using finite element method. The study area is Iga-ri pumping station in the estuary of Seojung-chun which runs to the East coast. First of aH, hydraulic properties and effects of different seawater levels are analyzed in this area. Variations of groundwater level caused by pumping and properties of seawater intrusion are analyzed, then compared the case of reinforcing the existing intake weir with the case of setting up an weir at the upper stream. The observed data of groundwater drawdown caused by pumping during drought period and seawater intrusion are compared with results of the analysis done by groundwater model using finite element method, and it is found that both are similar. Accordingly, groundwater model used in this study reflects well the variation of groundwater level caused by pumping.

• 한국환경과학회지
• /
• 제18권3호
• /
• pp.315-323
• /
• 2009

Hydraulic gradient of the landfill soils is estimated by Devlin (2003) method, and its variation characteristics from rainfall and permeability of the aquifer material are analyzed. The study site of 18 m $\times$ 12 m is located in front of the Environment Research Center at the Pukyong National University, and core logging, slug/bail test and groundwater monitoring was performed. The sluglbail tests were performed in 9 wells (except BH9 well), and drawdown data with elapsed time for bail tests were analyzed using Bouwer-Rice and Hvorslev methods. The average hydraulic conductivity estimated in each of the test wells was ranged $1.991{\times}10^{-7}{\sim}4.714{\times}10^{-6}m/sec$, and the average hydraulic conductivity in the study site was estimated $2.376{\times}10^{-6}m/sec$ for arithmetic average, $1.655{\times}10^{-6}m/sec$ for geometric average and $9.366{\times}10^{-7}m/sec$ for harmonic average. The permeability of landfill soils was higher at the east side of the study site than at the west side. Groundwater level in 10 wells was monitored 44 times from October 2 to November 7, 2007. The groundwater level was ranged 1.187$\sim$1.610 m, and the average groundwater level range in each of the well showed 1.256$\sim$1.407 m. The groundwater level was higher at the east side than at the west side of the study site, and this distribution is identify to it of hydraulic conductivity. The hydraulie gradient and the major flow direction for 10 wells were estimated 0.0072$\sim$0.0093 and $81.7618{\sim}88.0836^{\circ}$, respectively. Also, the hydraulic gradient and the major flow direction for 9 wells were estimated 0.0102$\sim$0.0124 and $84.6822{\sim}89.1174^{\circ}$, respectively. The hydraulic gradient of the study site increased from rainfall (83.5 mm) on October 7, causing by that the groundwater level of the site with high permeability was higher. The hydraulic gradient estimated on and after October 16 was stable, due to almost no rainfall. Thus, it was confirmed that the variation of the hydraulic gradient in the landfill soils was controlled by the rainfall.

• 한국지하수토양환경학회지:지하수토양환경
• /
• 제27권6호
• /
• pp.22-36
• /
• 2022

Water resources in Jeju Island are dependent virtually entirely on groundwater. For groundwater resources, drought damage can cause environmental and economic losses because it progresses slowly and occurs for a long time in a large area. Therefore, this study quantitatively evaluated groundwater level fluctuations using principal component and cluster analyses for 42 monitoring wells in Jeju Island, and further identified the types of groundwater fluctuations caused by drought. As a result of principal component analysis for the monthly average groundwater level during 2005-2019 and the daily average groundwater level during the dry season, it was found that the first three principal components account for most of the variance 74.5-93.5% of the total data. In the cluster analysis using these three principal components, most of wells belong to Cluster 1, and seasonal characteristics have a significant impact on groundwater fluctuations. However, wells belonging to Cluster 2 with high factor loadings of components 2 and 3 affected by groundwater pumping, tide levels, and nearby surface water are mainly distributed on the west coast. Based on these results, it is expected that groundwater in the western area will be more vulnerable to saltwater intrusion and groundwater depletion caused by drought.

• 한국지하수토양환경학회지:지하수토양환경
• /
• 제25권3호
• /
• pp.1-11
• /
• 2020

In this study, leakage ratios of Jeju Island's public agricultural groundwater were calculated by utilizing field measurements of groundwater level and surface reservoir water level. The average leakage ratios were 75.6% at groundwater well A and 57.5% at well B, with the ratio inversely proportional to agricultural water usage. The level of agricultural reservoirs varied at constant intervals at night, and the amount of water leakage associated with the variation was estimated as 0.1 - 16.3 ㎥/h. The leakage ratio was also influenced by pipeline length, average slope, and number of farmhouses. Currently, the estimation of agricultural water leakage on Jeju Island is based upon field inspection which is very labor- and cost intensive. The leakage ratio estimated by monitoring the reservoirs associated with the well A and B were 73.3 and 54.7%, respectively, consistent with the values obtained by field measurements.

• 한국지하수토양환경학회지:지하수토양환경
• /
• 제19권4호
• /
• pp.12-22
• /
• 2014

This study evaluated the pattern of groundwater fluctuation in cyrstalline rock using time series and factor analyses. From the results, groundwater level for the 18 wells was classified into 4 types reflecting the hydrogeological properties and rainfall event. Type 1 (DB1-5, DB1-6, DB2-2, KB-10, KB-13) was significantly influenced by groundwater flow through water-conducting features, whereas type 2 (DB1-3, DB1-7, KB-1~KB-3, KB-7, KB-11, KB-14, KB-15) was affected by minor fracture network as well as rainfall event. Type 3 (DB1-1, DB1-2) was mainly influenced by surface infiltration of rainfall event. Type 4 (DB1-8, KB-9) was reflected by the irregular variation of groundwater level caused by anisotropy and heterogeneity of crystalline rock.

• 한국환경과학회지
• /
• 제17권2호
• /
• pp.257-267
• /
• 2008

Groundwater recharge characteristics in a fractured granite area, Mt. Geumjeong, Korea. was interpreted using bedrock groundwater and wet-land water data. Time series analysis using autocorreclation, cross-correlation and spectral density was conducted for characterizing water level variation and recharge rate in low water and high water seasons. Autocorrelation analysis using water levels resulted in short delay time with weak linearity and memory. Cross-correlation function from cross-correlation analysis was lower in the low water season than the high water season for the bedrock groundwater. The result of water level decline analysis identified groundwater recharge rate of about 11% in the study area.

• 대한토목학회논문집
• /
• 제33권6호
• /
• pp.2285-2291
• /
• 2013

본 연구에서는 Kim et al. (2013)이 개발한 부정류상태에서의 지하수위 변동 해석모델을 이용하여 단위시간별 함양량을 산정하였다. 지하수위 변동 해석모델은 지하수의 흐름과 관련된 반응계수와 함양에 관련된 비산출율을 매개변수로 이용한다. 해석모델의 적용을 위해 제주도 한천지역의 관측정 중 강우에 대한 반응이 양호한 JW 공항과 JD 용담1지점에 대한 5년(2006-2010)간의 지하수위 관측자료를 이용하였다. 지하수위 해석모델은 국지적인 규모에서의 단위시간별(일별) 함양량 산정이 가능하며 지하수위 자료와 비교한 결과, 함양량과 매개변수가 지하수위를 잘 반영하고 있었다. 함양량 산정결과 JD 용담1 지점은 520.4~904.0mm (32.7~61.8%), JW 공항지점은 447.4~633.4mm (24.0~45.1%)로 나타났다. 개발된 해석모델은 지하수위 자료를 이용하여 매개변수를 추정과 단위시간별 함양량을 산정하는 방안으로 활용되기를 기대된다.

• 한국수자원학회논문집
• /
• 제53권4호
• /
• pp.273-283
• /
• 2020

지하수자원의 변동성 및 취약성 평가를 위한 지하수위의 정량적 예측은 매우 중요하다. 이를 위해 다양한 시계열 분석 기법과 머신러닝 기법 등이 사용되어 왔다. 본 연구에서는 제주도 한경면 지역에 설치된 11개 지하수위 관측정의 일 수위자료를 대상으로 인공신경망 알고리즘의 하나인 Long short term memory (LSTM)에 기반한 예측 모델을 개발하였다. 제주도의 지하수위는 일반적으로 조석에 의한 자기상관성이 높고 강수에 의한 영향이 잘 반영되는 것으로 알려져 있다. 이러한 자료 특성을 고려한 입출력 텐서를 구성하기 위해 각 지하수 관측정의 수위변동 관측 자료와 같은 기간의 강수량 자료를 추가 입력자료로 선택하였다. 4계절을 나타내는 초기 365일 자료를 이용하여 LSTM 모델을 학습시켰으며 나머지 자료를 검증에 활용하여 예측 모델의 적합도를 평가하였다. 모델의 개발은 Python기반 딥러닝 프레임워크인 Keras를 이용하였고, 학습속도를 향상시키고자 NVIDIA CUDA 아키텍처를 도입하였다. LSTM 모델을 이용하여 지하수위 변화를 학습시키고 검증한 결과 결정계수가 평균 0.98로 나타나 개발된 예측모델의 적합성이 매우 높은 것으로 확인되었다.

• 지질공학
• /
• 제18권2호
• /
• pp.167-176
• /
• 2008

본 연구에서는 제주도 지역의 257개 공에서 갈수기(1998년 12월) 풍수기(1999년 7월)에 관측된 지하수위 자료에 의해 해안선에서 이격거리에 따른 지하수위 변동특성이 규명되었다. 갈수기 지하수위는 $7.46{\sim}203.88\;m$의 범위에 평균은 60.49 m, 풍수기에는 $4.01{\sim}204.10\;m$의 범위에 평균 57.66 m 이었다. 제주도에서는 1998년 $6{\sim}10$월에 내린 강우에 의한 영향으로 갈수기의 지하수위가 풍수기에 비해 높았다. 본 연구에 이용된 지하수공의 표고와 지하수위 사이의 상관계수는 0.86 이상이었으며, 갈수기와 풍수기 지하수위 간의 상관계수는 0.95 이상으로 매우 높았다. 본 연구에 이용된 257개 지하수공 자료를 해안선에서 500 m 간격으로 재정리한 후, 17개구간의 이격거리, 표고, 갈수기 및 풍수기 지하수위에 대한 평균을 산정하였다. 해안선에서 $0{\sim}4\;km$ 구간의 표고와 지하수위는 이격거리에 따라 증가하는 경향을 보였으나, $4{\sim}9\;km$ 구간에서는 증가와 감소가 반복되면서 불규칙적이었다. 제주도 지역의 해안선에서 이격거리에 대한 지하수위의 선형함수가 추정되었으며, 전체 구간($0{\sim}9\;km$)보다 $0{\sim}4\;km$ 구간 자료를 이용한 경우의 결정계수가 높게 나타났다. 그리고, 제주도 지역의 해안선에서 이격거리가 $0{\sim}4\;km$ 구간이 전체 구간의 지하수위 증가율에 비해 2배 이상 높았다. 이는 해안선에서 4 km 정도까지는 지하수위가 선형적으로 증가하지만, $4{\sim}9\;km$ 지역들에서는 증가와 감소가 반복되면서 증가하기 때문이다.