Search

국내 설치 운영 중인 지하수 관측정은 계측 센서의 고장 또는 전원 장치의 이상 등으로 결측이 발생하기도 한다. 지하수위 자료를 활용한 수문 평가, 지표수와 지하수의 연계 분석 및 지하수 함양량 산정 등을 위하여 결측 자료의 처리는 필수적이다. 본 연구에서는 가평가평 국가지하수 관측정의 1개월의 지하수위 결측 기간을 채우기 위하여 다변량 ARIMA와 인공신경망 모델인 MLP 및 LSTM을 적용하였으며, RMSE 및 MAE를 활용하여 모델의 정확도를 비교 분석하였다. 자기회귀 특성의 지하수위와 인근의 강우 자료가 입력변수로 사용되는 다변량 ARIMA 및 LSTM 모델이 MLP 모델에 비하여 정확한 것으로 평가되었다. LSTM 모델의 경우, 입력 변수로 사용되는 주변 강우 및 지하수 관측정의 수를 증가하더라도 오차가 거의 줄어들지 않는 것으로 나타났다. LSTM 모델은 훈련 및 검증 자료에 의하여 결정되는 만큼, 지하수위 결측 처리의 일관성을 위해서 입력 변수의 선정 방법에 대한 추가 연구가 수행될 필요가 있다.

Time series of natural groundwater level considering rainfall effects are usually used for the estimation of groundwater recharge, long-term trend analysis, and assessment of interactions between surface water and groundwater along rivers. However, anthropogenic activities, such as groundwater pumping, land excavation, and barrier construction, may induce abnormal changes in water levels. This study aimed to develop a universal long short-term memory (LSTM) model for predicting natural water level variations in a time-series record that has been affected by groundwater abstraction or other anthropogenic factors. This model uses past and present groundwater levels, rainfall and representative principal components of groundwater level time series as input variables. For this purpose, 17 cases of the developed LSTM model were tested using 13 monitoring wells, of which the case with the highest prediction performance was selected. Among the test cases, case 6 was found to achieve the highest performance, with average RMSE and MAE values of 0.061 and 0.027, respectively. The case 6 model used rainfall, groundwater level of monitoring wells, and four main principal components (1–4) as input variables. Also, its optimum window size was found to be 5. The accuracy of the LSTM model was found to be more strongly affected by window size than by input variables. Although the case 6 LSTM model may have errors for some monitoring wells, it has high potential as a universal model that can consistently determine natural groundwater levels in South Korea. This LSTM model makes it universally available for water level prediction, even with long missing periods in groundwater level time series or in the absence of adjacent observations for model input.

Artificial recharge is increasingly necessary to ensure sustainable groundwater supplies and meet the needs of a growing population and crop cultivation in the age of climate change. In this study, artificial groundwater recharging using surplus water resources within the local area is considered for a water shortage area during a drought season, Shingok village, Hongsung-gun, South Korea, that lacks water sources. Based on an analysis of the monthly supply and demand with the water budget method, the current water supply is determined to be entirely dependent on groundwater wells, and water shortages occur in May (approximately 6446 m3) when rice planting begins. The surplus groundwater, which is the amount of natural recharge minus the present usage, and stream discharge are sufficient from December to June, so they are considered source waters for artificial recharge. A depth of 6 ~ 19 m is suggested as the main target region for recharge based on the composition of strata and the vertical distribution of the hydraulic conductivity, which was investigated with electric resistivity survey, drilling, and pumping test methods. Considering the permeability of layers, a vertical well, by which water can infiltrate directly into the target strata, is proposed as the main injection structure, and a ditch is also suggested as an auxiliary system to improve the effect of injection over a broad area. In the future, a detailed design for the injection scenario, injection facility, well and ditch arrangements, and pumping facility will be established through numerical modeling, field injection tests, etc. If it is impossible to obtain water from a long distance, an artificial recharge system based on local water sources may be an appropriate method of securing water considering the storage capacity of permeable strata.

Because the mixture of seawater and freshwater in the Gyeongin-Ara Waterway in South Korea can lead to the intrusion of saline water into surrounding aquifers, systematic management through the establishment of a groundwater protection area is required. The analytic hierarchy process (AHP) model is used to delineate this protection area based on two primary factors and five secondary factors related to saline water intrusion. The study area is divided into 987 gridded cells with a unit size of 100 × 100 m, and the final evaluation score for each cell is calculated using the AHP model. Consequently, several artificial neural network models based on a multilayer perceptron are developed using the AHP’s secondary criteria and the evaluation score. Comparing the evaluation scores of ANN and AHP, more than 180 samples are required in the ANN model to insure high R2 between the original and estimated values. The ANN model is more consistent than the AHP model when determining groundwater protection area, because it can be re-constructed due to the changes in some secondary criteria and also changed due to a standardization process. The final evaluation score by the ANN model based on 300 samples, with the highest R2, is calculated and the regions with a score higher than 2.0 are selected as the groundwater protection area, accounting for 15% of the total cells. This area is similar to the range within approximately 200 m of the GA Waterway and also includes some changing sites in hydrogeochemistry and electric conductivity, which is produced by saline water intrusion. If the land-use type, groundwater levels, and some other criteria change at any cell, the ANN model can be re-executed to verify whether the cell belongs to a groundwater protection area. Considering that salinity of groundwater near the waterway can be affected by various factors including well depth, pumping conditions, and groundwater levels, the ANN model, which is a non-linear model, can be more effective for prediction than the AHP model.

본 연구에서는 기계학습 기법인 인공신경망(ANN)과 지지벡터기계(SVM) 기반 지하수위 시계열 예측모델을 개발하고 이를 국가지하수관측망 216개소 지하수위 자료에 적용하였다. 시계열 예측 모델 적용의 목적은 결측 및 이상 자료를 보정하고 지하수위 자료에 포함된 양수 및 하천수위 변동 영향을 제거하여 강우에 의한 지하수위 변동 모의에의 적용성을 평가하는 것이다. 먼저 각 관측소에 대해 1일 직접예측 모델을 구성한 뒤 이를 이용한 반복예측 모델을 구성하고 각 모델의 예측 및 모의 능력을 검토하였다. 시계열 예측 모델 적용 결과 결측, 이상 자료, 양수 및 하천수위 영향을 효과적으로 보정하고 강우에 의한 지하수의 변동 패턴을 모의하는 것으로 나타났다. 오차지표 분석 결과 ANN 모델이 SVM 모델보다 다소 높은 직접예측능력을 보여주지만 반복예측모델에 대한 안정성은 SVM이 높은 것을 알 수 있다. 또한 본 연구에서는 이용된 연구자료에 대한 ANN, SVM 기반 시계열 모델 적용 시 적절한 모델 파라미터 값의 범위를 제시하였다. 본 연구에서 적용한 방법론과 결과는 지하수관측망 수위 자료 이상값을 탐지하는데 이용되어 지하수관측망을 효과적으로 관리하는데 활용될 수 있다. 또한 강우에 의한 지하수위 변동을 모의함으로써 지하수위 변동법을 적용하여 지하수 함양률을 산정하는 등 지하수 자원을 관리하는데 활용될 수 있을 것으로 기대한다.

This is conducted to observe underground temperature and to analyze its change affected by climate condition and soil infiltration in the mountainous area, Yesan region, Chungcheong-namdo province. Additionally, underground temperature change is also simulated using air temperature and soil thermal properties with a numerical model. Soil temperature monitoring data acquired from each depth, 20 cm, 50 cm, and 100 cm, indicates that the data within 50 cm in depth shows peak-shaped big fluctuation directly affected by air temperature and it at 100 cm has open-shaped small fluctuation. Underground temperature variation, a difference between high and low values, during monitoring period is weakly proportional to hydraulic conductivity of the sediment and it is assumed that water plays a part in delivering air temperature in soil. The underground temperature estimated by a numerical model is very similar to the observed data with an average value of 0.99 cross-correlation coefficient. From the result of this study, the aquifer unsaturated hydraulic conductivity of the soil and the groundwater recharge is likely to be able to estimate with underground temperature profile calculated using a numerical model.

The aim of the Four Major Rivers Restoration Project (4MRRP) was to mitigate flood damage by controlling surface water levels via the construction of large barrages along the main rivers of South Korea. The increase in surface water levels behind such barrages can lead to a rise in groundwater levels upstream, which may increase the likelihood of flooding after heavy rainfall or during non-pumping periods. A total of 900 groundwater monitoring sites, comprising 446 automatic and 454 manual stations, are proposed on the low-lying plains alongside the rivers based upon the following criteria: groundwater flow directions, topography (determined using GIS analysis), the desire for equal spacing of monitoring wells, economic efficiency, field accessibility, and cost. A numerical model (MODFLOW) is used to estimate future groundwater flow directions and the area likely to be affected by a rise in surface water levels after barrage construction. Since 2011, 137 automatic monitoring wells have been constructed, and recent time series data show that groundwater levels are closely linked to surface water levels, and increased rapidly for the first 1–2 years after barrage construction. Groundwater monitoring wells will be essential for assessing the depth to groundwater, forecasting the extent of any potential flooding, and developing countermeasures such as drainage system.

/d의 지하수가 수직구내로 유입되는 것으로 예상되었으며, 실제 그라우팅을 거친 후 굴착 3과정에서 배출된 지하수량을 측정한 결과 30~100 m/d의 범위를 보여 예측 수량과 유사한 것으로 나타났다. 수직구의 작업 안정성 확보 및 원활한 공정을 위해서는 수치모델을 활용한 그라우팅 효과의 사전 평가가 필요하며 이를 근거로 그라우팅 시공을 선행한 후 굴착이 진행되는 것이 필요하다.주요어 : 선택취수설비 , 굴착, 그라우팅 , 지하수 유입량, 수치모델The vertical shaft of a selective intake structure, which is constructed in a large reservoir, is required to be imper-meable and to employ a grouting technology to prevent water inflow from the reservoir or surrounding ground. Inthis study, groundwater inflow is estimated using a numerical model for two cases (i.e., grouting or non-groutingcases at the exterior of a vertical shaft) and compared with data measured during an excavation at the constructionsite of a selective intake structure in the Soyang reservoir, Korea. Groundwater inflow is estimated to range from444 to 754 m

A geological barrier such as a mountain ridge surrounding a large reservoir generally plays some part in preventing water leakage from the reservoir, but water flows out naturally according to the hydraulic gradient because the reservoir water level is typically higher than that outside. The hydrogeological, geochemical, and isotopic properties of water in the Daechung Dam area were analysed to characterize the source and the phenomenon of the groundwater flowing through the geological barrier. Our geochemical analysis showed that the concentrations of anions and cations tended to increase, due to the longer residence time of the water, as the distance between the sampling site and reservoir increased. Elevation differences (calculated using oxygen and hydrogen isotopes) between the highest and lowest sampling sites ranged between 350 and 420 m based on the altitude effect on isotope concentration, but the actual difference was only 35 m. These results, including analysis of the rate of water seepage from the reservoir (ca. 2.3% to 14.4% of natural recharge), indicate the possibility that groundwater in the study area originated from reservoir water that was derived from upstream water and rainfall in the high mountainous area of the upper watershed. Copyright © 2013 John Wiley & Sons, Ltd.

and Mn become too high for water treatment. A total of ninewells, including four shallow wells with a depth of 3 m and five deeper wells with a depth of 6 m, were installed inthe study area at Naeseong-cheon in Yecheon-gun. The change in hydrogeochemical features of groundwater andthe concentrations of Fe

Changes in groundwater level have been recognized by the earthquakes at various epicentral distances. The M9 Sumatra earthquake resulted in changes in the groundwater level, electrical conductivity, and temperature in monitoring wells on Jeju Island, South Korea. In regions of different groundwater type (basal, lower parabasal, upper parabasal, and high-level groundwater), the changes in the groundwater levels at 25 monitoring wells ranged between 4.0 and 49.5 cm; changes in the electrical conductivity at six monitoring wells ranged between 1 and 27,975 μS/cm; and the changes in water temperature at three wells ranged between 0.02 and 1.37 °C. The irregular groundwater level changes at different locations on the island due to the earthquake reflect various interactions between hydrological properties and seismological processes. The impact of the earthquake was successfully recognized via transfer function modeling between the time series of groundwater level and the tidal oscillation. On the basis of the theoretical aquifer response to the earthquake, storage coefficient estimates for aquifers, which could not be determined from the single-well pumping tests, were determined to be within the range of 1.22·10−4-3.51·10−6.

Though the geologic barrier surrounding a dam reservoir plays an important role in preventing water seepage to the outside of a reservoir, a detailed survey to assess bedrock conditions of the whole reservoir boundary is seldom performed when designing a dam. In this study, a variety of investigations and analyses have been conducted to reveal a seepage phenomenon through a reservoir boundary composed of highly fractured metamorphic rocks at the Daechung multipurpose dam. At the study site, groundwater levels at monitoring wells located close to a reservoir tend to change proportionally to reservoir water levels, usually with a rapid response time of 1 day or less. Soil moisture content also changes with respect to the reservoir water level, and the area where seepage from the reservoir is a possibility has a relatively high soil moisture content. The similarity of measured seepage rates at the outflow site and the seepage rate estimated by Darcy’s law suggests strong connectivity and high density of the fractures. The estimated seepage rate using a numerical model for the northernmost valley (Zone 1) is approximately 127.2 m3/day in the case of high reservoir water level and 24.3 m3/day in the case of low water level. Continuous monitoring to obtain time series data for water levels, quality, and bedrock displacement is recommended to assess the sustainable stability of this geologic barrier.

We performed numerical modeling to estimate the groundwater level around a riverside area following the construction of an artificial structure. The groundwater level of the alluvial deposit responded more rapidly to the river water level than to the rainfall event itself, indicating that the groundwater and river water are directly interrelated through the riverbed. Furthermore, transient modeling showed raised groundwater levels at the southern part of Mt. Dok and the eastern part of Mt. Dummit in an area of low plains. The artificial structure caused a rise in groundwater level of up to approximately 6 m.

A statistical analysis of time series of water level at 27 groundwater monitoring wells was conducted to analyze the surface water-groundwater connectivity in the wide alluvial plains surrounding the Nakdong River, Korea. Change in groundwater level is strongly related to river water level, yielding an average cross-correlation coefficient of 0.601, which is much higher than that between rainfall and groundwater level (0.125). Principal component analysis of groundwater level indicates that wells in the study area can be classified into two groups: wells in Group A are located close to a river, have water levels closely related to river level, and generally show a large increase in groundwater level during heavy rainfall. On the other hand, wells in Group B located far from a river are relatively less related to river level. Including hydrologic and statistical analyses, geochemical analysis and temperature monitoring are additionally required to reveal the relationship between surface water level and groundwater level, and to assess the possibility of groundwater flooding.

K-water Institute, Korea Water Resources Corporation미고결대수층에서 우물 굴착 시 대수층 교란이 수리특성에 미치는 영향을 파악하기 위하여 우물굴착 이후의 우물개량공법인 서지블록과 에어서징을 실시하였다. 그리고 우물 굴착 및 개량의 대수층 변화를 비교·분석하기, 위하여 단계 및장기 양수시험을 수행하였다. 그 결과 우물 3굴착 및 개량 이후의 단계양수시험에서 동일 양수율 700 m/day에 대한 수위강하와 비교하였을 때 수위강하는 각각 26.01 m, 21.62 m로 4.39 m정도 감소되었다. 이때 우물특성을 파악하기 위한우물 굴착 및 개량 이후의 스킨계수는 7.92에서 5.04로 감소되어 우물상태가 개선된 것으로 나타났다. 우물 굴착 및 개량 이후의 장기양수시험에서 양수정을 중심으로 -3주변 MW-2, 3, 4호정의 투수량계수 범위는 각각 1.684×10

Soil deformations such as settlement, heaving and lateral flow have frequently happened on marine reclaimed soft grounds due to embankment filling or banking. The electrical resistivity survey was applied to investigate on ground surface such soil deformation without disturbance of ground. A test embankment was performed to assess soil deformation in marine reclaimed soft grounds, where was located at Sihwa area in western coast of Korean peninsula. The soft ground was composed of clayey sediments. After embankment filling, the boundary of soil deformation affected by the filling could be investigated with application of the electrical resistivity survey. The result of electric resistivity survey shows that the extent of deformation is about 5 m laterally to the southern direction of embankment and about 5~6 m vertically in depth, which is about 1-1.2 times of embankment height. This shows that the electric resistivity survey can be applied to interpret the ground deformation in a soft ground region.

We examined the interactions between surface and groundwater through (1) flowmeter logging, (2) measurements of seasonal and vertical changes in temperature within a well, and (3) geochemical analyses of water samples from nine groundwater-monitoring wells. At two wells adjacent to a stream, subsurface water was found to flow from the stream to a surrounding alluvial fan, and the seasonal change in groundwater temperature is similar to those of surface water and air. Geochemical analyses at two wells indicated hydro-geochemical features affected by streamwater inflow, showing seasonal variations. Accordingly, these two wells are located in an area with active interaction between surface water and groundwater. The Thermochron I-button used in the present study is useful for this type of study of groundwater?surface water interaction because of its low cost and small size.

Groundwater has played an important role in economic development in Southeast Asian countries, but some problems caused by nature or human actions such as contamination, over pumping, and land subsidence bring the necessity of more systematic groundwater monitoring wells. The analytical hierarchy process with pairwise comparison was used to allocate and organize the regional groundwater monitoring wells in five regions, Thailand, Cambodia, East/West Malaysia, and South Korea. Five different multi criteria decision models, which were composed of three primary criteria and eight secondary criteria, were developed based on the answers of the questionnaire from 76 groundwater experts in Thailand, 100 in Cambodia, 101 in East Malaysia, 87 in West Malaysia, and 93 in South Korea. It was revealed that the weights of model criteria for each country, which also represent relative importance on groundwater monitoring, were different according to the diverse groundwater situation. The most important factor to determine the number of monitoring well was ‘number of households using only groundwater as a water source’ for Thailand and South Korea, ‘number of contamination sources’ for Cambodia, ‘amount of groundwater use for drinking-water supply’ for East Malaysia, and ‘number of wells with contaminated water’ for West Malaysia.

A set of lineament maps has been constructed, comprising a lineament distribution map, a lineament length density map, a lineament counts density map and a lineament cross-point density map. The ArcView SHP file is used as input format. Two Avenue(TM) script files (Remove-node and Generalize) have been developed, in which input errors can be minimized, to optimize the original digitized lineament map data. For analyzing the lineament characteristics, five Avenue(TM) script files (L-STAT, Dist-to-Line, Dist-to-Cross-point, L-Selection and PL-DENS) calculating the lineament statistics (lineament length density, counts density, cross-point density, distance between a well and the nearest lineament, and distance between a well and the nearest lineament cross-point) have been developed with Avenue(TM) language using ArcView 3.1 GIS software. It is very easy to calculate the lineament-related values with the above five script files. They are also useful for drawing lineament density maps and for revising and compiling existing lineament density maps with ArcView 3.1 GIS software. Lineament alignment or lineament density distribution can play an important role in groundwater exploration or productivity analysis. These script files, together with well data, can facilitate analysis of the relationship between groundwater characteristics and lineament distribution.

Lineament density maps can be used for the quantitative evaluation of relationships between lineaments and the occurrence of groundwater. This paper reports the usefulness of the representative elementary area (REA) concept for lineament analysis. This concept refers to the area of the unit circle needed to calculate the lineament density factors distributed within the circle: length, counts and cross-points counts. The circle is a unit circle where one calculates the sum of the lineament length, lineament counts and the number of cross-points within it. The REA is needed to obtain the best representative lineament and ground-water well yield or other groundwater characteristics. A lineament map for the Yongsangang-Seomjingang watershed of Korea was used for demonstrating the concept. It is shown that the REA concept can be efficiently applied to lineament density analysis and mapping. In the demonstration case, the lineament densities are inversely proportional to the size of the REA and the REA can be calculated with this inversely linear regression model. If the average lineament density values for the whole study area are known, the most accurate density maps can be drawn using the REAs obtained from each linear regression model.

강우는 하천 수위의 상승 및 주변 충적 분지내 지하수위의 상승을 야기하여 토양수분 함량 증가를 가져올 수 있는데, 이 현상들은 지표 홍수의 규모에 영향을 미치게 된다. 본 연구에서는 지하수 관측정에서 측정된 지하수위 시계열 자료의 수위 상승 이벤트를 대상으로 수위 상승 기울기와 최고 지점에서의 지하수위 깊이를 산정하여 지하수 홍수강도를 계산하였으며, 고수위가 유지되는 기간과의 상관관계를 활용하여 홍수취약성 평가를 위한 지수를 고안하였다. 또한, 지수 산정 결과를 현장의 토양 습윤화 및 침수 현상과 비교해 본 결과 수변지역 홍수취약성 평가에 적용할 수 있을 것으로 나타났다. 향후 지하수위 미계측 지역에서의 적용성을 높이기 위해서 토양, 토지이용, 수문, 지형 등의 지역별 인자와의 상관관계로부터 홍수취약성 평가 모델을 개발할 필요가 있겠다.