설악산국립공원 주변 지하수관측망의 수위변동 해석

Groundwater (GW) level prediction is important for effective GW resource management. It is hypothesized that using precipitation data in GW level modelling will increase the overall accuracy of the results and that the distance of the observation well to the weather station (where precipitation data are obtained) will affect the model outcome. Here, genetic programming (GP) was used to predict GW level fluctuation in multiple observation wells under three scenarios to test these hypotheses. In Scenario 1, GW level and precipitation data were used as input data. Scenarios 2 only had GW level data as inputs to the model, and in Scenarios 3, only precipitation data were used as inputs. Long-term GW level time series data covering a period of 8 years were used to train and test the GP model. Further, to examine the effect of data from previous time periods on the accuracy of GW level prediction, 12 models with input data up to 12 months prior to the current period were investigated. Model performance was evaluated using two criteria, coefficient of determination (R2) and root mean square error (RMSE). Results show that when predicting GW levels through GP, using GW level and precipitation data together (Scenario 1) produces results with higher accuracy compared to only using GW level (Scenario 2) or precipitation data (Scenario 3). Additionally, it was found that model accuracy was highest for the well located closest to the weather station (where precipitation data were collected), demonstrating the importance of weather station location in GW level prediction. It was also found that using data from up to six previous time periods (months) can be the most efficient combination of input data for accurate predictions. The findings from this study are useful for increasing the prediction accuracy of GW level variations in unconfined aquifers for sustainable GW resource management.

Groundwater is the primary source for drinking, irrigation, and industrialpurposes. Groundwater provides about 24 per cent of global water supply.Groundwater level fluctuates depending on the groundwater recharge anddischarge from the aquifer. The climate change, changing rainfall pattern andgrowing water demand lead to the groundwater level variation year by year.Rainfall is the major source for groundwater recharge. As rainfall increases,groundwater level also increases. Theni District also called Cardamom citylocated in the foot of Western Ghats is chosen as the study area. The majorsource of irrigation was well. For analysis of groundwater level status, theAverage - Ground Water Level data and Categorization of the firkas in Thenidistrict was collected from State Ground and Surface Water Resources DataCenter. Arithmetic averages of Rainfall of stations under the Theni districtwas collected from India Meteorological Department. Groundwater levelrise and fall analysis of Average-Groundwater level data (2011 to 2019)was carried out using the line chart. Comparative analysis was done usingRainfall data Vs Average – Groundwater level data (2014-2018). Consideringthe last five years, the groundwater level in Theni District has increased. Thestatistical correlation between the rainfall and groundwater level fluctuationwas poor. Categorization of firkas was used as a performance indicator.It was observed that the Safe Firkas increased from 18% to 41% and theCritical Firkas increased from 6% to 12%. In Periyakulam and Theni Talukthere is no change in Firkas category and remains the same as on 2011.This helps in focusing groundwater development in those taluks which arein the fringe of groundwater status.

Groundwater is an important water resource that is widely used worldwide for agricultural, industrial, and domestic purposes. In the case of Jeju Island, located in southern South Korea, groundwater is an indispensable water resource that accounts for 82% of the total water supply. Therefore, scientific prediction and management of groundwater levels are very important for the sustainable use of groundwater by citizens. This study additionally used precipitation data from the Baekrokdam Climate Change Observatory located on the summit of Jeju Island in artificial intelligence (AI) models to accurately predict one-month-ahead future groundwater levels for the mid-mountainous areas of Jeju Island, where groundwater levels are highly variable. In other words, the AI models compared and analyzed the improvement effect of the monthly groundwater level prediction performance for 1) using precipitation data from two rainfall stations, groundwater withdrawal data from two groundwater sources, and groundwater level data from two monitoring wells in the study area, and 2) adding precipitation data from Baekrokdam Climate Change Observatory. The study subjects are two groundwater level monitoring wells located at 435-471m above mean sea level in the southeast of Jeju Island. The AI models used to predict groundwater levels are Artificial Neural Network (ANN) and Long Short-Term Memory (LSTM), a deep learning AI model.As a result, when the Baekrokdam precipitation data were not used, the two AI models showed excellent groundwater level prediction performance with Nash-Sutcliffe efficiency (NSE) values of 0.871 or higher. The LSTM model showed relatively higher prediction performance for high and low groundwater levels than the ANN model. This means that the LSTM model adequately incorporates the seasonal effects of wet and dry periods into groundwater level simulations. The more volatile the observed groundwater level, the more difficult it is for the AI models to interpret the characteristics of groundwater level fluctuations, and the lower the performance of predicting future groundwater levels. When additional Baekrokdam precipitation data were used, the two AI models showed improved groundwater level prediction performance by having NSE values of 0.907 or higher. This means that the additional use of precipitation data located in the uppermost region provides more information to help interpret groundwater levels, allowing AI models to better interpret the characteristics of groundwater level fluctuations. In addition, the use of Baekrokdam precipitation data was more helpful in improving groundwater level prediction for the monitoring well, which has highly variable groundwater levels that are difficult to predict, and the ANN model with relatively low groundwater level prediction performance. When additional Baekrokdam precipitation data was used for a specific monitoring well, the groundwater level prediction performance of the ANN model was improved to a level comparable to that of the LSTM model, which is a deep learning AI, even with a relatively simple ANN model structure. This is an example of how important it is to use additional useful data in research using AI models.

<p>Groundwater represents an important source of drinking water for 25% of the population in the province of Quebec (Canada) and for 80% of its rural population. The deployment of the Quebec Groundwater Observation Network (Réseau de suivi des eaux souterraines du Québec – RSESQ) since the start of the millennia provides important data on the dynamics of piezometric heads throughout southern Quebec. This study aims to use the wealth of available groundwater data available to better understand the resilience of groundwater resources to changes in meteorological and hydrological conditions. The study area is located between the St. Lawrence River and the Canada-USA border, and between the Quebec-Ontario border and Quebec City (36,000 km²). Available data consist of groundwater level time series from 81 observation wells (2000-2018; 43 in confined aquifers, 15 in semi-confined aquifers and 23 in unconfined aquifers), total flow rates from 179 hydrometric stations (1960-2017), and meteorological data from a spatially interpolated 10 km x 10 km grid (1960-2017). Statistical analyses (Mann Kendall and Sen’s slope) were used to understand if groundwater levels and flow rates are declining or rising, what is their short-, medium- and long-term memory and what are the geomorphological, land use, and climate controls of this reactivity. The results show that groundwater levels since 2007 exhibit statistically significant negative annual trends for most observation wells. Since 1960, river flow rates, total precipitation and air temperature all show significant increases. Trends calculated on five-year sliding windows confirm that groundwater levels and river flow rates are significantly correlated to the climate indices Southern Oscillation index (SOI), NINO-3 and Pacific Decadal Oscillation index (PDO). Autocorrelations of flow rates and groundwater level data indicate that rivers and aquifers have a short hydrological memory rarely extending beyond the hydrological year. Cross-correlations of flow rates and groundwater levels with temperature show high correlation coefficients with a lag of up to 60 days, indicating a season-long effect of temperature changes. As expected, cross-correlation analysis of the two data sets with precipitation shows smaller correlation coefficients and a shorter reaction time (10 days). Standard deviations of daily groundwater levels are significantly higher in shallower wells and in wells where groundwater levels are closer to the ground. This confirms the presence of highly dynamic shallow aquifers reacting rapidly to surface processes.  Analyses are under way to test if spatially distributed parameters (e.g., geological setting, slope, land use) and well-related parameters (e.g.: depth, confined or unconfined) are explaining factors of trends and variations in groundwater levels and flow rates. One key observation from this study is that the RSESQ is highly valuable to understand groundwater dynamics and should be maintained on a long-term horizon. This detailed analysis has allowed to identify external influences (e.g., pumping) on some observation wells that do not reflect natural conditions and could be removed from the observation network. Recommendations also include the need for new observation wells in specific locations to improve the representativity of groundwater flow conditions in the study area.</p>

Groundwater plays a crucial role in meeting the demands of domestic, agricultural, and industrial sectors worldwide. The growing reliance on groundwater resources has led to excessive withdrawal that exceeds natural recharge for an extended period due to population growth, urban development, and other environmental factors. This study analyzed groundwater level data from Mymensingh, Bangladesh, to determine the spatiotemporal pattern, the rate of groundwater level depletion, and the state of predicted groundwater level by 2050. Weekly groundwater level data from seven monitoring wells were collected from the Bangladesh Water Development Board (BWDB) from 2002 to 2023. Spatial interpolation using inverse distance weighting (IDW) and an ARIMA time series model was employed for the analysis of long-term water level forecasting. The results showed groundwater level fluctuations and depletion at different monitoring wells of the study area. Of the 7 monitoring wells, GT6152021 experienced the highest depletion (0.24 m/y), and GT6152020 was the lowest (0.09 m/y); both were in Mymensingh Sadar. The depletion rates of other wells fluctuated between 0.15 and 0.19 m/y. The groundwater level data from all monitoring wells revealed a declining trend, indicating that the groundwater resources were used indiscriminately in the studied region. The fitted ARIMA (0, 1, 0) forecasting model for well GT6152021 observed the groundwater level to be at 25 m by 2050, and the other wells be between 17.5 and 19.5 m. These results will help planners and policymakers allocate groundwater resources among agricultural, domestic, and industrial uses.

Modeling spatial groundwater level patterns of Bangladesh using physio-climatic variables and machine learning algorithms

Abstract. Uncertainties in hydrological simulations can be quantified and reduced through data assimilation (DA). This study explores strategies for assimilating soil moisture (SM) data from Cosmic-Ray Neutron Sensors (CRNS) and groundwater level (GWL) data into the Terrestrial System Modeling Platform (TSMP), which integrates both land surface and subsurface processes. DA experiments incorporating both state and parameter estimation were performed using the localized Ensemble Kalman Filter (LEnKF) within a representative catchment in Germany over the period 2016 to 2018, with cross-validation conducted on non-overlapping years. Univariate assimilation of SM reduced the unbiased root mean square error (ubRMSE) by approximately 50 %, while univariate assimilation of GWL achieved up to a 70 % reduction in ubRMSE at assimilation sites. Improvements in GWL estimates extended up to 5 km from the assimilation points, with ubRMSE reductions ranging between 2 % and 50 %. However, assimilating GWL independently had a negative effect on SM representation, and similarly, assimilating SM alone degraded GWL predictions. To address these issues, a novel multivariate DA framework was developed, enabling SM and GWL to be assimilated independently through separate modules. Groundwater data were used to constrain the water table position, thereby improving the estimation of the boundary between unsaturated and saturated zones and allowing updates to hydraulic conditions within the saturated zone. Meanwhile, SM data improved the representation of hydrological processes in the unsaturated zone. The multivariate assimilation approach resulted in comparable improvements in GWL, SM, and evapotranspiration (ET) at the assimilation sites. Moreover, including parameter estimation alongside state updating further reduced the ubRMSE by up to 17 %.

Groundwater is a critical lifeline for sustaining water resources in Upper Godavari Sub Basin, India's arid regions. However, due to impeding water requirement and demand in these regions along with anthropogenic complexities has raised serious concerns for this vital resource. Along with anthropogenic activities, Climate change also threatens this precious resource due to scarcity of surface water mostly during the summer season of the year. Hence, to comprehend this important issue, the groundwater resource assessment needs to be done for present as well as future scenarios. Therefore, the present study assesses the groundwater resource using a SWAT-MODFLOW model which is a combination of advanced hydrological model with cutting-edge numerical groundwater model. Individual surface and groundwater models are developed in SWAT and MODFLOW respectively and then are linked using the linkages files to get the more enhanced surface and groundwater interaction in the form of recharge, groundwater level and interaction of rivers with sub surface. The surface and groundwater models are calibrated and validated using the streamflow and groundwater level data. The calibrated model thus presents the current scenario of groundwater allocation which is then simulated with different bias corrected climate variables for getting the status of groundwater for future SSPs scenarios. From a range of CMIP6 climate models, the best model is selected based on the statistical index such as NSE, the correlation coefficient, R2, MAE, RMSE, MSE, and NRMSE which was NESM3 in the present case with a highest correlation and R2 with IMD precipitation and temperature dataset. The best selected climate model (NESM3) is then bias corrected using the empirical quantile method. Along with the numerical approach, to map the groundwater level data, soft computing approach using RFR and GBR is also employed to predict the groundwater level data for future scenarios. The optimization of these models was done by the Particle PSO. The study findings in Upper Godavari Sub Basin, India, revealed significant changes in groundwater levels across different seasons, with particularly significant increases observed during the dry season. The study showed that MODFLOW-GBR-PSO is more accurate in predicting groundwater level than MODFLOW-GBR, MODFLOW-RFR-PSO and MODFLOW-RFR. The result also predicted decreased rainfall for the SSP 585 scenario which in turn lead to drop in groundwater level and recharge in the distinct parts of the sub basin. Hence, from the above result a proper mitigation and framework needs to be prepared to counterfort the diminishing groundwater resource for the betterment of environment.Key words: climate change, hydrological model, SWAT, Nash-Sutcliffe efficiency (NSE), Root mean square error (RMSE), Random Forest regression (RFR) and Gradient Boosting Regression (GBR), Swarm Optimization method (PSO).

Climate change poses uncertainties to the supply and management of water resources under the observed increase in surface temperatures all over Africa. The aim of this study is to assess the impact of temperature changes on groundwater levels in Nzoia River Basin. Temperature and groundwater level variability and trends has been analyzed using the parametric test of Linear regression and the non-parametric Mann-Kendall statistical test. Temperature data was obtained from the Kenya meteorological department (KMD) whereas groundwater level data was collected from Water resources management agency (WRMA). Linear regression of the annual groundwater levels in Nzoia River Basin between 2011 and 2017 revealed a decreasing trend ranging from -0.49 ft/year (Kitale Golf Club) to -0.03 ft/year (Kakamega Tande School). Mann-Kendall statistical test also showed decreasing groundwater levels for all observation wells with the results for Kitale Golf Club and Mois Bridge Quarry observation wells being statistically significant, whereas those for Kapsabet Boys High School, Kakamega Mwikalikha School, Kakamega Tande School and Busia Town Prison were statistically insignificant at 5% significance level. The highest decline in groundwater levels was observed in the upper catchment of the basin.There are significant increases in annual tempratures for Kitale and Kakamega stations in the period 1979 - 2014. Kitale showed annual maximum temprature rising at 0.0006260C/year; annual minimum temperature rising at 0.0011630C/year and the annual mean temprature rising at 0.0008940C/year. Kakamega had annual maximum temprature rising at 0.0007710C/year; annual minimum tempratures rising at 0.0004710C/year and the annual mean tempratures rising at 0.0006230C/year. Eldoret showed falling maximum temprature at - 0.002020C/year; rising minimum temperature at 0.0008130C/year and falling mean temperatures at - 0.001420C/year. The results for Kitale and Eldoret stations showed statistically significant trends whereas those for Kakamega station had a statistically insignificant trend. In Nzoia River Basin, Kitale and Eldoret, annual minimum tempratures are rising faster than the maximum whereas in Kakamega it’s the annual maximum tempratures that are rising faster than the minimum. Kitale and Kakamega stations showed rising annual mean temperatures whereas Eldoret showed falling annual mean tempratures. As one would expect, temperatures in Nzoia River Basin are expected to be rising; however, the case of falling temperatures recorded at Eldoret international airport might occur because this region of Rift valley has highly protected natural resources and a high forest cover is available all the year round. Another possible explanation to this could be the changing cloudness around Eldoret station. Kitale and Kakamega showed annual mean tempratures rising at about 0.10C per century and Eldoret showed mean temperatures falling at about -1.40C per century. The findings for Kitale and Kakamega stations compare well with IPCC Third Assessment Report estimated global warming rate of 0.60C during the twentieth century and other studies from the African continent and East African region.The decreasing trend in groundwater levels in the basin appears to be linked to climate change. Increases in temperature have an impact on the hydrologic cycle because they enhance evaporation of accessible surface water and vegetation transpiration. As a result, these changes have an impact on precipitation volumes, timings, and intensity rates, as well as indirect effects on water flux and storage in surface and subsurface reservoirs. While changes in important long-term climatic factors such as air temperature, precipitation, and evapotranspiration directly affect surface water supplies, the interaction between changing climate variables and groundwater is more intricate and little understood. For efficient and long-term groundwater resource management, understanding long-term temperature variability and trends, as well as the corresponding reaction of groundwater levels, is critical. Despite the fact that groundwater level records are only available for a short period of time, they include essential information that may be utilized to establish strategies for managing the basin's limited groundwater resources.

Predicting groundwater level is necessary for environmental protection. Thus, our research suggests a hybrid artificial intelligence model that integrates the particle swarm optimization (PSO) algorithm, the coati optimization (COO) algorithm, the gated recurrent unit (GRU), and the adaptive neuro-fuzzy inference system (ANFIS) to address these challenges. The new model, called PSO-COO-GRU-ANFIS (PCGA), operates in several steps. First, the model uses PSO-COO to set ANFIS and GRU parameters. Then, the model uses GRU to extract hidden patterns from the data. In the final step, the extracted patterns are input into ANFIS to generate predictions. Specifically, the proposed model is used to precisely forecast monthly groundwater levels (GWLs) in the Ardabil Plain, Iran. The PCGA is also compared with several benchmark models, and its accuracy is evaluated using multiple evaluation metrics. Our findings show that the PCGA model achieves a mean absolute error (MAE) of 1.90 and a Nash-Sutcliffe efficiency (NSE) of 0.90. The PCGA model enhances the MAE and NSE of all other prediction models by 14-77% and 1-20%, respectively. The results also highlight the effectiveness of the PSO-COO algorithm and the GRU model in improving forecast precision, as the hybrid optimization approach significantly reduced error fluctuations during parameter tuning, and the GRU demonstrated strong capability in capturing long-term temporal dependencies in groundwater level data. Overall, the findings of the current study show that the PCGA model is a robust tool for forecasting monthly groundwater levels. The PCGA model demonstrates superior predictive capability compared to conventional and standalone models. By combining the strengths of optimization algorithms, deep learning, and fuzzy inference systems, the model effectively captures nonlinear and dynamic relationships in groundwater level (GWL) data.

Quantitative estimates on the magnitude and dynamics of vadose and groundwater zones in subarctic conditions are rare. Furthermore, knowledge on the constraints of different data on numerical groundwater recharge estimates (i.e., data worth analyses) is limited. We built a process-based three-dimensional hydrological model to describe and analyze the dynamics and magnitude of groundwater recharge, vadose zone, and related key hydrological components in subarctic conditions. Three-year time series of daily groundwater level and soil moisture (five different depths) data were used to identify plausible model realizations. Thereafter, long-term simulations over 6 years were conducted. The model adequately corresponded with the magnitude and dynamics of the observations, even though the simulations and observations occasionally differed. Simulations with several different plausible parameterizations showed how soil moisture and groundwater level fluctuations, as well as recharge estimates, are sensitive to water retention parameter variations, whereas precipitation and evapotranspiration controlled the recharge magnitude. Recharge showed clear interannual variability (annual mean 0.28–0.60 m). The fraction of annual mean recharge over precipitation was 39%–66% during our simulation period. We showed how groundwater level data can set more constraints on the recharge estimates than soil moisture data, and how both of the data types together constrained the recharge estimates the most. We showed how the vadose zone had modest dynamics and storage properties (typically 2%–8% of the total soil water) in terms of the bulk soil water storage. However, the average amount of the air-filled pore space (0.70 m) and water (0.38 m) in the vadose zone were proportional to annual precipitation and evapotranspiration, respectively. Thus, the vadose zone can form an important boundary between the soil surface and groundwater zone, and its role is not highly sensitive to hydrometeorological fluctuations. Data worth analyses provide a potential opportunity to move toward increasingly reliable model-based groundwater recharge estimates and system understanding.

Groundwater is an important water resource, and groundwater level (GWL) forecasting is a useful tool for supporting the sustainable management of water resources. Existing studies have shown that GWLs can be accurately predicted by combining an artificial neural network model with meteorological and hydrological factors. However, GWL data are typically geographic spatiotemporal series data, and current studies have considered only the spatial distance factor when predicting GWLs. In karst aquifers, the GWL is affected by the developmental degree of the karst, topographic factors, structural features, and other factors; considering only the spatial distance is not enough, and the real spatial connectivity characteristics need to be considered. Thus, in this paper, we proposed a new method for forecasting GWLs in karst aquifers while considering connectivity characteristics using a neural network prediction model. The connectivity of a karst aquifer was analyzed by a multidimensional feature clustering method based on the distance index and hydrogeological characteristics recorded at observation wells, and a convolutional long short-term memory (ConvLSTM) conjunction model was constructed. The proposed approach was validated through GWL simulations and predictions in karst aquifers in Jinan, China, and four experiments were conducted for comparison. The experimental results show that the proposed method provided the most consistent results with the measured observation well data among the analyzed methods. These findings demonstrate that the proposed method, which considers connectivity characteristics in karst aquifers, has a higher simulation accuracy than other methods. This method is therefore effective and provides a new idea for the real-time prediction of the GWLs of karst aquifers.

Analysis of groundwater level records from Ireland’s South Eastern River Basin District (SERBD) allowed fundamental information about the nature of bedrock and gravel aquifers to be investigated. The hydrogeological setting of a monitoring point (with respect to, for example, recharge area, discharge area or proximity to a river) is the dominant factor influencing hydrograph character in bedrock aquifers, with aquifer type and subsoil properties producing secondary effects. Analysis of seasonal groundwater levels showed that the fractured bedrock aquifers recharge more quickly and typically have a longer recession period than gravel aquifers. The calculated recession periods for bedrock aquifers are longer than previous estimates for similar aquifers. Hydrograph analysis identified a number of notable phenomena including a gravel aquifer’s interaction with surface water and evidence of rejected recharge. Short-term groundwater level fluctuations caused by global seismic events, recorded via chart recorders, are discussed. Specific yield values were calculated, for a number of settings, from annual average groundwater level variations. The values supported estimates from previous research on similar aquifers. An analysis to investigate if any impacts of climate change were evident showed no consistent change in the timing of groundwater level minima or maxima.

Specific yield (Sy) is defined as the ratio of the volume of water that saturated rock or soil yields by gravity to the total volume of the rock or soil. Sy is often taken as a constant that when multiplied to groundwater level change provides water volume change, hence it is crucial in validating Gravity Recovery And Climate Experiment (GRACE)-derived groundwater changes, providing estimates of available groundwater resources, and in modelling groundwater aquifers. In this study, we used GRACE data and available quality-controlled in-situ well data to estimate Sy instead. Our hope was that it would match the available Sy, however we observed a time-varying Sy. Upon further investigation we found a negative correlation between water level and Sy. Hence the time-evolution of Sy was due to changes in the water level. We processed available well data and GRACE(-FO) in India, the United States, Europe, and Australia, spanning from January 2004 to December 2022. We also developed a general law/empirical relationship between Sy and groundwater level. All regional-specific empirical relationships exhibit a decrease in Sy as the average groundwater level depth drops, but the decay rate of Sy is notably faster in India (0.17 ± 0.04 m-1) compared to the United States (0.03 ± 0.01 m-1), Australia (0.06 ± 0.02 m-1), and European countries (0.04 ± 0.03 m-1). This empirical expression allows for the estimation of Sy based on readily available groundwater level data, thus supporting large-scale groundwater assessments and modelling efforts. At the global scale, a 50% decrease in Sy results in a groundwater level depletion of ~17 meters. However, in India, due to a faster decay rate, the same 50% reduction in Sy causes a groundwater level depletion of ~4 meters. This relationship can be utilized by hydrologists, water resource managers, and policymakers to predict Sy and assess changes in groundwater levels over time, aiding in more effective and sustainable water resource management.

As a significant portion of the available water resources in volcanic terrains such as Jeju Island are dependent on groundwater, reliable groundwater level forecasting is one of the important tasks for efficient water resource management. This study aims to propose deep-learning-based methods for groundwater level forecasting that can be utilized in actual management works and to assess their applicability. The study suggests practical forecasting methodologies through the Gyorae area of Jeju Island, where the groundwater level is highly volatile and unpredictable. To this end, the groundwater level data of the JH Gyorae-1 point and a total of 12 kinds of daily hydro-meteorological data from 2012 to 2021 were collected. Subsequently, five factors (i.e., mean wind speed, sun hours, evaporation, minimum temperature, and daily precipitation) were selected as hydro-meteorological data for groundwater level forecasting through cross-wavelet analysis between the collected hydro-meteorological data and groundwater level data. The study simulated the groundwater level of the JH Gyorae-1 point using the long short-term memory (LSTM) model, a representative deep-learning technique, with the selected data to show that the methodology is adequately applicable. In addition, for its better utilization in actual practice, the study suggests and analyzes (i) a derivatives-based groundwater level learning model which is defined as derivatives-based learning to forecast derivatives (gradients) of the groundwater level, not the target groundwater time series itself, and (ⅱ) an ensemble forecasting methodology in which groundwater level forecasting is performed repetitively with short time intervals.