Groundwater Potential Prediction Research Articles

Prediction and Verification of Groundwater Potential in Qingyang Area Based on Reliability Test, Validity Test and ROC Curve Method

The combination of remote sensing and GIS has become a common method to locate groundwater potential. Selecting reasonable and effective evaluation factors has become the most important thing in groundwater potential prediction. In this paper, reliability test and validity test are used to evaluate the rationality and validity of evaluation factors. The weight is determined by comprehensive weight method, and the groundwater potential map is obtained by visualization processing based on GIS.

Modeling of geophysical derived parameters for groundwater potential zonation using GIS-based multi-criteria conceptual model

This study modeled geophysical derived parameters and multi-critically synthesized their themes based on geospatial and analytical hierarchy processes (AHP) approaches for groundwater potentiality prediction mapping. These methodologies were investigated in a typical hard rock geologic terrain, southwestern, Nigeria. Considering the spatially acquired 96 vertical electrical sounding (VES) data in the area, geoelectric sections revealing five subsurface layers including the topsoil, laterite, weathered layer, fractured basement and fresh basement rock were produced mindful of the 2-D resistivity structure subsurface imaging data interpreted results. The correlative results of the 2-D resistivity structure images and VES data interpretation results delineated major low resistivity vertical discontinuity typical of fractured zones characterized with width range of 25–40 m, while the depth vary from about 40 to > 60 m. Themes of groundwater potential conditioning factors (GPCFs), namely: regolith, bedrock relief, hydraulic head, coefficient of anisotropy, aquifer resistivity and aquifer thickness were prepared from the re-analyzed hydrogeological and geophysical data. The produced themes were appropriately weighted in the context of AHP data mining technique. The groundwater potential prediction index (GPPI) mathematical modeling equation for the area was established via applying the weight linear average algorithm involving the AHP weightage results. The synthesized results of the applied GPPI model equation on the GPCFs’ hydrogeologic themes give GPPI values in the range 1.59–3.65 for the study area. The geospatial modeling of the GPPI estimated values result produced groundwater potential prediction index map for the area. The produced GPPI model map zoned the area into low (1.59–2.30), medium (2.30–2.61), medium–high (2.61–3.02) and high (3.02–3.65) groundwater potential classes. The area analysis of the GPPI map indicates that more than 70% of the study area has ‘low to medium groundwater potential. The GPPI map result verification using reacting operating characteristics technique results gave 86% and 81% success and prediction rates, respectively. The findings of this study are useful to water managers and decision-makers for locating appropriate positions of new productive wells in the study area and other areas with similar geologic settings.

Application of hybrid model-based machine learning for groundwater potential prediction in the north central of Vietnam

Application of hybrid model-based machine learning for groundwater potential prediction in the north central of Vietnam

Spatial predictions of groundwater potential using automated machine learning (AutoML): a comparative study of feature selection and training sample size in Qinghai Province, China.

Predicting groundwater potential is crucial for identifying the spatial distribution of groundwater in a region. It serves as an essential guide for the development, utilization, and protection of groundwater resources. Previous studies have primarily emphasized finding the most accurate prediction model for groundwater potential while giving less attention to the selection of training features and sample sizes. This study aims to predict groundwater potential within Qinghai Province using automated machine learning technology and assess the influence of sample sizes and feature selection on prediction accuracy. Sixteen groundwater conditioning factors were categorized into categorical and numerical variables. Four feature selection modes were utilized as input in training the model. The results indicated that, except for correlations between evaporation and landforms (- 0.8) and precipitation and normalized difference vegetation index (0.8), the Pearson correlation coefficients among the remaining sixteen factors were ≤ 0.5 or ≥ - 0.5. The models XGB-ALL, RF-Entropy, ET-CRITIC, and XGB-PCA yielded accuracy scores of 0.783, 0.685, 0.745, and 0.703, and area under curve (AUC) of 0.819, 0.724, 0.779, and 0.747, respectively. If enough samples are available with the tree model, an increased number of features can improve prediction accuracy. The principal component analysis method showed difficulty in reducing the dimensionality of the input space, while the Entropy method proved efficient. The accuracy and AUC value of the prediction model improved with an increasing number of samples. Training with 8 features and 200 data points achieved an accuracy of 0.745, sufficient to evaluate regional groundwater potential. As for training with 600 samples, the model's performance accuracy rose to 0.9, enabling precise groundwater potential prediction. The outputs of this research can provide decision-makers in groundwater resource management in Qinghai Province with crucial theoretical and practical support. The lessons learned can have future applications in similar situations.

Study on potential groundwater yield zone in sandstone aquifer based on a dual dynamic variable weight model: A case study in Shuangma Coal Mine of Ordos Basin

Accurate and effective assessment of groundwater potential in aquifers hundreds of meters deep remains a major challenge in mining and environmental protection. We analyzed the interaction mechanism of hydrogeological parameters, quantified it and incorporated it into the variable weight model (VWM) for secondary optimization, and constructed a dual dynamic variable weight model (PE-HVWM) that varies with multi-parameter entanglement and single-parameter state values to characterize the groundwater potential. Then, the pumping test data measured at the mining site were used to compare and analyze the results of different model predictions, and the optimized model was validated and tested using the Receiver Operating Curve (ROC) method and sensitivity analysis, which showed that the PE-HVWM method has a fairly high accuracy in groundwater potential prediction (AUC = 0.874). The PE-HVWM model can visualize the intrinsic correlation between hydrogeological parameters, and can respond sensitively to sudden changes in the state value of a hydrogeological parameter in the aquifer, and output groundwater potential results that are more in line with the actual situation after self-balancing by linking other parameters.

Assessment of groundwater potential zones for urban development site suitability analysis in Srivaikundam region, Thoothukudi district, South India

To demarcate groundwater potential zones (GWPZ) for the Srivaikundam region in Thoothukudi district, Tamil Nadu, India, an effort was made in this study to locate the subsurface basin for groundwater storage and to identify the suitable destination for the urban development, agricultural, and drinking needs. This study using remote sensing and geographical information system (GIS) with analytical hierarchy process (AHP). According to their respective importance for groundwater occurrence, Saaty's scale was appropriately weighted to give the thematic layers and their attributes. The selected thematic maps were then merged using the weighted linear combination method to produce the final groundwater potential zone map after the layers' assigned weights and their features had been standardized using the Analytic Hierarchy Process (AHP) and eigenvector approach. Based on Saaty's 9-point scale, each criterion or factor was given the proper weight, and the weights were then normalized using the analytic hierarchy approach (AHP). To create the groundwater potential prediction map for the research area, the method was incorporated into the GIS environment. For the purpose of identifying potential urban residential development sites, the generated GWPZ is combined with eight different thematic layers. Groundwater conditions have been used to define GIS analysis criteria, and each information layer has been assigned an appropriate rank and weightage. Finally, groundwater recharge zones have been selected and classified into low, medium, and high based on cumulative weighted value. The moderate groundwater potential zone covers the majority of the land, with areas of very low 0.5 km2 (0.1%), low 112.3 km2 (17.7%), medium 428.7 km2 (67.6%), high 87.8 km2 (13.8%), and very high 5.1 km2 (0.8%) accordingly. Furthermore, higher groundwater yields are estimated in medium to good zones, indicating greater potentiality of groundwater sources.

Comparison of Novel Hybrid and Benchmark Machine Learning Algorithms to Predict Groundwater Potentiality: Case of a Drought-Prone Region of Medjerda Basin, Northern Tunisia

Water scarcity is a severe problem in Tunisia, particularly in the northern region crossed by the Medjerda River, where groundwater is a conjoint water resource that is increasingly exploited. The aim of this study is to delineate the groundwater potential zones (GWPZs) in the Lower Valley of the Medjerda basin by using single benchmark machine learning models based on artificial neural network (ANN), random forest (RF), and support vector regression (SVR), and by developing a novel hybrid method, NB-RF-SVR, to reach the highest accuracy of groundwater potential prediction. Each model produced a spatial groundwater potential map (GPM) with the input of 26 groundwater-related factors (GRF) selected by the frequency ratio model and 70% of the transmissivity training data. The models’ effectiveness was assessed using the AUC-ROC curve, sensitivity, specificity, MAE, and RMSE metric indicators. The validation findings revealed that all the models performed successfully for the GWPZ mapping, where the AUC values for the ANN, RF, SVR, and NB-RF-SVR models were estimated as 71%, 79%, 87%, and 92%, respectively. The relative importance of the GWPZs revealed that land use followed by geology and elevation were the most important factors. Finally, these outcomes can provide valuable information for decision makers to effectively manage groundwater in water-stressed regions.

Evaluation of groundwater potential using Dempster-Shafer model and sensitivity analysis of effective factors: A case study of north Khuzestan province

In order to optimize the management of groundwater resources, accurate measurement of groundwater potential is necessary. Application of learning models and remote sensing is very encouraging in reducing the costs of groundwater potential measurement. In this research, Dempster-Shafer learning model and remote sensing data are used to evaluate the groundwater potential in the north of Khuzestan. The study area has a semi-arid climate and is mostly covered by quaternary sediments except for the north, which is mountainous and different geological formations are seen there. For this purpose, 9 parameter layers including: lithology, soil type, slope percent, land use, altitude, TWI index, lineament density, river density and distance from the river as the most effective parameters in groundwater potential are prepared. Additionally, 474 wells in high potential areas have been selected as suitable real data for training and testing of the model. In the training phase, the weights of the layers were calculated using Dempster-Shafer model based on real data of 70% of wells then the groundwater potential map was obtained by combining the weighted layers. To confirm the efficiency of the model and the accuracy of the groundwater potential map, the ROC curve is plotted on the basis of 30% of wells that were not used during the training phase. The results show that the final groundwater potential map is 86% accurate. After verifying the efficiency of the model, sensitivity analysis of effective parameters in predicting groundwater potential is performed using map removal sensitivity analysis method. The results show that all of the considered parameters have a positive effect on groundwater potential prediction accuracy. However, lithology, soil type, and slope percent are the most effective on the accuracy of the model.

Ensemble Boosting and Bagging Based Machine Learning Models for Groundwater Potential Prediction

Due to the rapidly increasing demand for groundwater, as one of the principal freshwater resources, there is an urge to advance novel prediction systems to more accurately estimate the groundwater potential for an informed groundwater resource management. Ensemble machine learning methods are generally reported to produce more accurate results. However, proposing the novel ensemble models along with running comparative studies for performance evaluation of these models would be equally essential to precisely identify the suitable methods. Thus, the current study is designed to provide knowledge on the performance of the four ensemble models i.e., Boosted generalized additive model (GamBoost), adaptive Boosting classification trees (AdaBoost), Bagged classification and regression trees (Bagged CART), and random forest (RF). To build the models, 339 groundwater resources’ locations and the spatial groundwater potential conditioning factors were used. Thereafter, the recursive feature elimination (RFE) method was applied to identify the key features. The RFE specified that the best number of features for groundwater potential modeling was 12 variables among 15 (with a mean Accuracy of about 0.84). The modeling results indicated that the Bagging models (i.e., RF and Bagged CART) had a higher performance than the Boosting models (i.e., AdaBoost and GamBoost). Overall, the RF model outperformed the other models (with accuracy = 0.86, Kappa = 0.67, Precision = 0.85, and Recall = 0.91). Also, the topographic position index’s predictive variables, valley depth, drainage density, elevation, and distance from stream had the highest contribution in the modeling process. Groundwater potential maps predicted in this study can help water resources managers and policymakers in the fields of watershed and aquifer management to preserve an optimal exploit from this important freshwater.

Groundwater potential evaluation using geoelectrical and analytical hierarchy process modeling techniques in Akure-Owode, southwestern Nigeria

The complexity and heterogeneity of hydrological conditions in Akure-Owode of southwestern Precambrian basement complex of Nigeria largely control the availability of groundwater supplies in the area. Akure-Owode, being one of the four hydrogeological provinces in Sub-Saharan Africa is worst hit by poor access to continuous supply of groundwater resource throughout the year. Analytical hierarchy process (AHP) modeling technique in conjunction with geo-electric measurements was employed to unravel subsurface complexities and groundwater potential zones in the study area. Geophysical-derived groundwater potential conditioning criteria, namely bedrock relief (B), resistivity (R) and aquifer thickness (T) were synthesized by AHP pairwise comparison weighting and ranking technique to generate groundwater potential indices (GPIs) of the area. Raster-weighted overlay analysis were performed on the GPIs to obtain groundwater potential model (GPM) of the study area. The produced GPM was binned into low, medium and high groundwater potential prediction zones. The performance evaluation of the groundwater potential model using the spatial locations of prolific boreholes and hand-dug wells in the area indicates 80% success rate. Emerging from the vertical electrical sounding curve analysis, weathered (unconfined) and weathered/fractured (unconfined) aquifer types were identified in the area. Hence, the developed groundwater potential map in this study can be harnessed for optimal groundwater resource management of Akure-Owode area.

Optimized Conditioning Factors Using Machine Learning Techniques for Groundwater Potential Mapping

Assessment of the most appropriate groundwater conditioning factors (GCFs) is essential when performing analyses for groundwater potential mapping. For this reason, in this work, we look at three statistical factor analysis methods—Variance Inflation Factor (VIF), Chi-Square Factor Optimization, and Gini Importance—to measure the significance of GCFs. From a total of 15 frequently used GCFs, 11 most effective ones (i.e., altitude, slope angle, plan curvature, profile curvature, topographic wetness index, distance from river, distance from fault, river density, fault density, land use, and lithology) were finally selected. In addition, 917 spring locations were identified and used to train and test three machine learning algorithms, namely Mixture Discriminant Analysis (MDA), Linear Discriminant Analysis (LDA) and Random Forest (RF). The resultant trained models were then applied for groundwater potential prediction and mapping in the Haraz basin of Mazandaran province, Iran. MDA has been successfully applied for soil erosion and landslide mapping, but has not yet been fully explored for groundwater potential mapping (GPM). Although other discriminant methods, such as LDA, exist, MDA is worth exploring due to its capability to model multivariate nonlinear relationships between variables; it also undertakes a mixture of unobserved subclasses with regularization of non-linear decision boundaries, which could potentially provide more accurate classification. For the validation, areas under Receiver Operating Characteristics (ROC) curves (AUC) were calculated for the three algorithms. RF performed better with AUC value of 84.4%, while MDA and LDA yielded 75.2% and 74.9%, respectively. Although MDA performance is lower than RF, the result is satisfactory, because it is within the acceptable standard of environmental modeling. The outcome of factor analysis and groundwater maps emphasizes on optimization of multicolinearity factors for faster spatial modeling and provides valuable information for government agencies and private sectors to effectively manage groundwater in the region.

An attribute recognition model to predict the groundwater potential of sandstone aquifers in coal mines

The groundwater potential prediction of sandstone aquifers is an important pre-requisite for the implementation of reasonable and effective measures to prevent mine water inrush disasters. In this study, an attribute recognition model was combined with entropy weighting to predict the groundwater potential of sandstone aquifers in coal mines. Five evaluation indices were selected to predict groundwater potential, such as sandstone thickness, sandstone lithology coefficient, flushing fluid consumption, fracture fractal dimension and fold fractal dimension. On the basis of data analysis, the groundwater potential was classified into four levels. Confidence and improved score criteria were applied to attribute recognition. The main advantages of this model are that it enables both the prediction and quantification of the groundwater potential of sandstone aquifers. The model’s results were compared with those from a comprehensive geographic information system evaluation. The final model results were in good agreement with the observed results, proving that this attribute recognition model is accurate and effective for groundwater potential prediction.

Groundwater potential prediction by using geoelectricity method a case study in Simpang Lima and around it

Water is an important natural resource, which is available both on surface as well as in recharge zone of weathered layer and in various other suitable water reservoir formations/structures below the surface. The study was conducted in Semarang City, Indonesian. Simpang Lima area were chosen since they have geology record of thick alluvium. The condition of Simpang Lima and around it which currently has developed into the main business area in Semarang City. This is marked by the presence like hotels, super market, that is more than one. This condition certainly requires a supply of water for various purposes that support economic efforts in the region. During this time the water needs are met by drilling wells and used as much as possible without taking into account the impact of water utilization. This study aims to determine the zones that have the potential existence of groundwater. The method used is Geoelectric method with Shlumberger configuration in 7 distributed spatial point. The results showed that there is a potential groundwater in eastern Pekunden, Erlanga, behind The Hotel Horizon, and Simpang Lima. Groundwater is found successively at an average depth of 15 meters under earth’s surface.

Delineation of groundwater potential zones using remote sensing (RS), geographical information system (GIS) and analytic hierarchy process (AHP) techniques: a case study in the Leylia–Keynow watershed, southwest of Iran

In this research, a standard methodology has been applied to delineate groundwater resource potential zonation based on integrated remote sensing (RS), geographic information system (GIS), and analytical hierarchy process (AHP) techniques in Leylia–Keynow watershed, southwest of Iran. A total of five sets of criteria/factors (including lineament density, rainfall, lithology, slope, and drainage density) believed to be influencing groundwater storage potential in the area were selected. Each criterion/factor was assigned appropriate weight based on Saaty’s 9-point scale and the weights were normalized through the analytic hierarchy process (AHP). The process was integrated in the GIS environment to produce the groundwater potential prediction map for the area. The fi9-p groundwater prospect map obtained was classified as excellent potential, very good potential, good potential, moderate potential, and poor potential zone. The obtained results indicated that only 21% (122 km2) of the study area exhibit poor groundwater potential, whereas most of the regions (326 km2) in the research showed good to excellent groundwater potential. Also, about 24% (141 km2) was classified as having moderate groundwater potential. The good to excellent potential zones are characterized by the higher lineament density, higher rainfall, and lithology type such as limestone, whereas the poor to moderate groundwater potential zones are characterized by the lesser lineament density, lower rainfall, lithology type of shale and marl as well as shale and limestone. Based on the obtained evidences, the tectonic structures had an important role in fracturing and crushing of limestone units in the area and so are vital for karst and ground water sources development. The demarcation of groundwater potential zones in the Leylia–Keynow watershed will be helpful for future planning, development and management of the groundwater resources.

Modeling of geoelectric parameters for assessing groundwater potentiality in a multifaceted geologic terrain, Ipinsa Southwest, Nigeria – A GIS-based GODT approach

Modeling of groundwater potentiality zones is a vital scheme for effective management of groundwater resources. This study developed a new multi-criteria decision making algorithm for groundwater potentiality modeling through modifying the standard GOD model. The developed model christened as GODT model was applied to assess groundwater potential in a multi-faceted crystalline geologic terrain, southwestern, Nigeria using the derived four unify groundwater potential conditioning factors namely: Groundwater hydraulic confinement (G), aquifer Overlying strata resistivity (O), Depth to water table (D) and Thickness of aquifer (T) from the interpreted geophysical data acquired in the area. With the developed model algorithm, the GIS-based produced G, O, D and T maps were synthesized to estimate groundwater potential index (GWPI) values for the area. The estimated GWPI values were processed in GIS environment to produce groundwater potential prediction index (GPPI) map which demarcate the area into four potential zones. The produced GODT model-based GPPI map was validated through application of both correlation technique and spatial attribute comparative scheme (SACS). The performance of the GODT model was compared with that of the standard analytic hierarchy process (AHP) model. The correlation technique results established 89% regression coefficients for the GODT modeling algorithm compared with 84% for the AHP model. On the other hand, the SACS validation results for the GODT and AHP models are 72.5% and 65%, respectively. The overall results indicate that both models have good capability for predicting groundwater potential zones with the GIS-based GODT model as a good alternative. The GPPI maps produced in this study can form part of decision making model for environmental planning and groundwater management in the area.

Application of a GIS-/remote sensing-based approach for predicting groundwater potential zones using a multi-criteria data mining methodology.

This study integrates the application of Dempster-Shafer-driven evidential belief function (DS-EBF) methodology with remote sensing and geographic information system techniques to analyze surface and subsurface data sets for the spatial prediction of groundwater potential in Perak Province, Malaysia. The study used additional data obtained from the records of the groundwater yield rate of approximately 28 bore well locations. The processed surface and subsurface data produced sets of groundwater potential conditioning factors (GPCFs) from which multiple surface hydrologic and subsurface hydrogeologic parameter thematic maps were generated. The bore well location inventories were partitioned randomly into a ratio of 70% (19 wells) for model training to 30% (9 wells) for model testing. Application results of the DS-EBF relationship model algorithms of the surface- and subsurface-based GPCF thematic maps and the bore well locations produced two groundwater potential prediction (GPP) maps based on surface hydrologic and subsurface hydrogeologic characteristics which established that more than 60% of the study area falling within the moderate-high groundwater potential zones and less than 35% falling within the low potential zones. The estimated uncertainty values within the range of 0 to 17% for the predicted potential zones were quantified using the uncertainty algorithm of the model. The validation results of the GPP maps using relative operating characteristic curve method yielded 80 and 68% success rates and 89 and 53% prediction rates for the subsurface hydrogeologic factor (SUHF)- and surface hydrologic factor (SHF)-based GPP maps, respectively. The study results revealed that the SUHF-based GPP map accurately delineated groundwater potential zones better than the SHF-based GPP map. However, significant information on the low degree of uncertainty of the predicted potential zones established the suitability of the two GPP maps for future development of groundwater resources in the area. The overall results proved the efficacy of the data mining model and the geospatial technology in groundwater potential mapping.

Appraising the accuracy of GIS-based Multi-criteria decision making technique for delineation of Groundwater potential zones

Increased demand of water in different sectors and restriction of water resources, necessitate the planning of groundwater recharge. In this study, groundwater potential zone are delineated by combining remote sensing, geographical information system (GIS) and multi-criteria decision making (MCDM) techniques in the Durg district, Chhattisgarh. Groundwater potential zones prepared using various thematic layers viz. geology, slope, land-use, lineament, drainage, soil, and rainfall. The thematic layers and their features were assigned suitable weights on the Saaty’s scale according to their relative significance for ground water occurrence. The assigned weights of the layers and their features were normalized by using Analytic Hierarchy Process (AHP) and eigenvector method finally; the selected thematic maps were integrated using weighted linear combination method to create the final ground water potential zone map. Each criterion/factor was assigned an appropriate weight based on Saaty’s 9 point scale and the weights were normalized through the analytic hierarchy process (AHP). The process was integrated in the GIS environment to produce the groundwater potential prediction map of the study area. The groundwater potential map of the Durg district was found to be 75 % and 56 % accurate for seven and four factors respectively. The ground water potential zone map was finally validated using the groundwater depth data from 16 pumping wells respectively in the study area.

Assessing the accuracy of GIS-based elementary multi criteria decision analysis as a spatial prediction tool – A case of predicting potential zones of sustainable groundwater resources

► We assess the effectiveness of MCDA as a predicting tool. ► We establish the dependence of the effectiveness of MCDA on coherence of criterial. ► Spatial prediction of groundwater potential zones was presented as a case study. ► We provide steps for proper data integration in spatial prediction of groundwater. ► The accuracy of the prediction was verified to be 81.25%. Inappropriate handling/integration of data from various sources is a problem that can make any spatial prediction tasking and inaccurate. Attempt was made in this study to offer solution to this problem by exploring the capability of GIS-based elementary MCDA as a spatial prediction tool. In order to achieve the set objectives, spatial prediction of potential zones of sustainable groundwater resources in a given study area was used as a case study. A total of five set of criteria/factors believed to be influencing groundwater storage potential in the area were selected. Each criterion/factor was assigned appropriate weight based on Saaty’s 9 point scale and the weights were normalized through the analytic hierarchy process (AHP). The process was integrated in the GIS environment to produce the groundwater potential prediction map for the area. The effect of coherence of criteria on the efficiency of MCDA as a prediction tool was also examined. The prediction map produced was found to be 81.25% accurate. The results of the examination of the effect of coherence of criteria revealed that the ability of the method to produce accurate prediction is dependent on the exhaustiveness of the set of criteria used. It was established in the study that the GIS-based elementary MCDA technique is capable of producing accurate and reliable prediction particularly if the set of criteria use for the prediction is coherent.