Log in

Using artificial neural network models for groundwater level forecasting and assessment of the relative impacts of influencing factors

Utilisation des modèles de réseaux neuronaux artificiels pour la prévision du niveau des eaux souterraines et l’estimation des impacts relatifs des facteurs influents

Uso de modelos de redes neuronales artificiales para el pronóstico del nivel del agua subterránea y evaluación de los impactos relativos de los factores influyentes

利用人工神经网络模型预测地下水位和评价影响因素的相关影响

Utilizando redes neurais artificiais para previsão de níveis de águas subterrâneas e avaliação dos impactos relativos dos fatores influenciadores

  • Paper
  • Published:
Hydrogeology Journal Aims and scope Submit manuscript

Abstract

Change in groundwater level is predicted for a special site where transient natural factors affecting the groundwater level are mixed with very irregular anthropogenic influences. When there is not enough hydrogeological information about the area to be analyzed, an artificial neural network (ANN) is a powerful tool for groundwater level forecasting in highly irregular and uncertain groundwater systems. In this study, groundwater levels were predicted by using ANN models with input variables composed of one natural factor and two anthropogenic factors in Yangpyeong riverside area, South Korea. Complex and irregular change of the groundwater level was monitored due to the operation of a groundwater heat pump system and winter intensive pum** for water curtain cultivation (by which greenhouses are warmed). The prediction results showed good performance with root mean square errors of 3–6 cm when the average groundwater level is about 25.59 m, the correlation coefficient is >0.9 and the Nash–Sutcliffe efficiency is >0.75, indicating that the ANN models are well suited for assessing complex groundwater systems. Along with the prediction, an extraction method was devised to calculate contributions and relative impacts of the input variables in the time-series-based ANN models. As a result, it was proved that the river level dominantly affects the groundwater level fluctuation, and the contributions of each influencing factor were obtained reliably according to spatial distribution and temporal variance. This makes the scheme effective for managing and using groundwater resources with consideration of every crucial influencing factor of the groundwater level fluctuation.

Résumé

La modification du niveau des eaux souterraines est prédit dans un site déterminé, où des facteurs naturels transitoires affectant le niveau des eaux souterraines sont intriqués avec des influences anthropiques très irrégulières. Quand il n’y a pas assez d’informations hydrogéologiques sur la zone à analyser, un réseau neuronal artificiel (RNA) est un outil puissant de prévision du niveau des eaux souterraines dans les systèmes hydrogéologiques hautement irréguliers et incertains. Dans la présente étude, le niveau des eaux souterraines a été prédit en utilisant des modèles RNA avec des données variables composées d’un facteur naturel et de deux facteurs anthropiques dans la zone riveraine du Yangpyeong, Corée du Sud. Le changement complexe et irrégulier du niveau des eaux souterraines a été contrôlé grâce à l’exploitation d’un système de pompe à chaleur sur eau souterraine et d’un pompage intensif en hiver pour une culture sous rideau d’eau (par lequel sont chauffées les serres). Les résultats de la prédiction se sont montrés performants avec une racine de l’erreur quadratique moyenne de 3–6 cm quand le niveau moyen des eaux souterraines est d’environ 25.59 m, le coefficient de corrélation est > 0.9 et le coefficient de performance de Nash–Sutcliffe est > 0.75, ce qui montre que les modèles RNA sont bien adaptés à l’évaluation des systèmes hydrogéologiques. En accompagnement de la prédiction, une méthode d’extraction a été conçue pour calculer les contributions et les impacts relatifs des données variables dans les séries chronologiques basées sur les modèles RNA. En conséquence, il a été prouvé que c’est le niveau de la rivière qui affecte principalement les fluctuations du niveau des eaux souterraines et les contributions de chaque facteur impactant ont été obtenues de manière fiable conformément à la distribution spatiale et à la variance temporelle. Ceci rend le système efficace pour gérer et utiliser les ressources souterraines avec prise en compte de chaque facteur décisif pesant sur la fluctuation du niveau des eaux souterraines.

Resumen

Se predice el cambio en el nivel del agua subterránea para un sitio especial donde los factores naturales transitorios que lo afectan se mezclan con influencias antropogénicas muy irregulares. Cuando no hay suficiente información hidrogeológica sobre el área a analizar, una red neuronal artificial (ANN) es una herramienta poderosa para el pronóstico del nivel del agua subterránea en sistemas de agua subterránea altamente irregulares e inciertos. En este estudio, los niveles de agua subterránea se predijeron utilizando modelos ANN con variables de entrada compuestas por un factor natural y dos factores antropogénicos en la zona ribereña de Yangpyeong, Corea del Sur. Se supervisó el cambio complejo e irregular del nivel del agua subterránea debido a la operación de un sistema de bomba de calor de agua subterránea y al bombeo intensivo de invierno para el cultivo de cortinas de agua (mediante el cual se calientan los invernaderos). Los resultados de predicción mostraron un buen rendimiento con errores cuadráticos medios de 3–6 cm cuando el nivel medio de agua subterránea es de aproximadamente 25.59 m, el coeficiente de correlación es > 0.9 y la eficacia de Nash–Sutcliffe es > 0.75, lo que indica que los modelos ANN son adecuados para evaluar sistemas complejos de aguas subterráneas. Junto con la predicción, se diseñó un método de extracción para calcular las contribuciones y los impactos relativos de las variables de entrada en los modelos ANN basados ​​en series de tiempo. Como resultado, se demostró que el nivel del río afecta predominantemente la fluctuación del nivel del agua subterránea, y las contribuciones de cada factor influyente se obtuvieron de manera confiable según la distribución espacial y la varianza temporal. Esto hace que el esquema sea efectivo para administrar y utilizar los recursos de agua subterránea con la consideración de cada factor de influencia crucial de la fluctuación del nivel del agua subterránea.

摘要

预测了一个特殊场地的地下水位变化情况,该场地影响地下水位的瞬时自然因素与非常不规则的人为因素混合在一起。当研究区没有足够的水文地质信息供分析时,人工神经网络就是预测高度不规则和不确定地下水系统的地下水位的一个**有力工具。在本研究中,利用人工神经网络模型,输入包括一个神经因素和两个人为因素的变量,预测了**国Yangpyeong河边地区地下水位。监测了由于运行地下水热泵系统和冬季**烈抽水用于水幕种植(温室被加温)造成的地下水位复杂和不规则的变化。预测结果显示了很好的性能,当**均地下水位为25.59 m时,均方根误差为3–6 cm,相关系数为>0.9, Nash–Sutcliffe效率 > 0.75,表明人工神经网络模型非常适合评价复杂的地下水系统。除了预测,还开发了萃取法用来计算基于时序的人工神经网络模型中输入变量的贡献率和相对影响。结果证明,河流水位主要影响地下水位波动,根据空间分布和时间变异,可靠地获取了每个影响因素的贡献率。这使管理和利用地下水资源中考虑地下水位波动的每个关键影响因素的计划非常有效。

Resumo

Mudanças nos níveis das águas subterrâneas são preditas para um local especial onde fatores naturais transientes afetando o nível das águas subterrâneas são misturados com influências antrópicas muito irregulares. Quando não existe informação hidrogeológica suficiente sobre a área a ser analisada, uma rede neural artificial (ANN) é uma poderosa ferramenta para predição de níveis de águas subterrâneas em sistemas de águas subterrâneas altamente irregulares e incertos. Neste estudo, os níveis das águas subterrâneas foram previstos usando modelos de ANN com variáveis ​​de entrada compostas de um fator natural e dois fatores antropogênicos na área ribeirinha de Yangpyeong, Coréia do Sul. Mudanças complexas e irregulares do nível do lençol freático foram monitoradas devido à operação de um sistema de bombas de calor subterrâneas e ao bombeamento intensivo de inverno para o cultivo de cortina de água (pelo qual as estufas são aquecidas). Os resultados de predição mostraram bom desempenho com erros quadráticos médios de 3–6 cm quando o nível médio de água subterrânea é de cerca de 25.59 m, o coeficiente de correlação é > 0.9 e a eficiência de Nash–Sutcliffe é > 0.75, indicando que os modelos ANN são adequados para avaliar sistemas complexos de águas subterrâneas. Juntamente com a previsão, foi desenvolvido um método de extração para calcular contribuições e impactos relativos das variáveis ​​de entrada nos modelos de ANN baseados em séries temporais. Como resultado, ficou comprovado que o nível do rio afeta predominantemente a flutuação do nível do lençol freático, e as contribuições de cada fator de influência foram obtidas de maneira confiável de acordo com a distribuição espacial e a variância temporal. Isso torna o esquema efetivo para o gerenciamento e uso dos recursos hídricos subterrâneos levando em consideração todos os fatores importantes que influenciam a flutuação do nível do lençol freático.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save

Springer+ Basic
EUR 32.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or Ebook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price includes VAT (Germany)

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

References

  • Alley WM, Reilly TE, Franke OL (1999) Sustainability of ground-water resources. US Geological Survey Circular 1186, pp 79. https://pubs.usgs.gov/circ/circ1186/. Accessed June 2017

  • Anctil F, Michel C, Perrin C, Andreassian V (2004) A soil moisture index as an auxiliary ANN input for stream flow forecasting. J Hydrol 286(1–4):155–167

    Article  Google Scholar 

  • ASCE Task Committee on Application of Artificial Neural Networks in Hydrology (2000a) Artificial neural networks in hydrology I: preliminary concepts. J Hydrol Eng 5:115–123

    Article  Google Scholar 

  • ASCE Task Committee on Application of Artificial Neural Networks in Hydrology (2000b) Artificial neural networks in hydrology II: hydrologic applications. J Hydrol Eng 5:124–137

    Article  Google Scholar 

  • Barzegar R, Fijani E, Moghaddam AA, Tziritis E (2017) Forecasting of groundwater level fluctuations using ensemble hybrid multi-wavelet neural network-based models. Sci Total Environ 599:20–31

    Article  Google Scholar 

  • Box GEP, Jenkins GM (1976) Time series analysis: forecasting and control. Holden-Day, San Francisco

  • Brunner P, Simmons CT (2012) HydroGeoSphere: a fully integrated, physically based hydrological model. Groundwater 50(2):170–176

    Article  Google Scholar 

  • Dimopoulos Y, Bourret P, Lek S (1995) Use of some sensitivity criteria for choosing networks with good generalization ability. Neural Process Lett 2(6):1–4

    Article  Google Scholar 

  • Daliakopoulos IN, Coulibaly P, Tsanis IK (2005) Groundwater level forecasting using artificial neural networks. J Hydrol 309(1–4):229–240

    Article  Google Scholar 

  • Ebrahimi H, Rajaee T (2017) Simulation of groundwater level variations using wavelet combined with neural network, linear regression and support vector machine. Glob Planet Change 148:181–191

    Article  Google Scholar 

  • Eriksson E (1970) Groundwater time series. Hydrol Res 1(3):181–205

    Article  Google Scholar 

  • Garson GD (1991) Interpreting neural network connection weights. AI Expert 6(4):46–51

    Google Scholar 

  • Gevrey M, Dimopoulos I, Lek S (2003) Review and comparison of methods to study the contribution of variables in artificial neural network models. Ecol Model 160(3):249–264

    Article  Google Scholar 

  • Gleeson T, Alley WM, Allen DM, Sophocleous MA, Zhou Y, Taniguchi M, VanderSteen J (2012) Towards sustainable groundwater use: setting long-term goals, backcasting, and managing adaptively. Groundwater 50(1):19–26

    Article  Google Scholar 

  • Hoque MA, Hoque MM, Ahmed KM (2007) Declining groundwater level and aquifer dewatering in Dhaka metropolitan area, Bangladesh: causes and quantification. Hydrogeol J 15(8):1523–1534

    Article  Google Scholar 

  • HRFCO (2017) The Han River flood control office. http://www.hrfco.go.kr Cited 15 June 2017

  • Hornik K, Stinchcombe M, White H (1989) Multilayer feedforward networks are universal approximators. Neural Netw 2(5):359–366

    Article  Google Scholar 

  • Hsu K, Gupta HV, Sorooshian S (1995) Artificial neural network modeling of the rainfall-runoff process. Water Resour Res 31(10):2517–2530

    Article  Google Scholar 

  • Jan CD, Chen TH, Lo WC (2007) Effect of rainfall intensity and distribution on groundwater level fluctuations. J Hydrol 332(3):348–360

    Article  Google Scholar 

  • Jha MK, Sahoo S (2015) Efficacy of neural network and genetic algorithm techniques in simulating spatio-temporal fluctuations of groundwater. Hydrol Process 29(5):671–691

    Article  Google Scholar 

  • Khan UT, Valeo C (2016) Dissolved oxygen prediction using a possibility theory based fuzzy neural network. Hydrol Earth Syst Sci 20(6):2267–2293

    Article  Google Scholar 

  • Khan UT, He J, Valeo C (2018) River flood prediction using fuzzy neural networks: an investigation on automated network architecture. Water Sci Technol 2017(1):238–247

    Article  Google Scholar 

  • Konikow LF, Kendy E (2005) Groundwater depletion: a global problem. Hydrogeol J 13(1):317–320

    Article  Google Scholar 

  • Lek S, Belaud A, Dimopoulos I, Lauga J, Moreau J (1995) Improved estimation, using neural networks, of the food consumption of fish populations. Mar Freshw Res 46(8):1229–1236

    Article  Google Scholar 

  • Maier HR, Dandy GC (1996) The use of artificial neural networks for the prediction of water quality parameters. Water Resour Res 32(4):1013–1022

    Article  Google Scholar 

  • Mohanty S, Jha MK, Kumar A, Sudheer KP (2010) Artificial neural network modeling for groundwater level forecasting in a river island of eastern India. Water Resour Manag 24(9):1845–1865

    Article  Google Scholar 

  • Nayak PC, Rao YRS, Sudheer KP (2006) Groundwater level forecasting in a shallow aquifer using artificial neural network approach. Water Resour Manag 20(1):77–90

    Article  Google Scholar 

  • Nourani V, Alami MT, Vousoughi FD (2015) Wavelet-entropy data pre-processing approach for ANN-based groundwater level modeling. J Hydrol 524:255–269

    Article  Google Scholar 

  • Quilty J, Adamowski J, Khalil B, Rathinasamy M (2016) Bootstrap rank ordered conditional mutual information (broCMI): a nonlinear input variable selection method for water resources modeling. Water Resour Res 52(3):2299–2326

    Article  Google Scholar 

  • Rumelhart DE, McClelland JL, The PDP Research Group (1986) Parallel distributed processing: explorations in the microstructure of cognition. MIT Press, Cambridge, MA

    Google Scholar 

  • Sahoo S, Russo TA, Elliott J, Foster I (2017) Machine learning algorithms for modeling groundwater level changes in agricultural regions of the US. Water Resour Res 53(5):3878–3895

    Article  Google Scholar 

  • Sung AH (1998) Ranking importance of input parameters of neural networks. Expert Syst Appl 15(3–4):405–411

    Article  Google Scholar 

  • Taormina R, Chau K, Sethi R (2012) Artificial neural network simulation of hourly groundwater levels in a coastal aquifer system of the Venice lagoon. Eng Appl Artif Intell 25(8):1670–1676

    Article  Google Scholar 

  • Trefry MG, Muffels C (2007) FEFLOW: a finite-element ground water flow and transport modeling tool. Groundwater 45(5):525–528

    Article  Google Scholar 

  • Wang S, Shao J, Song X, Zhang Y, Huo Z, Zhou X (2008) Application of MODFLOW and geographic information system to groundwater flow simulation in North China Plain, China. Environ Geol 55(7):1449–1462

    Article  Google Scholar 

  • Winter TC (1999) Relation of streams, lakes, and wetlands to groundwater flow systems. Hydrogeol J 7(1):28–45

    Article  Google Scholar 

  • Yang ZP, Lu WX, Long YQ, Li P (2009) Application and comparison of two prediction models for groundwater levels: a case study in western Jilin Province. China J Arid Environ 73(4–5):487–492

    Article  Google Scholar 

  • Yao J, Teng N, Poh HL, Tan CL (1998) Forecasting and analysis of marketing data using neural networks. J Inf Sci Eng 14(4):843–862

    Google Scholar 

  • Yi M, Kim G, Sohn Y, Lee J, Lee K (2004) Time series analysis of groundwater level data obtained from national groundwater monitoring stations. J Geol Soc Korea 40:305–329

    Google Scholar 

  • Yoon H, Jun SC, Hyun Y, Bae GO, Lee KK (2011) A comparative study of artificial neural networks and support vector machines for predicting groundwater levels in a coastal aquifer. J Hydrol 396(1–2):128–138

    Article  Google Scholar 

  • Yoon H, Hyun Y, Ha K, Lee KK, Kim GB (2016) A method to improve the stability and accuracy of ANN- and SVM-based time series models for long-term groundwater level predictions. Comput Geosci 90:144–155

    Article  Google Scholar 

Download references

Acknowledgements

This work was supported by the International Collaborative Energy Technology R&D Program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) (No. 20168510050070). This study was also supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2017R1A2B3002119).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Kang-Kun Lee.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lee, S., Lee, KK. & Yoon, H. Using artificial neural network models for groundwater level forecasting and assessment of the relative impacts of influencing factors. Hydrogeol J 27, 567–579 (2019). https://doi.org/10.1007/s10040-018-1866-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10040-018-1866-3

Keywords

Navigation