Predicting Groundwater Level Changes Using Deep Learning and Influencing Factors Through Remote Sensing

Heydarzadeh, Maryam; Siasar, Hadi

doi:10.22034/iwm.2024.2036252.1164

Predicting Groundwater Level Changes Using Deep Learning and Influencing Factors Through Remote Sensing

Document Type : Original Article

Authors

Maryam Heydarzadeh ¹

Hadi Siasar ²

¹ Department of Water Engineering, Faculty of Agriculture, Minab higher Education center, University of Hormozgan, Bandar-Abass, Hormozgan, Iran

² Department of Agriculture, Payame Noor University (PNU), Zabol, Sistan and Baluchestan, Iran

10.22034/iwm.2024.2036252.1164

Abstract

Extended Abstract
Introduction: Groundwater level (GWL) is of critical importance, especially in arid and semi-arid countries. In many areas, excessive exploitation of GWL has led to irreversible damage to groundwater resources. Predicting GWL is a key challenge in hydrogeological research, effective aquifer management, and assessing groundwater volumes. The aim of this research is to investigate and compare the efficiency of Deep Learning (DL), Decision Tree (DT), and Gradient Boosted Tree (GBoost) models in predicting the GWL of the Rudan aquifer.
Materials and methods: Monthly GWL data of the Rudan aquifer, along with precipitation, temperature, and evaporation data from the region's meteorological stations (2000-2020), were collected. The second part of the study involved satellite data accessed through the Google Earth Engine platform, where GWL data and key parameters- including the Palmer Drought Severity Index (PDSI), Standardized Precipitation Index (SPI), Potential Evapotranspiration (PET), precipitation (Pr), and vegetation indices such as NDVI, EVI, SAVI, NDWI, and GNDVI- were extracted and processed. Data normalization was performed to improve the performance of machine learning models, and the data were split into training (80%) and testing (20%) sets to evaluate model performance and prevent overfitting. To investigate the behavior and model the GWL of the Rudan aquifer, 10 parameters were used in 10 scenarios across three models: DL, DT, and GBoost.
Results and Discussion: In the DL model, increasing the number of parameters from the first to the third scenario decreased the model's accuracy. However, scenarios 5 to 9, which showed the highest correlation (0.86) and the lowest error (0.85) and percentage error (41%), were the most suitable for modeling GWL changes in the Rudan aquifer using cumulative precipitation, NDWI, PDSI, SAVI, NDVI, EVI, Pr, PET, and SPI variables. The DT model showed improved accuracy with an increasing number of parameters up to a certain point (from the first to the seventh scenario). The highest accuracy was achieved using a combination of cumulative precipitation, NDWI, PDSI, SAVI, NDVI, and EVI, with RMSE and MSE of 0.282 and 0.08, respectively, and a percentage error and correlation of 13.07% and 0.987, respectively. The GBoost model demonstrated relatively stable accuracy across all scenarios. Given the low error values and high correlation, the overall statistical criteria indicated that adding more parameters had a reduced sensitivity on the model's performance and did not significantly change its accuracy. Both the DL and DT models are more sensitive to input parameters. Additionally, this models exhibited similar responses to the input parameters in each scenario. Considering the four-month delay of precipitation on GWL, the DL model included precipitation, evapotranspiration, and drought indices in its selected scenarios. Therefore, it can be concluded that this model, considering a broader set of environmental parameters and examining their impact on modeling GWL changes, provides better efficiency, performance, and comprehensiveness for the Rudan aquifer.
Conclusion: The output results of the models showed that the DT and DL models, with high correlation values and lower error values, demonstrate highly accurate performance in predicting GWL. The scatter plots of predictions and actual values indicate a very close match between these two datasets, highlighting the high accuracy of both the DT and DL models. The accepted scenarios in both models include vegetation indices, indicating the significant impact of this parameter on groundwater resources, particularly in arid and semi-arid regions, where vegetation is a primary source of moisture. On the other hand, the DL model included meteorological drought indices in its selected scenarios, demonstrating the influence of these key factors on GWL changes in the region. Given the complexity of the DL model in selecting important parameters affecting GWL fluctuations, this model can be considered an efficient and suitable tool for investigating the factors influencing GWL changes in the Rudan aquifer. The use of these input parameters in the selected scenarios of both models can improve the accuracy and efficiency of groundwater resource management. Ground observational data for GWL further confirm the high importance of these parameters.

Keywords

Aquifer water level

Machine Learning

Deep Learning

Remote Sensing

Drought Indices

Subjects

Integrated Watershed Management

Abdi, E., Ali, M., Guimarães Santos, C. A., Olusola, A., & Ghorbani, M. A. (2024). Enhancing groundwater level prediction accuracy using interpolation techniques in deep learning models. Groundwater for Sustainable Development, 26,101213. https://doi.org/10.1016/j.gsd.2024.101213

Alfarrah, N., & Walraevens, K. (2018). Groundwater overexploitation and seawater intrusion in coastal areas of arid and semi-arid regions. Water, 10(2), 143. https://doi.org/10.3390/w10020143

Afzaal, H., Farooque, A.A., Abbas, F., Acharya, B., & Esau, T. (2020). Groundwater estimation from major physical hydrology components using artificial neural networks and deep learning. Water, 12 (1). 35- 49. https://doi.org/10.3390 /w12010005

Alizamir, M., Kisi, O., & Zounemat-Kermani, M. (2018). Modelling long-term groundwater fluctuations by extreme learning machine using hydroclimatic data. Hydrology Science Journal. 63 (1), 63–73. https://doi.org/10.1080/02626667.2017.1410891

Bowes, B.D., Sadler, J.M., Morsy, M.M., Behl, M., & Goodall, J.L. (2019). Forecasting groundwater table in a flood prone coastal city with long short-term memory and recurrent neural networks. Water, 11 (5). 1098. https://doi.org/10.3390/w11051098

Barzegar, R., Fijani, E., Moghaddam, A.A., & Tziritis, E. (2017). Forecasting of groundwater level fluctuations using ensemble hybrid multi-wavelet neural network-based models. Science of the Total Environment, 599, 20–31. https://doi.org/10.1016/j.scitotenv.2017.04.189

Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785-794. https://doi.org/10.1145/2939672.2939785

Chen, Y., Chen, W., Chandra P. S.,Saha, A., Chowdhuri, I., Adeli, B., Janizadeh, S., Dineva, A. A., Wang, X., & Mosavi, A. (2021). Evaluation efficiency of hybrid deep learning algorithms with neural network decision tree and boosting methods for predicting groundwater potential. Geocarto International journal, 37(19), 5564–5584. https://doi.org/10.1080/10106049.2021.1920635

Franses, P. H., & Van Dijk, D. (2000). Non-linear time series models in empirical finance. Cambridge, United Kingdom: Cambridge University Press. https://doi.org/10.1017/CBO9780511754067

Gong, Y., Zhang, Y., Lan, S., & Wang, H. (2016). A comparative study of artificial neural networks, support vector machines and adaptive neuro fuzzy inference system for forecasting groundwater levels near Lake Okeechobee, Florida. Water resources management, 30(1), 375–391. https://doi.org/10.1007/s11269-015-1167-8

Ghobadi, F., & Kang, D. (2023). Application of machine learning in water resources management: a systematic literature review. Water 15(4), 620. https://doi.org/10.3390/w1504062

Ghadami, G., & Poosti, M. (2020). Geochemistry, mineralogy and petrogenesis of Rudan Ultramafics North(Groom mountain), Hormozgan province. Scientific Quarterly Journal of Geosciences, 29(115), 313-324. https://doi.org/10.22071/gsj.2018.111764.1354 (In Persian)

Planning and Budget Organization. (2019). Hormozgan Management and Planning Organization Land planning, supervision and planning affairs. Center for Development Research and Foresight. (pp. 54).

Haas.J., Metz. M., & Neteler, M. (2021). MODIS NDVI and EVI, 16-day time series for Europe at 1 km resolution. https://doi.org/10.5281/zenodo.657385

Huang, X., Gao, L., Crosbie, R.S., Zhang, N., Fu, G., & Doble, R. (2019). Groundwater recharge prediction using linear regression, multi-layer perception network, and deep learning. Water, 11(9), 1879. https://doi.org/10.3390/w11091879

Jena, M., & Dehuri, S. (2020). Decision tree for classiﬁcation and regression: A state-of-the art review. Informatica Journal, 44(4). https://doi.org/10.31449/inf.v44i4.3023

Kochhar, A., Singh, H., Sahoo, S., Litoria, P.K., & Pateriya, B. (2022). Prediction and forecast of pre-monsoon and post-monsoon groundwater level: Using deep learning and statistical modelling. Modeling Earth Systems and Environment, 8(2), 2317–2329. https://doi.org/10.1007/s40808-021-01235-z

Kadra, A., Lindauer, M., Hutter, F., & Grabocka, J. (2021). Well-tuned Simple Nets Excel on Tabular Datasets. In M. Ranzato, A. Beygelzimer, Y. Dauphin, P. S. Liang, & J. W. Vaughan (Eds.). Advances in Neural Information Processing Systems. 34, pp. 23928–23941. Curran Associates, Inc. https://proceedings.neurips.cc/paper_files/paper/2021/file/c902b497eb972281fb5b4e206db38ee6-Paper.pdf

Kim,G.B. (2020). A study on the establishment of groundwater protection area around a saline waterway by combining artificial neural network and GIS-based AHP. Environmental Earth Sciences Journal, 79 (5), 1–17. https://doi.org/10.1007/s12665-020-8862-3

Khan, J., Lee, E., Balobaid, A.S., & Kim, K. (2023). A comprehensive review of conventional, machine leaning, and deep learning models for groundwater level (GWL) forecasting. Applied Sciences, 13, 2743. https://doi.org/10.3390/app13042743

Lo, W.C., Borja, R.I., Deng, J.H., & Lee, J.W. (2020). Analytical solution of soil deformation and fluid pressure change for a two-layer system with an upper unsaturated soil and a lower saturated soil under external loading. Journal of Hydrology, 588, 124997. https://doi.org/10.1016/j.jhydrol.2020.124997

Manna, T., & Anitha, A. (2023). Deep ensemble-based approach using randomized low-rank approximation for sustainable groundwater level prediction. Applied Sciences, 13(5), 3210. https://doi.org/ 10.3390 /app13053210

Malakar, P., Mukherjee, A., Bhanja, S. N., Ray, R. K., Sarkar, S., & Zahid, A. (2021). Machine-learning-based regional-scale groundwater level prediction using GRACE. Hydrogeology Journal, 29, 1027–1042. https://doi.org/10.1007/s10040-021-02306-2

Michael, W.J., Minsker, B.S., Tcheng, D., Valocchi, A.J., & Quinn, J.J. (2005). Integrating data sources to improve hydraulic head predictions: A hierarchical machine learning approach. Water Resources Research, 41, 1–14. https://doi.org/10.1029/2003WR002802

Miotto, R., Wang, F., Wang, S. Jiang, X., & Dudley, J.T. (2018). Deep learning for healthcare: Review, opportunities and challenges. Brief. Bioinform, 19(6), 1236–1246. https://doi.org/10.1093/bib/bbx044

Nguyen, P.T., Ha, D.H., Jaafari, Nguyen, H.D., Van Phong, T., Al-Ansari, N., Prakash, I., Le, H.V., & Pham, B.T. (2020). Groundwater potential mapping combining artificial neural network and real adaboost ensemble technique: the daknong province case-study, Vietnam. International Journal of Environmental Research and Public Health, 17 (7), 2473. https://doi.org/10.3390/ijerph17072473

Nadiri, A.A., Chitsazan, N., Tsai, F.T.C., & Moghaddam, A.A. (2014). Bayesian artificial intelligence model averaging for hydraulic conductivity estimation. Journal of Hydrologic Engineering, 19, 520–532. https://doi.org/10.1061/(asce)he.1943-5584.0000824

Paroon, S., Yavari, G., & Rezazadeh, M. (2019). Climate classification of Hormozgan province using classical methods. Geography (Regional Planning), 8(33), 115-127. (In Persian)

Priyan, K. (2021). Issues and challenges of groundwater and surface water management in semi-arid regions. Groundwater Resources Development and Planning in the Semi-Arid Region, 1–17. https://doi.org/10.1007/978-3-030-68124-11

Pham, B.T., Jaafari, A. Prakash, I., Singh, S.K., Quoc, N.K., & Bui, D.T. (2019). Hybrid computational intelligence models for groundwater potential mapping. Catena Journal, 182, 104101. https://doi.org/10.1016/j.catena.2019.104101

Pham, Q. B., Kumar, M., Nunno, F. D., Elbeltagi, A., Granata, F., Towfiqul Islam, A. R. M. d., Talukdar, S., Nguyen, X. C., Ahmed, A.N., & Duong Tran Anh, D. (2022). Groundwater level prediction using machine learning algorithms in a drought-prone area. Neural Computing and Applications, 34, 10751–10773. https://doi.org/10.1007/s00521-022-07009-7

Rahmati, O., & Melesse, A.M. (2016). Application of Dempster–Shafer theory, spatial analysis and remote sensing for groundwater potentiality and nitrate pollution analysis in the semi-arid region of Khuzestan, Iran. Science of Total Environment, 568, 1110–1123 https://doi.org/10.1016/j.scitotenv.2016.06.176

Rajaee, T., Ebrahimi, H. & Nourani, V. (2019). A review of the artificial intelligence methods in groundwater level modeling. Journal of hydrology, 572, 336–351. https://doi.org/10.1016/j.jhydrol.2018.12.037

Rathinasamy, M., Khosa, R., Adamowski, J., Ch, S., Partheepan, G., & Anand, J. (2014). Wavelet-based multiscale performance analysis: an approach to assess and improve hydrological models. Water Resources Research, 50, 9721–9737. https://doi.org/10.1002/2013wr014650

Radhakrishnan, N., & Pillai, A. (2020). Comparison of Water Quality Classification Models using Machine Learning. 5th International Conference on Communication and Electronics Systems (ICCES), 1183–1188. https://doi.org/10.1109/ICCES48766 .2020.9137903

Sahoo, A. K., Pradhan, C., Barik, R. K., & Dubey, H. (2019). DeepReco: Deep Learning Based Health Recommender System Using Collaborative Filtering. Computation, 7(2), 25. https://doi.org/10.3390/computation7020025

Sun, J., Hu, L., Li, D., Sun, K., & Yang, Z. (2022). Data-driven models for accurate groundwater level prediction and their practical significance in groundwater management. Journal of Hydrology, 608, 127630. https://doi.org/10.1016/j.jhydrol.2022.127630

Singh, S. K., Shirzadi, A., & Pham, B. T. (2021). Application of artificial intelligence in predicting groundwater contaminants. Water Pollut. International Journal of Management Practice, 71–105. https://doi.org/10.1007/978-981-15-8358-2_4

Singh, A., Patel, S., Bhadani, V., Kumar, V., & Gaurav, K. (2024). AutoML-GWL: automated machine learning model for the prediction of groundwater level. Engineering Applications of Artificial Intelligence, 127, 107405. https://doi.org/10.1016/j.engappai.2023.107405

Salamat, A., Ardestani, M., & Malekmohammadi, B. (2023). Modeling Groundwater Potential Using Machine Learning Models. Watershed Management Research, 36(4), 114-132. https://doi.org/10.22092/wmrj.2023.360675.1506 (In Persian)

Tao, H., Hameed, M . M., Marhoon, H. A., Zounemat-Kermani, M., Heddam, S., Kim, S., Sulaiman, S. O., Tan, M. L., Sa’adi, Z., Danandeh Mehr, A., Allawi, M .F., Abba, S.I., Mohamad Zain, M., Falah, M.V., Jamei, M., Bokde, N. D., Bayatvarkeshi, M., Al-Mukhtar, M., Bhagat, S. K., Tiyasha, T., Khedher, K. M., Al-Ansari, N., Shahid, S., & Yaseen, Z. M. (2022b). Groundwater level prediction using machine learning models: a comprehensive review. Neurocomputing Journal, 489, 271–308. https://doi.org/10.1016/j.neucom.2022.03.014

Tran, D.H., & Wang, S.J. (2020). Land subsidence due to groundwater extraction and tectonic activity in Pingtung Plain, Taiwan. Proceedings of the International Association of Hydrological Sciences, 382, 361–365. https://doi.org/10.5194/piahs-382-361-2020

Wu, T.H., Chen, P.Y., Chen, C.C., Chung, M.J., Ye, Z.K., & Li, M.H. (2024). Classification and Regression Tree (CART)-based estimation of soil water content based on meteorological inputs and explorations of hydrodynamics behind. Agricultural Water Management, 299, 108869. https://doi.org/10.1016/J.AGWAT.2024.108869

Yonghong, S. (2013). Response of NDVI to groundwater level change in the lower reaches of the Heihe River, China. Journal of Desert Research, 33(2), 574-582. https://doi.org/10.7522/j.issn.1000-694X.2013.00079

Zhao, R., Yan, R., Chen, Z., Mao, K., Wang, P., & Gao, R.X. (2019). Deep learning and its applications to machine health monitoring. Mechanical Systems and Signal Processing, 115, 213–237. https://doi.org/10.1016/j.ymssp.2018.05.050

Zhang, J., Zhang, X., Niu, J., Hu, B. X., Soltanian, M. R., Qiu, H., & Yang, L.( 2019). Prediction of groundwater level in seashore reclaimed land using wavelet and artificial neural network-based hybrid model, Journal of Hydrology, 577, 123948. https://doi.org/10.1016/j.jhydrol.2019.123948

Zhou, Z., Zhao, L., Lin, A., Qin,W., Lu, Y., Li, J., Zhong, Y., & He, L. (2021). Exploring the potential of deep factorization machine and various gradient boosting models in modeling daily reference evapotranspiration in China. Arabian Journal of Geosciences., 13, 1287. https://doi.org/10.1007 /s12517-020-06293-8

Zarei, M., Zandi, R., & Naemitabar, M. (2022). Assessment of flood occurrence potential using data mining models of support vector machine, Chaid and Random Forest (Case study: Frizi watershed). Journal of Watershed Management Research, 13(25), 133-144. https://doi.org/10.52547/jwmr.13.25.133 (In Persian)